WO2012122602A1

WO2012122602A1 - Lipidomic method for assessing diabetes, pre-diabetes and obesity

Info

Publication number: WO2012122602A1
Application number: PCT/AU2012/000271
Authority: WO
Inventors: Peter John Meikle; Dianna MAGLIANO; Benjamin Goudey; Jonathan Shaw; Justin Bedo; Jeremy Jowett; Izhak Haviv
Original assignee: Baker Idi Heart And Diabetes Institute Holdings Limited; Dairy Innovation Australia Limited
Priority date: 2011-03-15
Filing date: 2012-03-15
Publication date: 2012-09-20

Abstract

Disclosed are diagnostic and prognostic assays for diabetes. The assays employ lipid profiling for determining the susceptibility of a subject to developing diabetes, and in some embodiments, independent of other risk factors such as whether or not the subject is obese, has normal glucose tolerance (NGT) or is pre-diabetic. The assays of the present invention are also useful in the stratification of a subject with respect to diabetes, pre-diabetes, and obesity.

Description

Lipidomic Method for assessing Diabetes, Pre-Diabetes and Obesity

FIELD

[0001] The present invention relates generally to the field of diagnostic and prognostic assays for diabetes. In one particular aspect, the present invention provides assays employing lipid profiling for determining the susceptibility of a subject to developing diabetes, and in some embodiments, independent of other risk factors such as whether or not the subject is obese, has normal glucose tolerance (NGT) or is pre-diabetic. The assays of the present invention are also useful in the stratification of a subject with respect to diabetes, pre-diabetes, and obesity.

BACKGROUND

[0002] Bibliographic details of references provided in the subject specification are listed at the end of the specification.

[0003] Reference to any prior art is not, and should not be taken as an acknowledgment or any form of suggestion that this prior art forms part of the common general knowledge in any country.

[0004] The Australian population and indeed many nations in the developed world are in the midst of an obesity epidemic. Associated with this is a dramatic increase in the prevalence of type II diabetes and its precursor, impaired glucose tolerance (IGT). If the current pattern continues in Australia, the prevalence of diabetes is projected to rise from 7.6% in 2000 to 11.4% by 2025. More than a third of individuals will develop diabetes within their lifetime and there will be an additional 1 million Australians with diabetes by the year 2025.

[0005] Early detection of diabetes or the identification of those at increased risk of developing diabetes provides the opportunity for early intervention to prevent progression of the disease or to prevent onset. Obesity, particularly abdominal obesity, is an important risk factor for type II diabetes. However, not all obese individuals will go on to develop diabetes and improved predictive markers are required to identify those who are likely to develop diabetes.

[0006] A key unmet medical need in the prevention and management of type II diabetes is the ability to differentiate between those who are at increased risk versus those at low risk of disease development. The ability to make a well-informed assessment of risk would allow a clinician the option of commencing intervention strategies prior to disease onset or progression. Since.the clinical signs of type II diabetes are initially modest and progress slowly, it may take between 4 and 7 years for an individual to be diagnosed with type II diabetes. During this undiagnosed stage hyperglycemia has been shown to cause organ damage with increased morbidity and mortality.

[0007] Measurement of blood glucose, either fasting or following an oral glucose load, is currently used for differentiating between normal glucose tolerance (NGT), type II diabetes and the intermediate stages, impaired glucose tolerance (IGT) and impaired fasting glucose (IFG). However, once diagnosed with IFG or IGT (pre-diabetic states) it is difficult to predict whether an individual will remain in this state, return to NGT or progress to type Π diabetes. Thus, identification of EFG or IGT alone is not sufficient to predict the onset of type Π diabetes.

[0008] Existing risk calculators rely on a diverse range of measures including age, sex, family history, lifestyle, body mass index (BMI) and plasma lipids (total cholesterol, high density lipoprotein cholesterol (HDL-C) and triglycerides) and measures of blood glucose. In a recent study by Chen et al, Medical Journal of Australia 192(4): 197-202, 2010, the self-completed risk tool (AUSDRISK) was shown to have 'optimal' combination of 74% sensitivity and 68% specificity. While these are promising statistics, additional measures are required to improve their prognostic efficiency; this is further evidenced by ongoing vigorous debate in the field as to whether regular screening of at-risk individuals with current tests would have significant population health benefits in relation to their cost. In support of early intervention, it has been demonstrated mat both lifestyle and pharmaceutical intervention in the "pre-diabetic" state confers reduced risk of progression.to type H diabetes.

[0009] There is a need for assays that identify subjects at risk for developing diabetes.

SUMMARY

[0010] The present invention applies a lipid profiling approach to identify lipid classes, species and profiles associated with one or more of the various stages of diabetes through normal glucose tolerance (i.e., normoglycemic), possible obesity, pre-diabetes and diabetes. In one embodiment, multivariate models have been developed using these data, the models comprising different numbers and combinations of lipid analytes which are able to accurately discriminate between diabetes and non-diabetes better than traditional risk factors even when measures of total cholesterol, triglycerides and HDL-cholesterol are included in the traditional risk factors. Thus, individual lipid analytes according to the present invention are not lipoproteins (e.g., LDL and HDL) and do not represent broad lipid genera or classes (e.g., total triglycerides and total cholesterol) that are the subject of conventional assays for assessing traditional risk factors. Instead, they represent detectably distinct lipids or lipid molecular species, which provide significant diagnostic and/or prognostic potential, as described hereafter.

[0011] The Tables and Figures herein provide representative examples of lipid analyte levels that are differentially associated with diabetes and non-diabetes, diabetes and pre-diabetes, obesity and non-obesity, and NGT (normal glucose tolerant) and diabetes or pre-diabetes. This approach has been further used to develop and validate predictive assays, models, diagnostic rules and algorithms for determining risk of diabetes onset (incident diabetes). These assays and models have been developed from a longitudinal study of subjects, a proportion of whom developed diabetes over a five-year follow-up period, facilitating the identification of test subjects who are susceptible to developing diabetes. In some embodiments, the susceptibility of a test subject to developing diabetes is independent of other risk factors including, for example, whether the test subject is NGT, obese or pre-diabetic.

[0012] A list of Abbreviations used in this specification is set out in Table 14.

[0013] The identification of susceptible test subjects according to the present invention should facilitate the delivery of critical early intervention strategies. The lipid profiles are also instructive as to the stage prior to diabetes or the effectiveness of treatment, making them useful tools in pharmacotranslational studies and in the clinical management of patients. Thus, the present invention provides panels of lipid analytes for use in incident diabetes prediction and stage analysis.

[0014] The lipid profiling approach uses one or more of three groups of lipid analytes:

(i) one or more (e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) modified lipid analytes listed in Table 10; (ii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) non- modified lipid analytes listed in Table 10; and/or

(iii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10.

[0015] Accordingly, in one embodiment, the present invention provides an assay to stratify a test subject as susceptible or non-susceptible with respect to developing type II diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from the list consisting of:

(i) one or more (e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) modified lipid analytes listed in Table 10; .

(ii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) non- modified lipid analytes listed in Table 10; and/or

(iii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a ηόη-modified lipid analyte listed in Table 10;

and comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject, wherein the level of an individual lipid analyte listed in Table 10 is different between susceptible control subjects and non-susceptible control subjects and wherein the level of the lipid analytes in the test subject relative to a control identifies the subject as being susceptible or non-susceptible.

[0016] Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject can be carried out in different ways.

[0017] In an illustrative example of this type, the assays comprise comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a susceptible subject and a non- susceptible subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the non-susceptible control subject identifies the subject as being non- susceptible, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the susceptible control subject identifies the subject as being susceptible.

[0018] For some lipid analytes listed in Table 10, whose levels differ between susceptible or non-susceptible control subjects, a higher level is observed in the susceptible control subjects as compared to the non-susceptible control subjects. This relationship provides an odds ratio of >1 as set out in Table 10. Where a lower lipid analyte level is observed in susceptible control subjects as compared to the non-susceptible control subjects, this is represented in Table 10 as an odds ratio that is <1. The greater the difference between the odds ratio and 1 (whether less than 1 or more than 1) the greater the probability that difference between the observed levels in two groups under comparison is not due to chance alone.

(0019] In a further illustrative example, the level of a lipid analyte is higher in susceptible as compared to non-susceptible control subjects (it is said to be positively associated with susceptibility) and susceptibility is determined by detecting in a test subject a susceptibility- associated higher level of the lipid analyte as compared to the level of the same lipid analyte in a non-susceptible control subject. Conversely, the same lipid analyte is negatively associated with non-susceptibility, and non-susceptibility may be determined by detecting in a test subject a non- susceptibility-associated lower level of the lipid analyte as compared to the level of the same lipid analyte in a susceptible control subject.

[0020] In another illustrative example, the level of a lipid analyte is lower in susceptible compared to non-susceptible control subjects (it is said to be negatively associated with susceptibility) and susceptibility is determined by detecting in a test subject a susceptibility- associated lower level of the lipid analyte as compared to the level of the same lipid analyte in a non-susceptible control subject. Conversely, the same lipid analyte is positively associated with non-susceptibility, and non-susceptibility may be determined by detecting in a test subject a non- susceptibility-associated higher level of the lipid analyte as compared to the level of the same lipid analyte in a susceptible control subject.

[0021] In some embodiments, the selection of lipid analytes may conveniently be based on the smallest number of lipid analytes required to accurately differentiate between susceptible and non-susceptible test subjects. In some embodiments, differentiation is facilitated by selecting a combination of at least two lipid analytes listed in the Tables herein, whose levels associate or correlate with susceptibility or non-susceptibility to incident diabetes and wherein at least one lipid analyte is positively associated and at least one lipid analyte is negatively associated with susceptibility, or non-susceptibility, to incident diabetes.

[0022] The lipid analyte profiles are also instructive as to the stage of diabetes or the effectiveness of treatment, making them useful tools in pharmacotranslational studies and in the clinical management of patients.

[0023] In some embodiments, lipid profiles are selected to comprise one or more lipid analytes whose levels discriminate between susceptibility and non-susceptibility and/or one or more lipid analytes whose levels discriminate a test subject as NGT, obese, pre-diabetic or diabetic. [0024] In some embodiments, lipid profiles are selected to comprise one or more lipid analytes whose levels discriminate between susceptibility and non-susceptibility and/or one or more lipid analytes whose levels discriminate a test subject as non-diabetic (e.g., NGT or normoglycemic or pre-diabetic) or diabetic.

[0025] In particular embodiments, as illustrated in Table 6, lipid analytes were identified that exhibit different lipid analyte levels between non-obese and obese control subjects, NGT and pre-diabetic control subjects and NGT and diabetic control subjects. By ranking lipid analytes according to whether their comparative levels provide an odds ratio that is greater or less than 1 for any pair of control subjects, and in order of the extent of the difference in the odds ratio from 1, lipid analytes are identified that are differentially associated with NGT, diabetes, prediabetes or obesity.

[0026] Accordingly, in another embodiment, the present invention provides an assay to stratify a test subject as diabetic, obese, pre-diabetic or NGT, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from the list consisting of:

(i) one or more (e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1, 12, 13, 14, 15, 16) modified lipid analytes listed in Table 10;

(iii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10;

and comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject, wherein the level of an individual lipid analyte listed in Table 10 is different between diabetic, obese, pre-diabetic and NGT subjects and wherein the level of the lipid analytes in the subject relative to a control identifies the test subject as being at least one of diabetic, obese, pre-diabetic and NGT.

[0027] In an illustrative embodiment, the present invention provides an assay to stratify a test subject with respect to NGT, diabetes or pre-diabetes. In some embodiments, the assay comprises determining in a test subject the levels of at least two lipid analytes selected from dhCer 16:0, DHC 24: 1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.608.4/4.0, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 wherein the levels of the individual lipid analytes are different between diabetic and pre- diabetic control subjects, and wherein the level of the lipid analytes in the test subject compared to the respective levels of the same lipid analyte in a control subject selected from a diabetic control subject and a NGT control subject provides an indication that the test subject is NGT or diabetic and not pre-diabetic. (See Figure 7B)

[0028] In an illustrative embodiment of this type, the levels of dhCer 16:0, ModPC 608.4/4.0, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 are higher in diabetic control subjects than NGT control subjects and diabetes is detected by determining in a test subject a diabetes-associated higher level of one or more of dhCer 16:0 ModPC 608.4/4.0, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 as compared to the level of the same lipid analyte in a NGT control subject. Pre-diabetic subjects do not exhibit the same increased levels of these lipid analytes and therefore the assay is able to differentiate between subjects that are diabetic and those that are pre-diabetic. Conversely, in another illustrative embodiment, the levels of dhCer 16:0, ModPC 608.4/4.0, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 are lower in NGT control subjects than diabetic control subjects and NGT is detected by determining in a test subject an NGT-associated lower level of one or more of dhCer 16:0, ModPC 608.4/4.0, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 as compared to the level of the same lipid analyte in a diabetic control subject.

[0029] For lipid analyte DHC 24:1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.633.4/4.6, mod PC.773.6/6.5, and mod PC 827.7/6.8, these lipid analytes are present at a lower level in diabetic control subjects than NGT control subjects and in some embodiments, diabetes is identified (indicated) by detecting in a test subject a diabetes-associated lower level of one or more of DHC 24:1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.633.4/4.6, mod PC.773.6/6.5, and mod PC 827.7/6.8 as compared to the individual level(s) of the same lipid analyte(s) in a NGT control subject. Pre-diabetic subjects do not exhibit the same lower levels of these lipid analytes and therefore the assay is able to differentiate between test subjects that are diabetic and those that are pre-diabetic.

[0030] In some embodiments, diagnosis is facilitated by selecting a combination of at least two lipid analytes listed in the Tables herein wherein at least one lipid analyte is positively associated and at least one lipid analyte is negatively associated, with diabetes.

[0031] In another embodiment, the invention provides an assay to stratify a test subject with respect to diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 883.8/7.8, APC 34:1, COH, TG 14:0 16:0 18:2, TG 14:0 16:1 18: 1, TG 14:0 16:1 18:2, TG 14:0 18:2 18:2, TG 14:1 16:0 18: 1, TG 14:1 16: 1 18:0, TG 14:1 18:0 18:2, TG 14: 1 18:1 18: 1, TG 15:0 18:1 18: 1, TG 16:0 16: 1 18:1, TG 16:1 16:1 16: 1, TG 16:1 16:1 18:0, TG 16:1 16: 1 18:1, TG 16:1 18:1 18: 1, TG 16:1 18: 1 18:2, TG 17:0 18: 1 16: 1, and TG 17:0 18:2 16:0 wherein the individual levels of the lipid analytes are different between diabetic control subjects and pre-diabetic control subjects, and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is pre-diabetic and not diabetic. (Figure 7A).

[0032] In an illustrative embodiment, lipid analyte APC 34: 1 is present at a lower level in pre-diabetic control subjects than in NGT control subjects and the observation of a lower level of APC 34: 1 in a test subject compared to the level in an NGT control subject indicates that the test subject is pre-diabetic. Because a lower level of APC 34:1 is only weakly associated with diabetes, the lower level of APC 34:1 provides an indication that the test subject is pre-diabetic and not diabetic. In some embodiments, analysis of APC 34:1 is combined with the analysis of one or more of modCer 883.8/7.8, COH, TG 14:0 16:0 18:2, TG 14:0 16:1 18: 1, TG 14:0 16: 1 18:2, TG 14:0 18:2 18:2, TG 14:1 16:0 18:1, TG 14:1 16: 1 18:0, TG 14:1 18:0 18:2, TG 14: 1 18:1 18: 1, TG 15:0 18:1 18: 1, TG 16:0 16: 1 18: 1, TG 16:1 16:1 16:1, TG 16: 1 16:1 18:0, TG 16: 1 16:1 18: 1, TG 16: 1 18:1 18:1, TG 16:1 18:1 18:2, TG 17:0 18: 1 16: 1, and TG 17:0 18:2 16:0 which are present at higher levels in pre-diabetic control subjects compared to NGT control subjects and which are only weakly associated with diabetes.

[0033] In one embodiment, the invention is directed to an assay to stratify a test subject with respect to obesity or diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from THC 18:0, THC 24:0, THC 24:1, GM3 24:0, GM3 24:1, PC 32:2, PC 36:3, PC 39:7 PC 40:7, APC 32:1, APC 38:6, LPC 20:0, LPC 20:1, LPC 22:6, modPC.843.6/7.2, modPC.866.6/7.2, modPC.877.6/6.0, PE 32:2, CE 22:2, and TG 17:0 18:1 14:0 wherein the levels of the individual lipid analytes are different between obese control subjects and diabetic control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is obese independent of diabetes. (Figure 5B)

[0034] In an illustrative embodiment, lipid analytes PE 32:2, TG 17:0 18:1 14:0, PC 32:2 and PC 36:3 are present at higher levels in obese control subjects compared to non-obese control subjects and the assay comprises detecting in a test subject an obesity-associated higher level of one or more of these analytes compared to the respective level of the same lipid analyte(s) in a non-obese control subject. Lipid analyte levels of PE 32:2, TG 17:0 18: 1 14:0, PC 32:2 and PC 36:3 are weakly associated with diabetes and accordingly higher levels are indicative of obesity independent of diabetes.

[0035] In another embodiment, the assays are directed towards stratifying a test subject with respect to obesity or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from GM3 24:0, GM3 24: 1, PC 39:7, PC 40:7, APC 32:1, APC 38:6, LPAF 18:1, LPC 18:1, LPC 20:0, LPC 20:1, LPC 20:1, LPC 20:2, LPC 22:0, LPC 22:6, modPC.636.4/3.6, modPC.664.4/4.3, modPC.843.6/7.2, modPC.877.6/6.0, modPC 879.1/6.1, and PS 36:1 wherein the levels of the individual lipid analytes are different between obese control subjects and pre-diabetic control subjects and wherein the level of the respective lipid analytes in the test subject relative to a control provides an indication that the test subject is obese independent of pre-diabetes. (Figure 6B)

[0036] In an illustrative embodiment, modPC.636.4/3.6, modPC.664.4/4.3 and PS 36:1 are positively associated with obesity and a test subject is stratified as obese by detecting an obesity-associated increase in the level of one or more of modPC.636.4/3.6, modPC.664.4/4.3 and PS 36: 1 compared to the respective levels of the same lipid analytes in a non-obese control subject. These lipid analytes are very weakly associated with pre-diabetes and accordingly, an increase in the levels of one or more of these lipid analytes in a test subject compared to an NGT control subject indicates that the test subject is obese independent of pre-diabetes.

[0037] In yet another embodiment, the assays are directed to stratifying a test subject with respect to diabetes or obesity, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 921.8/9.1, oddPC 35:2, oddPC 37:3, oddPC 37:4, PC 40:6, modPC.633.4/4.6, modPC.773.6/6.5, modPC 827.7/6.8, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:6, PS 40:6, CE 22:6, CE 24:5, CE 24:6, modCE 790.8/6.6, DG 16:0 22:6, and TG 18:1 18:1 22:6 wherein the levels of the individual lipid analytes are different between diabetic control subjects and obese control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is diabetic independent of obesity. (Figure 5A)

[0038] In an illustrative embodiment, lower levels of oddPC 35:2, oddPC 37:3, oddPC 37:4, modPC.633.4/4.6, modPC.773.6/6.5, and modPC 827.7/6.8 are observed in diabetic control subjects compared to NGT control subjects and diabetes is diagnosed by detecting a diabetes associated decrease in the level of 1 to 6 lipid analytes selected from oddPC 35:2, oddPC 37:3, oddPC 37:4, modPC.633.4/4.6, modPC.773.6/6.5, and modPC 827.7/6.8 in a test subject compared to the respective level(s) of the same lipid analytes in a NGT control subject. Conversely, higher levels of 1 to 14 lipid analytes selected from modCer 921.8/9.1, PC 40:6, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:6, PS 40:6, CE 22:6, CE 24:5, CE 24:6, modCE 790.8/6.6, DG 16:0 22:6, and TG 18:1 18:1 22:6 in a test subject compared to the respective level(s) of the same lipid analyte in a NGT control subject indicates that the test subject is diabetic. As these lipid analytes are only weakly associated with obesity a higher level of one or more of these lipid analytes in a test subject compared to the level in a non-obese control subject indicates that the test subject is diabetic independent of obesity.

[0039] In still yet another embodiment, a test subject is stratified with respect to prediabetes or obesity, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 576.5/7.7, modCer 883.8/7.8, modCer 921.8/9.1, PC 40:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, COH, DG 16:0 22:5, DG 16:0 22:6, DG 18:0 18:1, DG 18: 1 18:1, TG 14: 1 18:0 18:2, TG 16:1 16:1 18: 1, TG 16:1 18: 1 18:1, TG 16:1 18:1 18:2, TG 18:1 18: 1 18:1, and TG 18: 1 1 :1 22:6 wherein the levels of the individual lipid analyte are different between obese and pre-diabetic subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the subject is pre- diabetic independent of obesity. (Figure 6A)

[0040] These lipid analytes are all observed at higher levels in pre-diabetic control subjects compared to NGT control subjects but they are all only weakly associated with obesity. Accordingly, in an illustrative embodiment, a higher level of 1 to 20 of modCer 576.5/7.7, modCer 883.8/7.8, modCer 921.8/9.1, PC 40:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, COH, DG 16:0 22:5, DG 16:0 22:6, DG 18:0 18:1, DG 18:1 18: 1, TG 14: 1 18:0 18:2, TG 16:1 16:1 18:1, TG 16:1 18:1 18:1, TG 16: 1 18:1 18:2, TG 18: 1 18:1 18:1, and TG 18: 1 18:1 22:6 in a test subject compared to the respective level(s) in a NGT control subject indicates that the test subject is pre-diabetic independent of obesity.

[0041] In some embodiments, the assays to stratify a test subject with respect to obesity or diabetes or pre-diabetes comprise determining the levels of at least two polyunsaturated lipid analytes selected from PC 38:6, PC 40:5, PC 40:6, PC 40:7, PC 44: 12, APC 38:6, LPC 22:6, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, PS 40:6, CE 22:5, CE 22:6, DG 16:0 22:5, DG 16:0 22:6, and TG 18: 1 18: 1 22:6 wherein the levels of the individual lipid analytes are different between obese control and diabetic control or pre-diabetic control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the subject is diabetic or pre-diabetic independent of obesity. (Figure 8)

[0042] In some embodiments, the assay is used to stratify a test subject as diabetic or non-diabetic (e.g. , pre-diabetic or NGT) and the assay comprises, consists or consists essentially of determining the levels of at least two lipid analytes selected from the list consisting of:

(i) one or more (e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) modified lipid analytes listed in Table 10;

(ii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) non- modified lipid analytes listed in Table 10; and/or - π -

and comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject, wherein the level of an individual lipid analyte listed in Table 10 is different between diabetic and non-diabetic subjects and wherein the level of the lipid analytes in the subject relative to a control identifies the test subject as diabetic or non-diabetic.

[0043] In representative examples of these embodiments, a subject that is identified as NGT or pre-diabetic according to an assay as broadly described above is stratified as non-diabetic.

[0044] Presymptomatic diagnosis will facilitate prevention of .diabetes, including the use of existing medical therapies. Lipidotyping of individuals is useful for (a) identifying a subject or metabolic state that will respond to particular drugs, (b) identifying types of subject that respond well to specific medications or medication types with fewer adverse effects and (c) guide new drug discovery and testing.

[0045] Even yet another aspect of the present invention relates to a method of treatment or prophylaxis of a test subject comprising, consisting or consisting essentially of stratifying the test subject with respect to diabetes or pre-diabetes, including susceptibility to diabetes onset, by determining the levels of a lipid analyte selected from the list consisting of:

(ii) two or more (e.g., at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) non- modified lipid analytes listed in Table 10, and

wherein the level or ratio of the lipid analyte or analytes relative to a control provides a stratification of the test subject with respect to diabetes or pre-diabetes including susceptibility to develop onset and exposing the subject to therapeutic or behavioral intervention on the basis that the subject tests positive to diabetes or pre-diabetes or susceptibility to diabetes.

[0046]

[0047] The above summary is not and should not be seen in any way as an exhaustive recitation of all embodiments of the present invention. BRIEF DESCRIPTION OF THE FIGURES

[0048] Some figures contain color representations or entities. Color photographs are available from the Patentee upon request or from an appropriate Patent Office. A fee may be imposed if obtained from a Patent Office.

[0049] Figure 1 is a graphical representation of logistic regression analysis of lipid classes against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid class against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals.

[0050] Figure 2 is a graphical representation of logistic regression analysis of sphingolipids against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals.

[0051] Figure 3 is a graphical representation of logistic regression analysis of CE, LPC and PE species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals.

[0052] Figure 4 is a graphical representation of logistic regression analysis of DG and TG species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals. [0053] Figure 5 is a graphical representation of logistic regression analysis of diabetic and obesity specific lipid species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals. Lipids were then ranked by the ratio of the odds ratio (diabetes) to the odds ratio (obesity). Panel A shows the top 20 ranked lipids that are strongly associated with diabetes but weakly associated with obesity. Panel B shows the bottom 20 ranked lipid species lipids that are strongly associated with obesity but weakly associated with diabetes.

[0054] Figure 6 is a graphical representation of logistic regression analysis of pre- diabetic and obesity specific lipid species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and prediabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals. Lipids were then ranked by the ratio of the odds ratio (pre-diabetes) to the odds ratio (obesity). Panel A shows the top 20 ranked lipids that are strongly associated with pre-diabetes but weakly associated with obesity. Panel B shows the bottom 20 ranked lipid species lipids that are strongly associated with obesity but weakly associated with pre-diabetes.

[0055] Figure 7 is a graphical representation of logistic regression analysis of pre- diabetic and diabetic specific lipid species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against pre-diabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and prediabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio^for each lipid class, whiskers show the 95% confidence intervals. Lipids were then ranked by the ratio of the odds ratio (pre-diabetes) to the odds ratio (diabetes). Panel A shows the top 20 ranked lipids that are strongly associated with pre-diabetes but weakly associated with diabetes. Panel B shows the bottom 20 ranked lipid species lipids that are strongly associated with diabetes but weakly associated with pre-diabetes.

[0056] Figure 8 is a graphical representation of logistic regression analysis of polyunsaturated fatty acid containing lipid species against obesity, pre-diabetes and diabetes. Logistic regression was performed for each lipid species against diabetes (red bars), against prediabetes excluding diabetes (orange bars) and against obesity (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Regression against diabetes and pre-diabetes were also adjusted for waist while regression against obesity was also adjusted for HbAlc. Bars show the adjusted odds ratio for each lipid class, whiskers show the 95% confidence intervals. Lipids containing either Docosahexaenoic acid (DHA) or Docosapentaenoic acid (DPA) are shown in the plot.

[0057] Figure 9 is a graphical representation of logistic regression analysis of lipid classes against incident diabetes and pre-diabetes progression. Logistic regression was performed for each lipid class against incident diabetes (red bars) and against pre-diabetes progressors analyzing only the pre-diabetes at baseline group (green bars). AH analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Bars show the adjusted odds ratio for each lipid, whiskers show the 95% confidence intervals. Lipids were then ranked by the odds ratio.

[0058] Figure 10 is a graphical representation of logistic regression analysis of lipids against incident diabetes and pre-diabetes progression. Logistic regression was performed for each lipid species against incident diabetes (red bars) and against pre-diabetes progressors analyzing only the pre-diabetes at baseline group (green bars). All analyses were adjusted for age, sex, systolic blood pressure, exercise and education. Bars show the adjusted odds ratio for each lipid, whiskers show the 95% confidence intervals. Lipids were then ranked by the odds ratio. Panel A shows the top 10 ranked lipids that were positively and negatively associated with incident diabetes. Panel B shows the top 20 ranked lipids that are positively associated with pre-diabetes progression (there were no lipids negatively associated with pre-diabetes progression).

[0059] Figure 11 provides graphical representations of the relationship between lipid associations in the cross sectional and longitudinal studies. Logistic regression was performed on the longitudinal cohort for each lipid species against incident diabetes. These values were plotted against the odds ratio from the logistic regression of the cross sectional cohort for the same lipid species against diabetes (panel A), against pre-diabetes (Panel B) and against obesity determined by waist measurement (panel C). The plots show a strong positive correlation between the lipids associated with diabetes, pre-diabetes and obesity in the cross sectional study and the lipids associated with incident diabetes in the longitudinal study.

[0060] Figure 12 provides graphical representations of the performance of multivariate models to classify new diabetes. Recursive feature elimination (RFE) models containing different numbers of lipids were created to discriminate between control (non-diabetic) and diabetes at baseline in the cross sectional study. C-statistics (panel A) and % accuracy (panel B) with 95% confidence intervals for each model are plotted against the number of variables in the model. Models were; created from six risk factors (age, sex, systolic blood pressure, education, exercise, and waist) (red squares), six risk factors with total cholesterol, HDL cholesterol and triglycerides (green triangles), plasma lipids (yellow circles) and lipids with risk factors (blue diamonds).

[0061] Figure 13 provides graphical representations of the performance of multivariate models to predict incident diabetes. Recursive feature elimination (RFE) models containing different numbers of lipids were created to discriminate between control (non-diabetic) and incident diabetes at follow-up in the longitudinal study. C-statistics (panel A) and % accuracy (panel B) with 95% confidence intervals for each model are plotted against the number of variables in the model. Models were; created from the longitudinal cohort (orange circles), created in the cross sectional cohort tested in the longitudinal cohort (red squares) and created in the longitudinal cohort using the top 64 lipids identified in the cross sectional cohort (green triangles).

BRIEF DESCRIPTION OF THE TABLES

[0062] Table 1 provides baseline characteristics of the cross sectional study cohort.

[0063] Table 2 provides baseline characteristics of the longitudinal study cohort.

[0064] Table 3 provides conditions for precursor ion scan and MRM acquisition methods for lipid identification and quantification.

[0065] Table 4 provides logistic regression of total lipid classes in the cross sectional study.

[0066] Table 5 provides mean and standard deviation of total lipid classes in the cross sectional study.

[0067] Table 6 provides logistic regression of lipids against obesity, pre-diabetes and diabetes in the cross sectional study.

[0068] Table 7 provides mean and standard deviation of individual lipid species in the cross sectional study.

[0069] Table 8 provides logistic regression of total lipid classes in the longitudinal study.

[0070] Table 9 provides mean and Standard Deviation of total lipid classes in the longitudinal study.

[0071] Table 10 provides logistic regression of lipids against incident diabetes and prediabetes progressors.

[0072] Table 11 provides mean and standard deviation of individual lipid species in the longitudinal study.

[0073] Table 12 provides ranked features in the recursive feature elimination models for prediction of diabetes in the cross sectional and longitudinal studies.

[0074] Table 13 provides ROC analysis of multivariate models and other risk scores for the prediction of incident diabetes in the longitudinal study. ^

[0075] Table 14 provides a list of abbreviations used. DETAILED DESCRIPTION

[0076] Throughout this specification, unless the context requires otherwise, the words "comprise," "comprises" and "comprising" will be understood to imply the inclusion of a stated step or element or group of steps or elements but not the exclusion of any other step or element or group of steps or elements. Thus, use of the term "comprising" and the like indicates that the listed elements are required or mandatory, but that other elements are optional and may or may not be present. By "consisting of is meant including, and limited to, whatever follows the phrase "consisting of. Thus, the phrase "consisting of indicates that the listed elements are required or mandatory, and that no other elements may be present. By "consisting essentially of is meant including any elements listed after the phrase, and limited to other elements that do not interfere with or contribute to the activity or action specified in the disclosure for the listed elements. Thus, the phrase "consisting essentially of indicates that the listed elements are required or mandatory, but that other elements are optional and may or may not be present depending upon whether or not they affect the activity or action of the listed elements.

[0077] As used in the subject specification, the singular forms "a", "an" and "the" include plural aspects unless the context clearly dictates otherwise. Thus, for example, reference to "a lipid analyte" includes a single lipid analyte, as well as two or more lipid analytes; reference to "an analyte" includes a single analyte or two or more analytes; reference to "the invention" includes single and multiple aspects of the invention; and so forth.

[0078] The use of numerical values in the various ranges specified in this application, unless expressly indicated otherwise, are stated as approximations as though the minimum and maximum values within the stated ranges were both preceded by the word "about". In this manner, slight variations above and below the stated ranges can be used to achieve substantially the same results as values within the ranges. Also, the disclosure of these ranges is intended as a continuous range including every value between the minimum and maximum values. In addition, the present invention extends to ratios of two or more lipid analytes providing a numerical value associated with a level of risk of diabetes development, etc.

[0079] Each embodiment described herein is to be applied mutatis mutandis to each and every embodiment unless specifically stated otherwise.

[0080] A rapid, efficient and sensitive assay is provided for the stratification of diabetes incidence in symptomatic and asymptomatic subjects.

[0081] The term "stratification" or "stratify" includes identification, diagnosis or clarification of the stage of diabetes development from possibly obesity, no diabetes (NGT), pre- diabetes (IGT and or IFG), through to diabetes and includes determining susceptibility to diabetes. Generally, this is based on comparing a knowledge base of levels or ratios of lipid analytes in body fluid or tissue extract of a test subject to another knowledge base of predetermined levels, statistically correlated (associated) to obesity, NGT, pre-diabetes or diabetes as defined herein or to the risk of subsequent development of diabetes (incident diabetes).

[0082] As used herein, the term "diabetes" refers to a disease or condition that is generally characterized by metabolic defects in production and utilization of glucose which result in the failure to maintain appropriate blood sugar levels in the body. A subject is identified as having diabetes if the subject has a fasting blood glucose level greater than 125 mg dL, a 2 hour post-load glucose reading of greater than 200 mg/dL, or a HbAlc (glycosylated hemoglobin) level greater than or equal to 6.5%.

[0083] The term "pre-diabetes" refers to a disease or condition that is generally characterized by impaired glucose tolerance or impaired fasting glucose and which frequently precedes the onset of diabetes in a subject. A subject is identified as having pre-diabetes if the subject has a fasting blood glucose level greater than 100 mg/dL but less than or equal to 125 mg/dL, a 2 hour post-load glucose reading of greater than 140 mg/dL but less than 200 mg/dL, or a HbAlc level greater than or equal to 6.0% but less than 6.5%. Pre-diabetes extends the definition of impaired glucose tolerance to include individuals with a fasting blood glucose within the high normal range >100 mg/dL (Meigs et al, Diabetes 2003 52:1475-1484) and fasting hyperinsulinemia (elevated plasma insulin concentration).

[0084] "Impaired glucose tolerance (IGT)" and "impaired fasting glucose (IFG)" are two conventional clinical definitions of "pre-diabetes." IGT is used herein to describe a subject who, when given a glucose tolerance test, has a blood glucose level that falls between normal and hyperglycemic. Such a subject is at a higher risk of developing diabetes although they are not considered to have diabetes. For example, impaired glucose tolerance refers to a condition in which a patient has a 2-hour postprandial blood glucose or serum glucose concentration greater than 140 mg/dL (7.78 mmol/L) and less than 200 mg/dL (11.11 mmol/L).

[0085] The term "impaired fasting glucose" refers to a fasting plasma glucose of between 6.1 mmol/1 (110 mg/dL) to 6.9 mmol/1 (125 mg/dL), or a fasting glucose of between 6.1 mmol 1 (110 mg/dL) to 6.9 mmol/1 (125 mg/dL).

[0086] The condition of "hyperglycemia" (high blood sugar) is a condition in which the blood glucose level is too high. Typically, hyperglycemia occurs when the blood glucose level rises above 180 mg/dL. Symptoms of hyperglycemia include frequent urination, excessive thirst and, over a longer time span, weight loss. [0087] The term "obese" or "obesity" refers to an individual who has a body mass index (BMI) of 30 kg m² or more due to excess adipose tissue. Obesity also can be defined on the basis of body fat content: greater than 25% body fat content for a male or more than 30% body fat content for a female. A "morbidly obese" individual has a body mass index greater than 35 kg/m².

[0088] The present invention identifies a correlation between the level or ratio of particular lipid analytes in a subject and the risk of developing diabetes (incident diabetes). The term "diabetes" as used herein means subjects having a plasma-fasting glucose level of equal to or less than 7.0 mmol 1. Reference herein to a "subject" includes a human which may also be considered an individual, patient, host, recipient or target. The subject may also be an animal or an animal model.

[0089] The present invention enables, therefore, a diabetes risk profile to be determined for a test subject based on a lipidomic profile. The profiling enables early diagnosis, conformation of a clinical diagnosis, treatment monitoring and treatment selection.

[0090] The term "susceptible" or "susceptibility", as described herein, refer to the proneness of an individual towards the development of a certain state (e.g., diabetes), or towards being less able to resist a particular state than the average individual. The term encompasses both increased susceptibility and decreased susceptibility. As determined herein, there are complex changes in lipid metabolism (which may result from environmental or genetic factors) in subjects, which are associated with different stages such as pre-diabetes and diabetes and which also act as harbingers of disease onset in both NGT and pre-diabetic subjects. Thus, a susceptible subject test or control may be pre-diabetic or NGT. In other words, a diagnosis of pre-diabetes, by whatever means, is not a diagnosis of susceptibility in accordance with the present invention.

[0091] The identification of susceptible test subjects according to the present invention should facilitate the delivery of critical early intervention strategies to these subjects. The lipid profiles are also instructive as to the stage prior to diabetes or the effectiveness of treatment, making them useful tools in pharmacotranslational studies and in the clinical management of patients. Thus, the present invention provides panels of lipid analytes for use in incident diabetes prediction and stage analysis. Some of the panels are illustrated in the Description and Figures. Others are found in the Tables such as Table 10 and Table 6 (which list essentially the same lipid analytes) and would be readily identified therein by the skilled artisan. In an illustrative example, lipid analytes are selected for use in the present assays if their levels are shown to be significantly different (p < 0.01) between subject groups.

[0092] The lipid profiling approach uses one or more of three groups of lipid analytes:

(i) one or more modified lipid analytes listed in Table 10; (ii) two or more non-modified lipid analytes listed in Table 10; and/or

(iv) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10.

[0093] In some embodiments, the approach does not involve determining the levels of total cholesterol and/or total triglycerides. In other embodiments, no more than 1, 2, 3, 4, 5, 6, 7, or 8 TG lipid analytes are employed. In other embodiments, where TG lipid analyte levels are determined, the assays further comprise or consist essentially of determining a decrease in a level of one or more non-TG lipid analytes in a test subject relative to the level in a NGT control subjects.

[0094] There are many methods of lipid analysis, which may be used to detect lipid analyte levels including mass spectrometry. In an illustrative embodiment, liquid chromatography, electrospray ionization-tandem mass spectrometry is used.

[0095] In some embodiments, in order to detect a lipid analyte, a biological sample from a subject is prepared and subjected to lipid analysis.

[0096] The term "biological sample" as used herein refers to a sample that may be extracted, untreated, treated, diluted or concentrated from an animal. The biological sample may include a biological fluid such as whole blood, serum, plasma, saliva, urine, sweat, ascitic fluid, peritoneal fluid, synovial fluid, amniotic fluid, cerebrospinal fluid, tissue biopsy, and the like. In certain embodiments, the biological sample is blood, especially peripheral blood and plasma therefrom.

[0097] In some embodiments, the levels of lipid analytes are determined e.g., by comparing the individual levels of lipid analytes in a biological sample to the respective levels of the same lipid analytes in a biological sample obtained from a control subject. Reference to a "control subject" includes a single control subject and a population or cohort of control subjects.

[0098] The control level (reference level or concentration) may be expressed as a mean or mode level or a range from a cohort of subjects or a mean together with a standard deviation to determine threshold levels. Generally, levels or concentrations are determined from blood or a blood derivative such as plasma or serum and expressed as pmol per mL after comparison with internal standards.

[0099] The term "level" or "levels" encompasses absolute or relative amounts or concentrations of lipid analytes in a biological sample, including ratios of levels of lipid analytes, and odds ratios of levels or ratios of odds ratios. Lipid analyte levels in cohorts of subjects may be represented as mean levels and standard deviations as shown in the Tables and Figures herein. The term "level" includes an increase in a level or a ratio of two or more levels and a disease in a level or a ratio of two or more levels.

[0100] Reference to a "control" broadly includes data that the skilled person would use to facilitate the accurate interpretation of technical data. In an illustrative example, the level or levels of lipid analyte(s) from a subject are compared to the respective reference level or levels of ^< the same lipid analyte(s) in one or more cohorts (populations groups) of control or reference subjects whose disease status or risk is known or established. In some assays, control or reference subjects include a susceptible subject cohort wherein the subjects do not have diabetes (they may include pre-diabetic and NGT control subjects) when their lipid levels are established but subsequently developed diabetes over a follow up period. In the illustrative examples described herein the follow-up period was five years. Also, control or reference levels may be established using a non-susceptible subject cohort wherein the subjects do not have diabetes (they may include pre-diabetic and NGT control subjects) when their lipid levels were established and had no incident diabetes (that is they did not develop diabetes over a follow up period). Further, control subjects include NGT control subjects, pre-diabetic control subjects, obese or non-obese subjects. In some embodiments, the control may be the level or ratio of one or more lipid analytes in a sample from the test subject taken at an earlier time point. Thus, a temporal change in analyte levels can be used to identify susceptibility or provide a correlation as to the state of diabetes. In some embodiments, the relative levels of two or more lipid analytes provide a useful control.

[0101] In some embodiments, a control subject is a group of control subjects. The level of analytes in a control subject group may be a mean value or a preselected level, threshold or range of levels that define, characterize or distinguish a particular group. Thresholds may be selected that provide an acceptable ability to predict diagnostic or prognostic risk, treatment success, etc. In illustrative examples, receiver operating characteristic (ROC) curves are calculated by plotting the value of one or more variables versus its relative frequency in two populations (called arbitrarily "diabetes" and "normal" or "susceptible" and "pre-diabetic" groups for example). The area under the curve provides the C-statistic, which is a measure of the probability the measurement will allow correct identification of a condition or risk. For any particular lipid analyte(s) or class(es), a distribution of level(s) for subjects in two control populations will likely overlap. Under such conditions, a test level may not absolutely distinguish "diabetes" and "normal" or "susceptible" and "non-susceptible" with 100% accuracy and the area of overlap indicates where the test cannot distinguish between groups. Accordingly, in some embodiments, a threshold or range is selected, within which the test is considered to be "indicative" i.e., able to discriminate between disease status and without which the test is considered to be "non-indicative" i.e., unable to discriminate. [0102] In some embodiments, two or more lipid analytes are selected to discriminate between susceptible and non-susceptible subjects or between disease status groups (diabetic, pre- diabetic, obese or NGT) with at least about 60%, 65%, or 70% accuracy or having a C-statistic of at least about 0.60, 0.65, 0.70, or 0.75.

[0103] Alternatively, or in addition, thresholds may be established by obtaining an analyte level from the same patient, to which later results may be compared. In these embodiments, the individual in effect acts as their own "control group". For lipid analytes that increase with prognostic risk, an increase over time in the same patient can indicate a development of risk of diabetes or a failure of a treatment regimen, while a decrease over time can indicate remission of risk or success of a treatment regimen. The skilled artisan may routinely apply various further controls if required. In an illustrative example, the levels of a range or panel of lipid analytes within one or more lipid classes are determined and compared to predetermined levels in one or more control subject groups. In some embodiments, lipid analytes that are determined not to be correlated with diabetes risk of stratification can be included as internal controls and are therefore also useful.

[0104] In some embodiments, lipid analyte levels in control groups are used to generate a profile of lipid analyte levels reflecting difference between levels in two control groups. Thus, a particular lipid analyte may be more abundant or less abundant in one control group compared to another control group. The data may be represented as an overall signature score or the profile may be represented as a barcode or other graphical representation to facilitate analysis or diagnosis. The lipid analyte levels from a test subject may be represented in the same way and the similarity with the signature scope or level of "fit" to a signature barcode or other graphical representation may be determined. In other embodiments, the levels of a particular lipid analyte or lipid class are analyzed and a downward or an upward trend in analyte level determined.

[0105] Accordingly, in one embodiment, the present invention provides an assay to stratify a test subject as susceptible or non-susceptible with respect to developing type II diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from the list consisting of:

(i) one or more modified lipid analytes listed in Table 10;

(ii) two or more non-modified lipid analytes listed in Table 10; and/or

(hi) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10;

[0106] In some embodiments, the individual level of a lipid analyte in susceptible subjects is at least 101%, 102%, 103%, 104%, 105%, 106%, 107% 108%, 109%, 110%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900% or 1000% (i.e. an increased or higher level), or no more than about 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.01%, 0.001% or 0.0001% (i.e. a decreased or lower level) of the level of the same lipid analyte in non-susceptible subjects.

[0107] In an illustrative example of this type, the assays comprise comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a susceptible control subject and a non-susceptible control subject, wherein a similarity.in the respective levels of the at least two lipid analytes between the test subject and the non-susceptible control subject identifies the subject as being non-susceptible, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the susceptible control subject identifies the subject as being susceptible.

[0108] In a further illustrative example, the level of a lipid analyte is positively associated with susceptibility, and susceptibility is determined by detecting in a test subject a susceptibility-associated increase in the level of the lipid analyte as compared to the level of the same lipid analyte in a non-susceptible control subject.

[0109] In another illustrative example, the level of a lipid analyte is negatively associated with susceptibility, and susceptibility is determined by detecting in a test subject a susceptibility-associated decrease in the level of the lipid analyte as compared to the level of the same lipid analyte in a non-susceptible control subject.

[0110] In some embodiments, the test subject is non-diabetic and may have normal glucose tolerance. In other embodiments, test subject is insulin sensitive or pre-diabetic exhibiting impaired glucose tolerance and/or impaired fasting glucose. In some embodiments, the test subject is obese with or without any one or more of the foregoing conditions.

[0111] In. an illustrative example, lipid analyte levels from susceptible control subjects include levels from.NGT and/or pre-diabetic subjects who are prone to developing diabetes and lipid analyte levels from non-susceptible control subjects include levels from NGT and/or pre- diabetic subjects who are not prone to developing diabetes. [0112] Suitably, the test subject is identified as being susceptible ("high risk") to developing type II diabetes when the level of at least two lipid analytes in the subject varies from the level of the same lipid analytes in a susceptible control subject or in a susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%.

[0113] Alternatively, the test subject is identified as being non-susceptible ("low risk") to developing type II diabetes when the level of at least two lipid analytes in the subject varies from the level of the same lipid analytes in a NGT control subject or in a non-susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%.

[0114] In another illustrative embodiment, the at least two lipid analytes include 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1, 12, 13, 14, 15 or 16 lipid analytes listed in Table 10 wherein the level of an individual lipid analyte listed in Table 10 is different between susceptible control subjects and non- susceptible control subjects. In other embodiments, up to 20, 25, 30, 35, 40, 45, 50, 55, 60, 70, 80, 90, 100, 110, 120, or 130 or more lipid analytes listed in Table 10 are employed wherein the level of an individual lipid analyte listed in Table 10 is different between susceptible control subjects and non-susceptible control subjects. Suitably, internal controls may include lipid analytes such as but not limited to those listed in Table 6 or 10 whose levels are not different between susceptible control subjects and non-susceptible control subjects.

[0115] In some illustrative embodiments, assays are provided wherein the or each modified lipid analyte in (i) (above) is selected from a modified cholesterol ester (modCE), a modified ceramide (modCER) and a modified phosphatidylcholine (modPC). As shown, for example, in Figure 10, a decrease in the level of modCer 614.6/5.7 is strongly associated with incident diabetes. This lipid analyte also ranks highly in the models illustrated in Table 12. ModPC and modCE lipid analytes are also represented in Table 12, as well as Table 10.

[0116] In some embodiments, lipid analytes are selected that fall within a single lipid class. In other embodiments, the level of two or more lipid analytes in one or more lipid classes are determined and compared.

[0117] In particular embodiments, the assayed levels of lipid analytes are used in combination with one or more traditional risk factors to thereby identify the subject as being susceptible or non-susceptible.

[0118] In other embodiments, assays are provided wherein the non-modified lipid analytes in (ii) above are selected from a dihydroceramide (dhCer), a ceramide (Cer), a dihexosylceramide (DHC), a phosphatidylglycerol (PG), a phosphatidylethanolamine (PE), a phosphatidylinositol (PI), a cholesterol ester (CE), a diacylglycerol (DG) a triacylglycerol (TG), a (LPAF), a lysophosphatidylcholine (LPC), an alkylphosphatidylcholine (APC), a phosphatidylcholine (PC), a lysophosphatidylethanolamine (lysoPE), and an odd chain phosphatidylcholine (oddPC).

[0119] In an illustrative example assays are provided wherein the or each modified lipid analyte in (iii) above is selected from a modified cholesterol ester (modCE), a modified ceramide (modCER) and a modified phosphatidylcholine (modPC) and the or each non-modified lipid in (iii) is selected from a dihydroceramide (dhCer), a ceramide (Cer), a dihexosylceramide (DHC), a phosphatidylglycerol (PG), a phosphatidylethanolamine (PE), a phosphatidylinositol (PI), a cholesterol ester (CE), a diacylglycerol (DG) a triacylglycerol (TG), a (LPAF), an alkylphosphatidylcholine (APC), a lysophosphatidylcholine (LPC), a phosphatidylcholine (PC), a lysophosphatidylethanolamine (lysoPE), and an odd chain phosphatidylcholine (oddPC).

(0120] In yet another embodiment, any number of lipid analytes selected from between 1 and 24 lipid classes including 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 or 23 lipid classes are analyzed.

[0121] In some embodiments, total lipid class analysis is informative. As shown in Table 8, susceptible subjects are characterized by a decreased level of total DHC and by increased levels of total dhCer, total Cer, total PG, total PI, total CE total modCE, total DG and total TG. Levels of total COH were not significantly predictors of susceptibility to incident diabetes.

[0122] In another illustrative example, the at least two lipid analytes are selected from dhCer 18:0, dhCer 22:0, dhCer 24:0, dhCer 24:1, Cer 18:0, Cer 20:0, Cer 22:0, DHC 16:0,modCer 576.5/7.7, modCer 614.6/5.7, modCer 886.8/9.1, modCer 910.8/9.0, PC 33:0, PC 35:2, PC 39:7, PC 44:12, APC 30:0, APC 36:2, APC 34:1, APC 34:2b, APC 36:36, LPAF 22:0, LPAF 24:1, LPAF 14:0, LPAF 22: 1., LPAF 16:0, LPC 18:0, modPC.610.4/1.7, PG 16:0 18:1, PG 16:1 18:1, PG 18:0 18:1, PG 18: 1 18:1 , PE 32:0, PE 32:1, PE 34:1, PE 34:2, PE 36: 1, PE 36:2, PE 36:3, PE 36:4, PE 36:5, PE 38:1, PE 38:2, PE 38:3, PE 38:4, PE 38:5, PE 38:6, PE 40:4, PE 40:6, PE 40:5, lyso PE 16:0, lyso PE 18:0, PI 36:4, CE 14:0, CE 16:1, CE 16:2, CE 18: 1, CE 18:2, CE 18:3, CE 20:3, CE 20:4, CE 20:5, CE 22:0, CE 22:4, CE 22:5, CE 22:6, CE 24:0, modCE 558.5/7.74, modCE 588.5/7.94, modCE 682.7/8.76, modCE 790.8/6.57, G 14:0 14:0, DG 14:0 16:0, DG 14:0 18:1, DG 14:0 18:2, DG 14: 1 16:0, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18: 1, DG 16:0 18:2, DG 16:0 20:3, DG 16:0 20:4, DG 16:0 22:5, DG 16:0 22:6, DG 16:1 18:1, DG 18:0 16: 1, DG 18:0 18:0, DG 18:0 18:1, DG 18:0 18:2, DG 18:0 20:4, DG 18:1 18:1, DG 18:1 18:2, DG 18:1 18:3, DG 18:1 20:3, DG 18: 1 20:4, TG 14:0 16:0 18:2, TG 14:0 18:0 18: 1, TG 14:1 16:0 18:1, TG 14:1 16:1 18:0, TG 15:0 18:1 16:0, TG 16:0 16:0 16:0, TG 16:0 16:0 18:0, TG 16:0 16:0 18:1, TG 16:0 16:0 18:2, TG 16:0 16:1 18:1, TG 16:0 18:0 18: 1, TG 16:0 18: 1 18: 1, TG 16:0 18:1 18:2, TG 16:0 18:2 18:2, TG 16:1 16: 1 18: 1, TG 16: 1 16: 1 18:0, TG 17:0 16:0 16:1, TG 17:0 16:0 18:0, TG 17:0 18:1 14:0, TG 17:0 18: 1 16:0, TG 17:0 18: 1 18:1, TG 17:0 18:2 16:0, TG 18:0 18:0 18:0, TG 18:0 18:0 18:1, TG 18:0 18:1 18:1, TG 18:0 18:2 18:2, TG 18:1 14:0 16:0, TG 18: 1 18: 1 20:4, and TG 18:2 18:2 20:4. (",LIST A"Xsee Table 10)

[0123] The selection of lipid analytes may conveniently be based on the smallest number of lipid analytes required to differentiate between susceptible and non-susceptible test subjects. In some embodiments, differentiation is facilitated by selecting a combination of at least two lipid analytes listed in the Tables which exhibit levels associated with susceptibility to incident diabetes and wherein at least one lipid analyte is positively associated and at least one lipid analyte is negatively associated, with susceptibility to incident diabetes.

[0124] . In an illustrative example, the assay comprises determining a decrease relative to a control in the level of a lipid analyte selected from DHC 16:0, modCer 614.6/5.7, PC 33:0, PC 35:2, PC 39:7, PC 44: 12, APC 30:0, APC 36:2, APC 34: 1, APC 34:2b, APC 36:36, LPAF 22:0, LPAF 24:1, and LPAF 22: 1.

[0125] In a further illustrative example, susceptibility to incident diabetes is determined by detecting in a test subject a decrease (suitably a susceptibility-associated decrease) in the level of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, or 14 lipid analyte(s) selected from the group consisting of DHC 16:0, modCer 614.6/5.7, PC 33:0, PC 35:2, PC 39:7, PC 44: 12, APC 30:0, APC 36:2, APC 34:1, APC 34:2b, APC 36:36, LPAF 22:0, LPAF 24: 1, and LPAF 22: 1 as compared to the respective level(s) of the same lipid analyte(s) in a non-susceptible control subject or cohort of non-susceptible control subjects.

[0126] In another illustrative embodiment, the assays comprise determining an increase relative to a control in the level of a lipid analyte selected from dhCer 18:0, dhCer 22:0, dhCer 24:0, dhCer 24:1, Cer 18:0, Cer 20:0, Cer 22:0, modCer 576.5/7.7, modCer 886.8/9.1, modCer 910.8/9.0, LPAF 14:0, LPAF 16:0, LPC 18:0, modPC.610.4/1.7, PG 16:0 18: 1, PG 16: 1 18:1, PG 18:0 18: 1, PG 18:1 18:1, PE 32:0, PE 32:1, PE 34: 1, PE 34:2, PE 36:1, PE 36:2, PE 36:3, PE 36:4, PE 36:5, PE 38:1, PE 38:2, PE 38:3, PE 38:4, PE 38:5, PE 38:6, PE 40:4, PE 40:6, PE 40:5, lyso PE 16:0, lyso PE 18:0, PI 36:4, CE 14:0, CE 16:1, CE 16:2, CE 18: 1, CE 18:2, CE 18:3, CE 20:3, CE 20:4, CE 20:5, CE 22:0, CE 22:4, CE 22:5, CE 22:6, CE 24:0, modCE 558.5/7.74, modCE 588.5/7.94, modCE 682.7/8.76, modCE 790.8/6.57, DG 14:0 14:0, DG 14:0 16:0, DG 14:0 18:1, DG 14:0 18:2, DG 14: 1 16:0, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18: 1, DG 16:0 18:2, DG 16:0 20:3, DG 16:0 20:4, DG 16:0 22:5, DG 16:0 22:6, DG 16: 1 18: 1, DG 18:0 16: 1, DG 18:0 18:0, DG 18:0 18: 1, DG 18:0 18:2, DG 18:0 20:4, DG 18:1 18:1, DG 18:1 18:2, DG 18:1 18:3, DG 18:1 20:3, DG 18:1 20:4, TG 14:0 16:0 18:2, TG 14:0 18:0 18:1, TG 14: 1 16:0 18:1, TG 14: 1 16: 1 18:0, TG 15:0 18:1 16:0, TG 16:0 16:0 16:0, TG 16:0 16:0 18:0, TG 16:0 16:0 18:1, TG 16:0 16:0 18:2, TG 16:0 16: 1 18:1, TG 16:0 18:0 18: 1, TG 16:0 18:1 18:1, TG 16:0 18:1 18:2, TG 16:0 18:2 18:2, TG 16:1 16: 1 18:1, TG 16:1 16: 1 18:0, TG 17:0 16:0 16: 1, TG 17:0 16:0 18:0, TG 17:0 18: 1 14:0, TG 17:0 18:1 16:0, TG 17:0 18:1 18:1, TG 17:0 18:2 16:0, TG 18:0 18:0 18:0, TG 18:0 18:0 18:1, TG 18:0 18: 1 18: 1, TG 18:0 18:2 18:2, TG 18: 1 14:0 16:0, TG 18:1 18:1 20:4, and TG 18:2 18:2 20:4. (see Table 10)

[0127] In another illustrative example, susceptibility to incident diabetes is determined by detecting in a test subject a increase (suitably a susceptibility-associated increase) in the level of 1 or at least 2 to at least 108, and all integers in between, lipid analyte(s) selected from the group consisting of dhCer 18:0, dhCer 22:0, dhCer 24:0, dhCer 24: 1, Cer 18:0, Cer 20:0, Cer 22:0, modCer 576.5/7.7, modCer 886.8/9.1, modCer 910.8/9.0, LPAF 14:0, LPAF 16:0, LPC 18:0, modPC.610.4/1.7, PG 16:0 18:1, PG 16:1 18:1, PG 18:0 18:1, PG 18: 1 18:1, PE 32:0, PE 32:1, PE 34:1, PE 34:2, PE 36:1, PE 36:2, PE 36:3, PE 36:4, PE 36:5, PE 38: 1, PE 38:2, PE 38:3, PE 38:4, PE 38:5, PE 38:6, PE 40:4, PE 40:6, PE 40:5, lyso PE 16:0, lyso PE 18:0, PI 36:4, CE 14:0, CE 16:1, CE 16:2, CE 18:1, CE 18:2, CE 18:3, CE 20:3, CE 20:4, CE 20:5, CE 22:0, CE 22:4, CE 22:5, CE 22:6, CE 24:0, modCE 558.5/7.74, modCE 588.5/7.94, modCE 682.7/8.76, modCE 790.8/6.57, G 14:0 14:0, DG 14:0 16:0, DG 14:0 18: 1, DG 14:0 18:2, DG 14:1 16:0, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18:1, DG 16:0 18:2, DG 16:0 20:3, DG 16:0 20:4, DG 16:0 22:5, DG 16:0 22:6, DG 16:1 18:1, DG 18:0 16: 1, DG 18:0 18:0, DG 18:0 18:1, DG 18:0 18:2, DG 18:0 20:4, DG 18:1 18:1, DG 18: 1 18:2, DG 18:1 18:3, DG 18:1 20:3, DG 18:1 20:4, TG 14:0 16:0 18:2, TG 14:0 18:0 18:1, TG 14:1 16:0 18: 1, TG 14:1 16: 1 18:0, TG 15:0 18: 1 16:0, TG 16:0 16:0 16:0, TG 16:0 16:0 18:0, TG 16:0 16:0 18:1, TG 16:0 16:0 18:2, TG 16:0 16:1 18:1, TG 16:0 18:0 18: 1, TG 16:0 18:1 18:1, TG 16:0 18:1 18:2, TG 16:0 18:2 18:2, TG 16:1 16: 1 18:1, TG 16: 1 16: 1 18:0, TG 17:0 16:0 16:1, TG 17:0 16:0 18:0, TG 17:0 18: 1 14:0, TG 17:0 18: 1 16:0, TG 17:0 18: 1 18:1, TG 17:0 18:2 16:0, TG 18:0 18:0 18:0, TG 18:0 18:0 18: 1, TG 18:0 18: 1 18:1, TG 18:0 18:2 18:2, TG 18:1 14:0 16:0, TG 18:1 18: 1 20:4, and TG 18:2 18:2 20:4 as compared to the respective level(s) of the same lipid analyte(s) in a non-susceptible control subject or cohort of non-susceptible control subjects.

[0128] In one embodiment, the assays comprise determining the levels of at least two lipid analytes selected from the group consisting of modCer 614.6/5.7, PC 33:0, PC 39:7, PC 44:12, APC 34:1, APC 34:2b, APC 36:2, APC 36:3b, LPAF 22:0, LPAF 24:1, PE 40:4, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18: 1, DG 16:0 18:2, DG 16:0 20:4, DG 18:0 16:1, DG 18:0 18: 1, TG 16:0 16:0 18:1, TG 16:0 18:0 18:1, OddPC 35.2, OddPC 37.3, and OddPC 37.4. ("LIST B")(See Figure 10A)

[0129] In an illustrative example, the assay comprises determining in a test subject a decrease relative to a control in the level of a lipid analyte selected from modCer 614.6/5.7, PC 33:0, PC 39:7, PC 44:12, APC 34: 1, APC 34:2b, APC 36:2, APC 36:3b, LPAF 22:0 and LPAF 24:1.

[0130] In a further illustrative example, susceptibility to incident diabetes is determined by detecting in a test subject a decrease (suitably a susceptibility-associated decrease) in the level of 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 lipid analyte(s) selected from the group consisting of modCer 614.6/5.7, PC 33:0, PC 39:7, PC 44:12, APC 34: 1, APC 34:2b, APC 36:2, APC 36:3b, LPAF 22:0, LPAF 24: 1, as compared to the level of the same lipid analyte in a non-susceptible control subject or cohort of non-susceptible control subjects.

[0131] In another illustrative example, the assay comprises determining an increase relative to a control in the level of a lipid analyte selected from PE 40:4; DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18:1, DG 16:0 18:2, DG 16:0 20:4, DG 18:0 16:1, DG 18:0 18:1, TG 16:0 16:0 18:1, TG 16:0 18:0 18:1, OddPC 35.2, OddPC 37.3, and OddPC 37.4.

[0132] In a further illustrative example, susceptibility to incident diabetes is determined by detecting in a test subject an increase (suitably a susceptibility-associated increase) in the individual level(s) of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, or 13 lipid analyte(s) selected from the group consisting of PE 40:4, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18: 1, DG 16:0 18:2, DG 16:0 20:4, DG 18:0 16: 1, DG 18:0 18: 1, TG 16:0 16:0 18: 1, TG 16:0 18:0 18:1, OddPC 35.2, OddPC 37.3, and OddPC 37.4., as compared to the individual level(s) of the same lipid analyte(s) in a non-susceptible control subject or a cohort of non-susceptible control subjects.

[0133] In another aspect, the assays comprise determining an increase in the individual levels of at least two lipid analytes selected from PC 35:3, PC 35:4, PC 37:5, PC 38:4, APC 36:3a, APC 36:4, APC 38:4, APC 38:5, DG 16:0 18:0, DG 16:0 18:2, DG 16:0 20:0, DG 16:0 20:4, DG 16:0 22:5, DG 18:0 18:2, DG 18:0 20:4, DG 18: 1 20:4, TG 16:0 16:0 18:2, TG 18:0 18:2 18:2, TG 18: 1 18:1 20:4, and TG 18:2 18:2 20:4 as compared to the individual levels of the same lipid analytes in a non-susceptible control subject or cohort of non-susceptible control subjects wherein the levels are positively associated with susceptibility to pre-diabetes progression to diabetes (LIST C") (See Figure 10B).

[0134] In some embodiments, differentiation is facilitated by selecting a combination of at least two lipid analytes from LIST A and/or LIST B and/or LIST C (e.g., LIST A, or LIST B, or LIST C, or LIST A and LIST B, or LIST A and LIST C, or LIST B and LIST C, or LIST A and LIST B and LIST A), which exhibit levels associated with susceptibility to incident diabetes and wherein at least one lipid analyte is positively associated and at least one lipid analyte is negatively associated with susceptibility to incident diabetes. [0135] In some embodiments, levels of the lipid analyte(s) may be assayed alone or in combination with other lipid analytes or diabetes risk factors.

[0136] The determination of the levels of the lipid analytes enables establishment of a diagnostic rule based on the concentrations relative to , controls. In some embodiments, the diagnostic rule is based on the application of a statistical and machine learning algorithm. Such an algorithm uses relationships between lipid analyte profiles and diabetes susceptibility or non- susceptibility, obesity, pre-diabetes etc. observed in control subjects or typically cohorts of control subjects (sometimes referred to as training data, which provides control or reference levels or ratios for comparison with lipid analyte levels determined in a test subject. The data are used to infer relationships that are then used to predict the status, including susceptibility of patients with unknown status.

[0137] The present invention provides a diagnostic rule based on the application of statistical and machine learning algorithms. Practitioners skilled in the art of data analysis recognize that many different forms of inferring relationships in the training data may be used without materially changing the present invention. The data presented in the Tables herein has been used to generate illustrative minimal combinations of lipid analytes (models) that differentiate between susceptible and non-susceptible subjects using recursive feature elimination strategies using support vector machine learning. Table 12 provides illustrative lists of lipid analytes ranked according to the frequency (1 being most frequent) of their incorporation into a suitable model. These models comprising about 16 lipid analytes were able to determine susceptibility almost as well as models comprising 256 lipid analytes (see Figure 12).

[0138] In some further illustrative embodiments, the assays comprise determining in a test subject the levels of at least 8 to 32 or at least 5 to 40 or at least 10 to 20, or at least 1 to at least 64, and respectively all integers in between, lipid analytes selected from the group consisting of dhCer 18:1, CE 24:5, modPC 536.3, PI 32:1, Cer 22:0, LPAF 24:1, CE 20:5, CE 16:2, APC 32:0, CE 16:1, PC 30:2, dhCer 18:0, DG 18:2 18:2, modPC 773.6, modPC 538J and oddPC 35:4., DG 16:0 20:4, modCer 614.6, dhCer 22:0, LPC 18:2, PC 39:7, modPC 610.4, TG 18:2 18:2 20:4, PC 35:2, dhCer 24:0, modCer 886.8, APC 36:3b, PE 38:1, modCer 910.8, APC 34:2b, LPAF 22:0 and DG 16:0 18:0., CE 16:1, CE 20:4, CE 16:2, CE 24:5, dhCer 18:0, DG 16:0 18:2, LPC 15:0, LPC 18:2, CE 18:2, modCer 614.6, dhCer 18:1, LPC 18:1, modCE 588.5, PI 32:1, CE 20:5 and SM 16: 1., modCer 614.6, LPC 18:2, DG 16:0 18:1, DG 16:0 16:0, PE.36.1, PE.38.1, CE 20:4, dhCer 18:0, DG 18:0 18: 1, DG 16:0 18:2, DG 18:0 20:4, Cer 18:0, Cer 22:0, CE 20:3, SM 16: 1, and CE 24:0, and comparing individual levels of the selected lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a susceptible control subjects or a non-susceptible control subject, wherein the level of the individual lipid analyte is different between susceptible control subjects and non-susceptible control subjects and wherein the level of the lipid analytes in the test subject relative to a control identifies the subject as being susceptible or non-susceptible.

[0139] The lipid analyte profiles are also instructive as to the stage of diabetes or the effectiveness of treatment, making them useful tools in pharmacotranslational studies and in the clinical management of patients.

[0140] Accordingly, the present invention additionally provides an assay to stratify a test subject with respect to diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in a test subject selected from dhCer 16:0, DHC 24:1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.608.4/4.0, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 wherein the individual level of a lipid analyte. is associated with diabetic control subject and is not associated with pre-diabetic control subjects, and wherein the individual levels of the at least two lipid analytes in the test subject compared to the respective levels of the same lipid analytes in at least one control subject selected from a diabetic control subject or a NGT control subject provides an indication that the subject is diabetic and not pre-diabetic. (See Figure 7B). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0141] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes to corresponding reference levels associated with a control subject selected from a diabetic and a NGT subject.

[0142] In a further illustrative example, assays to stratify a subject with respect to diabetes or pre-diabetes may .comprise determining by detecting in a test subject a decrease (suitably a diabetes-associated decrease) in the level(s) of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 lipid analyte(s) selected from the group consisting of DHC 24:1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, CE 16:2, CE 20:4, CE 20:5, CE 24:5, and modCE 588.5/7.9 as compared to the level of the same lipid analyte in a NGT control subject or cohort of NGT control subjects.

[0143] In a further illustrative example, assays to stratify a subject with respect to diabetes or pre-diabetes, comprise determining (e.g., by detecting) in a test subject an increase (suitably a diabetes-associated increase) in the level of 1, 2, 3, or 4 lipid analyte(s) selected from the group consisting of dhCer 16:0, PC.608.4/4.0, PS 40:6, raodCE 682.7/8.8, as compared to the level of the same lipid anal te in a NGT control subject or cohort of NGT control subjects.

[0144] In some embodiments, the level of an individual lipid analyte in a diabetic subject is at least 101%, 102%, 103%, 104%, 105%, 106%, 107% 108%, 109%, 110%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900% or 1000% (i.e. an increased or higher level), or no more than about 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.01%, 0.001% or 0.0001% (i.e. a decreased or lower level) of the level of the same lipid analyte in a NGT subject.

[0145] Suitably, the test subject is identified as being susceptible ("diabetic") with respect to type Π diabetes when the level of at least two lipid analytes in the test subject varies from the level of the same lipid analytes in a diabetic control subject or in a susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%.

[0146] In another embodiment, the invention provides an assay to stratify a test subject with respect to diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 883.8/7.8, APC 34:1, COH, TG 14:0 16:0 18:2, TG 14:0 16:1 18:1, TG 14:0 16: 1 18:2, TG 14:0 18:2 18:2, TG 14:1 16:0 18: 1, TG 14:1 16:1 18:0, TG 14: 1 18:0 18:2, TG 14: 1 18:1 18: 1, TG 15:0 18:1 18:1, TG 16:0 16:1 18:1, TG 16: 1 16:1 16: 1, TG 16:1 16:1 18:0, TG 16:1 16:1 18: 1, TG 16:1 18: 1 18:1, TG 16:1 18: 1 18:2, TG 17:0 18:1 16:1, and TG 17:0 18:2 16:0 wherein the levels of the individual lipid analytes are different between diabetic control subjects and pre-diabetic control subjects and wherein the level of the lipid analytes in the subject relative to a control provides an indication that the subject is pre-diabetic and not diabetic. (Figure 7A). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0147] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic control subject or a NGT control subject.

[0148] In some embodiments, the level of an individual lipid analyte in a pre-diabetic subject is at least 101%, 102%, 103%, 104%, 105%, 106%, 107% 108%, 109%, 1 10%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900% or 1000% (an increased or higher level), or no more than about 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.01%, 0.001% or 0.0001% (i.e. a decreased or lower level) of the level of the same lipid analyte in a NGT subject.

[0149] Suitably, the test subject is identified as being susceptible ("pre-diabetic") with respect to type Π diabetes when the level of at least two lipid analytes in the test subject varies from the level of the same lipid analytes in a pre-diabetic control subject or in a susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%.

[0150] In one embodiment, the invention is directed to an assay to stratify a subject with respect to obesity or diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from THC 18:0, THC 24:0, THC 24: 1, GM3 24:0, GM3 24:1, PC 32:2, PC 36:3, PC 39:7 PC 40:7, APC 32:1, APC 38:6, LPC 20:0, LPC 20:1, LPC 22:6, modPC.843.6/7.2, modPC.866.6/7.2, modPC.877.6/6.0, PE 32:2, CE 22:2, and TG 17:0 18:1 14:0, wherein the levels of the individual lipid analytes are different between obese subjects and diabetic subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is obese independent of diabetes. (Figure 5B). Comparison of the individual levels of the at least two lipid analytes in the te^*st subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0151] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in a at least one control subject selected from an obese control subject or a non-obese control subject.

[0152] In some embodiments, the level of an individual lipid analyte in an obese subject is at least 101%, 102%, 103%, 104%, 105%, 106%, 107% 108%, 109%, 110%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900% or 1000% (an increased or higher level), or no more than about 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.01%, 0.001% or 0.0001% of the level of 'the same lipid analyte in a non-obese subject.

[0153] Suitably, the test subject is identified as being susceptible ("obese") with respect to, type II diabetes when the level of at least two lipid analytes in the test subject varies from the level of the same lipid analytes in a obese control subject or in a susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%. [0154] In another embodiment, the assays are directed towards stratifying a test subject with respect to obesity or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from GM3 24:0, GM3 24:1, PC 39:7, PC 40:7, APC 32: 1, APC 38:6, LPAF 18:1, LPC 18: 1, LPC 20:0, LPC 20: 1, LPC 20:1, LPC 20:2, LPC 22:0, LPC 22:6, modPC.636.4/3.6, modPC.664.4/4.3, modPC.843.6/7.2, modPC.877.6/6.0, modPC 879.1/6.1, and PS 36: 1 wherein the levels of the individual lipid analytes are different between obese control subjects and pre-diabetic test subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is obese independent of pre-diabetes. (Figure 6B). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0155] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from an obese control subject or a non-obese control subject.

[0156] In yet another embodiment, the assays are directed to stratifying a test subject with respect to obesity or diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 921.8/9.1, oddPC 35:2, oddPC 37:3, oddPC 37:4, PC 40:6, modPC.633.4/4.6, modPC.773.6/6.5, modPC 827.7/6.8, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:6, PS 40:6, CE 22:6, CE 24:5, CE 24:6, modCE 790.8/6.6, DG 16:0 22:6, and TG 18: 1 18:1 22:6 wherein the levels of the individual lipid analytes are ^" different between diabetic control subjects and obese control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is diabetic independent of obesity. (Figure 5A). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0157] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a diabetic control subject or a NGT control subject.

[0158] In still yet another embodiment, a test subject is stratified with respect to obesity or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from modCer 576.5/7.7, modCer 883.8/7.8, modCer 921.8/9.1, PC 40:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, COH, DG 16:0 22:5, DG 16:0 22:6, DG 18:0 18: 1, DG 18: 1 18: 1, TG 14:1 18:0 18:2, TG 16: 1 16:1 18: 1, TG 16: 1 18:1 18:1, TG 16:1 18: 1 18:2, TG 18:1 18: 1 18:1, and TG 18:1 18: 1 22:6 wherein the level of an individual lipid analyte is different between obese control subjects and pre-diabetic control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the subject is pre-diabetic independent of obesity. (Figure 6A). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0159] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic or an NGT subject.

[0160] In some embodiments, the assays to stratify a test subject with respect to obesity or diabetes or pre-diabetes comprise determining the levels of at least two polyunsaturated lipid analytes selected from PC 38:6, PC 40:5, PC 40:6, PC 40:7, PC 44: 12, APC 38:6, LPC 22:6, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, PS 40:6, CE 22:5, CE 22:6, DG 16:0 22:5, DG 16:0 22:6, and TG 18: 1 18:1 22:6 wherein the levels of the individual lipid analytes are different between obese control subjects and diabetic control subjects or pre-diabetic control subjects and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the subject is diabetic or pre-diabetic independent of obesity. (Figure 8). Comparison of the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control in this embodiment can be carried out in different ways.

[0161] In an illustrative example, the assays comprise comparing individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic, diabetic or NGT subject.

[0162] The present invention further provides assays to stratify a test subject as normal with respect to diabetes or diabetes and pre-diabetes, the assays comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes selected from dhCer 16:0, DHC 24: 1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.608.4/4.0, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 and comparing the individual levels of the at least two lipid analytes in the subject to the respective levels of the same lipid analytes in at least one control subject selected from a normal control subject and a diabetic control subject or pre-diabetic control subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the diabetic or pre-diabetic disease control subject identifies the subject having diabetes or pre-diabetes, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the normal control subject identifies the test subject as a normal (NGT) subject with respect to diabetes or diabetes and pre-diabetes. [0163] In another embodiment, assays to stratify a test subject as normal with respect to diabetes comprise determining the individual levels of at least two lipid analytes (or 2 to 20 lipid analytes, including every integer between) selected from modCer 883.8/7.8, APC 34: 1, COH, TG 14:0 16:0 18:2, TG 14:0 16: 1 18:1, TG 14:0 16:1 18:2, TG 14:0 18:2 18:2, TG 14:1 16:0 18:1, TG 14: 1 16: 1 18:0, TG 14: 1 18:0 18:2, TG 14:1 18:1 18:1, TG 15:0 18:1 18: 1, TG 16:0 16:1 18:1, TG 16:1 16:1 16: 1, TG 16:1 16: 1 18:0, TG 16: 1 16:1 18:1, TG 16:1 18:1 18: 1, TG 16:1 18:1 18:2, TG 17:0 18: 1 16:1, and TG 17:0 18:2 16:0 and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a normal control subject and a diabetic control subject or pre-diabetic control subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the diabetic or pre-diabetic disease control subject identifies the test subject as diabetic or pre-diabetic, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the normal control subject identifies the test subject as being a normal (NGT) subject with respect to diabetes.

[0164] Suitably, the test subject is identified as being non-susceptible ("NGT") with respect to type Π diabetes when the level of at least two lipid analytes in the test subject varies from the level of the same lipid analytes in a NGT control subject or in a non-susceptible control population of subjects by no more than about 20%, 18%, 16%, 14%, 12%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1% or 0.1%.

[0165] In some embodiments, the individual level of a lipid analyte in a NGT subject is at least 101%, 102%, 103%, 104%, 105%, 106%, 107% 108%, 109%, 1 10%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900% or

I.000% (i.e. an increased or higher level), or no more than about 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.01%, 0.001% or 0.0001% (i.e. a decreased or lower level) of the level of the same lipid analyte in a diabetic or pre-diabetic subject.

[0166] In some illustrative examples of any of the above assays, the assays comprise, consist or consist essentially of determining and comparing the levels of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,

I I, 12, 13, 14, 15 or 16 lipid analyte classes or species identified in Table 4, 6 or 10.

[0167] In some further illustrative examples of any of the above assays, the assays comprise or consist of determining or determining and comparing the levels of at least 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30 or more lipid analytes including 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287, 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347 or 348 lipid analytes.

[0168] Presymptomatic diagnosis may be used to facilitate prevention or delay the onset of diabetes, including the use of existing medical therapies. Lipidotyping of individuals is useful for (a) identifying a subject or metabolic state that will respond to particular drugs, (b) identifying types of subject that respond well to specific medications or medication types with fewer adverse effects and (c) guide new drug discovery and testing.

[0169] The lipidomic profile further enables determination of endpoints in pharmacotranslational studies. For example, clinical trials can take many months or even years to establish the pharmacological parameters for a medicament to be used in treating or preventing diabetes onset or pre-diabetes care. However, these parameters may be associated with a lipidomic profile associated with a health state. Hence, the clinical trial can be expedited by first selecting a medicament and pharmaceutical parameters which result in a lipidomic profile associated with the desired health state (e.g. NGT).

[0170] Accordingly, another aspect of the present invention contemplates a method for determining the pharmacoefficacy of a medicament for use in diabetes or pre-diabetes treatment or prophylaxis, the method comprising, consisting or consisting essentially of selecting a medicament and its concentration and/or formulation parameters which provide a lipidomic profile associated with or characteristic of a non-susceptible individual as described herein. As described herein, the lipidomic profile is broadly identified by determining the levels of a lipid analyte selected from the list consisting of:

(i) one or more modified lipid analytes listed in Table 10;

(ii) two or more non-modified lipid analytes listed in Table 10, and (iii) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10.

[0171] Another aspect of the present invention provides a method for conducting a clinical trial for a medicament for the treatment or prophylaxis of diabetes or pre-diabetes, the method comprising, consisting or consisting essentially of conducting the clinical trial using a formulation of the medicament, which generates a lipidomic profile associated or characteristic of a non-susceptible individual as described herein. The lipidomic profile is identified by determining the levels of a lipid analyte selected from the list consisting of:

(i) one or more modified lipid analytes listed in Table 10;

(ii) two or more non-modified lipid analytes listed in Table 10, and

(iii) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10.

[0172] The lipidomic profile, therefore, can be used as a marker to define a desired state of health in an individual. It can be considered, therefore, a defined surrogate endpoint or desired endpoint in clinical management of subjects having diabetes or pre-diabetes or diabetes susceptibility treatment.

[0173] Accordingly, in an illustrative embodiment, a test subject group comprising diabetic and/or pre-diabetic subjects and/or subjects identified as susceptible to diabetes as described herem are administered a medicament for use in treatment or prevention of diabetes or pre-diabetes for a time and under conditions proposed to effect treatment or prevention. During the administration period or thereafter, the test subject group is subjected to lipid profiling as described herein to stratify individual subjects as susceptible or non-susceptible, diabetic or non-diabetic, obese, NGT etc. If the lipid analyte profile of a test subject indicates an improvement in the disease status (e.g., from diabetic to pre-diabetic or NGT) or an improvement in susceptibility (i.e., non- susceptibility) of the test subject to diabetes onset as a result of the treatment, then the medicament and/or its formulation is are selected for further testing or use in the treatment of diabetic, prediabetes or susceptible subjects (as defined herein). In some embodiments, the lipid profiling is conducted before administration, during administration and after cessation of administration of the medicament.

[0174] Similar clinical trials and tests are also conducted in respect of obesity treatments or prophylactic measures using lipid profiling as described herein as an indication of pharmacoefficiency or as a surrogate endpoint.

[0175] Pre-symptomatic diagnosis will enable better treatment of diabetes, including the use of existing medical therapies. Lipidotyping of individuals is useful for (a) identifying a subject or metabolic state which will respond to particular drugs, (b) identifying types of subject that respond well to specific medications or medication types with fewer adverse effects and (c) guide new drug discovery and testing.

[0176] Even yet another aspect of the present invention relates to a method of treatment or prophylaxis of a subject comprising, consisting or consisting essentially of stratifying the subject with respect to diabetes or pre-diabetes by determining the levels of a lipid analyte selected from the list consisting of:

(i) one or more modified lipid analytes listed in Table 10;

(ii) two or more non-modified lipid analytes listed in Table 10, and

(iii) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10;

[0177] The present invention further provides a system where data on levels of lipids are provided by a client server to a central processor which analyses and compares to a control and optionally considers other information such as patient age, sex, weight and other medical conditions and then provides a report, such as, for example, a risk factor for disease severity or incidence or progression or status or an index of probability of diabetes in pre-diabetic or NGT individuals.

[0178] Hence, knowledge-based computer software and hardware also form part of the present invention.

[0179] In particular, the assays of the present invention may be used in existing or newly developed knowledge-based architecture or platforms associated with pathology services. For example, results from the susceptibility or stratification assays are transmitted via a communications network (e.g. the internet) to a processing system in which an algorithm is stored and used to generate a predicted posterior probability value which translates to the index of disease probability which is then forwarded to an end user in the form of a diagnostic or predictive report.

[0180] The assays may, therefore, be in the form of a kit or computer-based system which comprises the reagents necessary to detect the level (concentration) of the lipid analytes and the computer hardware and/or software to facilitate determination or determination and comparison and transmission of reports to a clinician.

[0181] In an illustrative example, the present invention contemplates a method of allowing a user to determine the status of a subject with respect to diabetes or diabetes susceptibility, obesity, pre-diabetes or risk of incident diabetes etc. the method including: (a) receiving data in the form of levels or concentrations of a lipid analyte selected from the list consisting of: (i) one or more modified lipid analytes listed in Table 10; (ii) two or more non- modified lipid analytes listed in Table 10, and (iii) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non-modified lipid analyte listed in Table 10; wherein the level or ratio of the lipid analyte or analytes relative to a control provides a correlation to the presence, state, classification, susceptibility, incidence or progression of diabetes; from the user via a communications network; (b) processing the subject data via multivariate analysis to provide a disease index value; (c) determining the status of the subject in accordance with the results of the disease index value in comparison with predetermined values; and (d) transferring an indication of the status of the subject to the user via the communications network reference to the multivariate analysis includes an algorithm which performs the multivariate or univariate analysis function.

[0182] Conveniently, the method generally further includes: (a) having the user determine the data using a remote end station; and (b) transferring the data from an end station to a base station via the communications network.

[0183] The base station can include first and second processing systems, in which case the method can include: (a) transferring the data to the first processing system; (b) transferring the data to the second processing system; and (c) causing the first processing system to perform the multivariate analysis function to generate the disease index value.

[0184] The method may also include: (a) transferring the results of the multivariate analysis function to the first processing system; and (b) causing the first processing system to determine the status of the subject. In this case, the method also includes at lest one of: (a) transferring the data between the communications network and the first processing system through a first firewall; and (b) transferring the data between the first and the second processing systems through a second firewall.

[0185] The second processing system may be coupled to a database adapted to store predetermined data and/or the multivariate analysis function, the method include: (a) querying the database to obtain at least selected predetermined data or access to the multivariate analysis function from the database; and (b) comparing the selected predetermined data to the subject data or generating a predicted probability index. The second processing system can be coupled to a database, the method including storing the data in the database. Thus, the base station attempts to identify individuals having similar parameter values to the test subject and once the status has been determined on the basis of that identification, the base station provides an indication of the diagnosis to the end station. [0186] The method can also include having the user determine the data using a secure array, the secure array of elements capable of determining the level of lipid analytes and having a number of features each located at respective position(s) on the respective code. In this case, the method typically includes causing the base station to: (a) determine the code from the data; (b) determine a layout indicating the position of each feature on the array; and (c) determine the parameter values in accordance with the determined layout, and the data.

[0187] The method can also include causing the base station to: (a) determine payment information, the payment information representing the provision of payment by the user; and (b) perform the comparison in response to the determination of the payment information.

[0188] Still another aspect of the present invention contemplates the use of a panel of lipid analytes selected from the list consisting of: (i) one or more modified lipid analytes listed in Table 10; (ii) two or more non-modified lipid analytes listed in Table 10, and (Hi) two or more lipid analytes wherein at least one is a modified lipid analyte listed in Table 10 and at least one is a non- modified lipid analyte listed in Table 10; in the manufacture of an assay to identify a subject who is susceptible or non-susceptible with respect to diabetes onset and/or to stratify a subject as diabetic, pre-diabetic, obese or NGT.

[0189] The present invention is further described by the following non-limiting Examples.

EXAMPLES

EXAMPLE !. Cross sectional studies

(0190] Logistic regression analysis was performed on the cross sectional study to determine the association of each lipid class as well as individual lipid species with different disease states as outcomes. Using the combined NGT and diabetes groups with diabetes as the outcome 8 lipid classes were identified that were significantly (p<0.01) associated with diabetes independent of age, sex, waist (a measure of obesity), education, exercise and systolic blood pressure (Table 4). Also, 132 individual lipid species were identified that were similarly associated (Table 6). Comparison of the NGT and pre-diabetes groups identified 10 lipid classes and 117 individual lipid species that were significantly (p<0.01) associated with pre-diabetes as an outcome, independent of age, sex, waist, education, exercise and systolic blood pressure (Table 4 and 6).

[0191] In contrast there was only a single lipid species that was significantly associated with diabetes when analyzing the combined diabetes and pre-diabetes groups, although there were a further 23 lipid species that showed a significance at the (p<0.05) level (Table 6).

[0192] The data were analyzed to identify those species that were significantly associated with obesity independent of age sex, education exercise systolic blood pressure and diabetes status, which was adjusted for by including HbAlc in the analysis. In this analysis obesity was defined as (waist circumference >102, male; waist circumference >88, female). 10 lipid classes and 141 individual lipid species were identified that were significantly (pO.Ol) associated with obesity (Tables 4 and 6).

[0193] Mean concentrations and standard deviations for each lipid class and individual lipid species are shown in Tables 5 and 7.

[0194] The relative association of each lipid class to diabetes, pre-diabetes and obesity are shown in Figure 1. Selected lipid classes are also shown in Figures 2-4.

[0195] All lipid species were ranked on the basis of the strength of their association with diabetes relative to the strength of their association with obesity by selecting those species with an odds ratio of diabetes against NGT greater than 1.0 and dividing this by the odds ratio of obese against non-obese, where the odds ratio of diabetes against NGT was less than 1.0 then the odds ratio of obese against non-obese was divided by this value. The resulting ratios were then ranked and the lipid species with the highest values (representing those lipid species with the strongest association with diabetes relative to their association with obesity) were selected and the relative association to diabetes, pre-diabetes and obesity were plotted (Figure 5A). The lipid species with the lowest values of this ranking (representing those species with the strongest association to obesity relative to their association with diabetes) were selected and the relative association to diabetes, pre-diabetes and obesity were plotted (Figure 5B).

[0196] Using the same approach, all lipid species were ranked on the basis of their strength of association to pre-diabetes relative to the strength of their association with obesity and the highest and lowest ranked lipids were selected and the relative association to diabetes, prediabetes and obesity were plotted (Figure 6). >

[0197] To identify lipid species that were differentially associated with diabetes and pre-diabetes the ratio of the odds ratios for diabetes and pre-diabetes were calculated as described above, these were then ranked and the top and bottom 20 lipid species selected and their relative association to diabetes, pre-diabetes and obesity plotted (Figure 7).

[0198] The relative association of selected lipids, containing polyunsaturated fatty acids, to diabetes, pre-diabetes and obesity is shown in Figure 8.

EXAMPLE 2. Longitudinal studies

[0199] Logistic regression was performed on the longitudinal study to determine the association of each lipid class and individual lipid species with incident diabetes (those individuals who developed diabetes at follow-up). Logistic regression was also performed on only the prediabetes group (at baseline) to determine the association with pre-diabetes progression (individual with pre-diabetes at baseline who progress to develop diabetes at follow-up). There were 10 lipid classes that were significantly (pO.01) associated with incident diabetes independent of age, sex, waist (a measure of obesity), education, exercise and systolic blood pressure (Table 8). There were no lipid classes that were associated with pre-diabetes progression at this level of significance although two classes total PC and total DG did show an association with pre-diabetes progression with p<0.05. There were 124 lipid species that were associated with incident diabetes (p<0.01) and 7 species that were associated with pre-diabetes progressors (p<0.01) although there were a further 28 lipid species associated with pre-diabetes progression (p<0.05) (Table 10).

[0200] Mean concentrations and standard deviations for each lipid class and individual lipid species are shown in Tables 9 and 11.

[0201] The relative association of each lipid class to incident diabetes and pre-diabetes progression and obesity are shown in Figure 9. Lipid species were ranked by the strength of their association with incident diabetes and the 10 lipid species with the strongest positive association and 10 species with the strongest negative association were selected and their relative association with incident diabetes and pre-diabetes progression was plotted (Figure 10A). There were no lipid species negatively associated with pre-diabetes progression. The 20 lipid species with the strongest positive association with pre-diabetes progression were selected and their relative association with incident diabetes and pre-diabetes progression was plotted (Figure 10B).

EXAMPLE 3. Comparison of cross sectional and longitudinal cohorts

[0202] Of the 348 lipid species measured in the cross sectional study 321 were also measured in the longitudinal study. The 27 excluded species were those of very low abundance that did not provide useful information. To assess whether the same lipid species identified as associated with diabetes, pre-diabetes and obesity in the cross sectional study are also associated with incident diabetes in the longitudinal study the odds ratios determined from the cross sectional study for diabetes were plotted against NGT, pre-diabetes against NGT and obese against non- obese (Table 6) against the odds ratio determine in the longitudinal study for incident diabetes (Table 10). These plots show a strong correlation between the lipids associated with diabetes prediabetes and obesity in the cross sectional study and the lipid associated with incident diabetes in the longitudinal study (Figure 11).

EXAMPLE 4. Multivariate modeling

[0203] Recursive feature elimination using support vector machine learning was applied to the cross sectional study to develop multivariate models of varying feature size (e.g., 1, 2, 4, 8, 16..., 348) that included either lipids alone, risk factors alone or lipids with risk factors. The ranked list of the lipids risk factors according to the frequency of their recurrent incorporation in the generated models is shown in Tables 12. The C-statistic and % accuracy from each model was plotted against the .number of variables to assess the performance of the different models and identify the minimum number required for optimal discrimination (Figure 12). Models using lipids alone to discriminate diabetes from non-diabetes showed a maximum C-statistic of 0.750 (CI 0.742-0.758) with all lipids in the model. This was significantly better than the model created with nine risk factors alone (age, sex, waist, education, exercise, systolic blood pressure, total cholesterol, HDL and triglycerides) (C-statistic of 0.714, CI 0.706-0.722), Risk factor models excluding total cholesterol, HDL and triglycerides performed significantly worse with a maximum C-statistic of 0.663 (CI 0.652-0.673). The combination of lipids and risk factors performed similar to lipid alone.

[0204] The same recursive feature elimination approach was used to develop models to predict incident diabetes in the longitudinal study. These models were able to achieve a C-statistic of 0.703 and an accuracy of 71.1% (Figure 13). [0205] In order to assess if models created on the cross sectional study cohort could be used to predict incident diabetes in the longitudinal study each model (of varying feature size) created on the cross sectional data was tested on the longitudinal data for the ability to predict incident diabetes (Figure 13). These models were able to achieve a C-statistic of up to 0.695 and an accuracy of 64.5%.

[0206] To investigate the cross sectional study could be used to select the optimal lipid species for prediction of incident diabetes the top ranked 64 lipid species from the cross sectional study were selected and used the data for these lipid species in the longitudinal study to create predictive models of incident diabetes using the same recursive feature elimination approach with three fold cross validation (repeated 100 times). The outcome of these models gave a maximum C- statistic of 0.684 and accuracy of 63.8% (Figure 13).

[0207] The power of these models in the longitudinal cohort was compared with the existing risk scores (AUSDRISK (Chen et al., 2010 {supra)) and FINDRISC (Lindstrom and Tuomilehto, Diabetes Care 26(3): 725-31, 2003) San Antonio Heart Study (Mann et al, Am J Epidemiol 171(9): 980-8, 2010), Framingham Offspring study (Mann et al., 2010 (supra)) and the ARIC study (Schmidt et al:, Diabetes Care 25(8): 2013-8, 2005)) and the ROC analyses of these is shown in Table 13. The multivariate models based on plasma lipids performed similarly to other existing risk scores.

[0208] The ARIC score which included fasting glucose triglycerides and HDL cholesterol in addition to the clinical variables age, sex, ethnicity, history, systolic blood pressure, waist and height, performed significantly better than plasma lipids alone.

DISCUSSION OF RESULTS PRESENTED IN EXAMPLES 1-4

[0209] In the cross sectional studies it has been demonstrated that many plasma lipids are associated with diabetes independently of age, sex, education exercise systolic blood pressure and obesity. Importantly most of these lipids are also associated with pre-diabetes which suggests that the dyslipidemia which is reflected in the plasma lipids of the diabetes group precedes the onset of diabetes.

[0210] While many of the lipids that are significantly associated with diabetes and prediabetes are also associated with obesity, the effects are independent which would indicate that those obese individuals who also have diabetes or pre-diabetes will display a greater difference in the levels of these lipids than obese, non-diabetic individuals. This suggests that these lipids may be useful markers to for the identification and risk assessment of diabetes and pre-diabetes even in an obese population. [0211] Importantly a number of lipids have been identified that are associated with diabetes and pre-diabetes but show no association with obesity, suggesting that there are at least some unique metabolic changes associated diabetes. These lipids appear to contain a high proportion of polyunsaturated fatty acids suggesting some difference in the metabolism of these fatty acids in diabetes (Figures 5 and 8) and pre-diabetes (Figures 6 and 8).

[0212] Plasma lipids associated with obesity but not diabetes or pre-diabetes appear to include predominantly LPC, LPAF and other phosphocholine containing species (Figures 5 and 6).

[0213] A series of lipids has been identified, which shows a stronger association against pre-diabetes relative to diabetes and vice versa, these results indicate that there are specific metabolic processes associated with different stages of the disease. Lipids showing a stronger association with pre-diabetes contain a high representation of the fatty acid palmitoleate (16:1) which is associated with de novo fatty acid synthesis and a potent lipokine involved in insulin sensitivity and metabolic homeostasis (Cao et al., Cell 134(6): 933-44, 2008).

[0214] These observations indicate that there are complex differences in lipid metabolism associated with different stages of the disease continuum which spans from obesity through insulin sensitivity and pre-diabetes to frank diabetes and that these differences are reflected in plasma lipids. As such these plasma lipids are useful lipid analytes for the early detection of, and risk assessment for, pre-diabetes and diabetes.

[0215] The multivariate models of the cross sectional study defined herein indicate that plasma lipids are able to better discriminate diabetes from non-diabetes groups than traditional risk factors, even when the lipid measures of total cholesterol, triglycerides and HDL cholesterol are included as risk factors. It has also been demonstrated that as few as 16 lipid species can be used to create a multivariate model that has almost equal discriminating power to the models created with over 300 lipid species.

[0216] Validation of the cross sectional models on the longitudinal cohort further demonstrates the ability of these lipid based models to predict incident diabetes.

[0217] Creation of either de novo multivariate models on the longitudinal study or models using the lipids selected from the cross sectional study but modeled on the longitudinal cohort did not improve the performance over the models created solely on the cross sectional study.

[0218] These results demonstrate that multivariate models based on plasma lipid profiles are able to predict incident diabetes.

MATERIALS & METHODS FOR EXAMPLES 1-4

The Australian Diabetes Lifestyle and Obesity Study (AusDiab) [0219] AusDiab was established to estimate the prevalence of diabetes and risk factors for diabetes and CVD in a national population sample. Baseline testing (in 1999-2000) involved 11,247 adults aged >25 years residing in 42 randomly selected areas of the six states of Australia and the Northern Territory (Dunstan et al, Diabetes Res Clin Pract 57(2): 119-129, 2002). Demographic information, smoking history, alcohol intake, dietary intake, history of CVD and diabetes were collected by questionnaire, and blood pressure and anthropometrics were measured. A two-hour oral glucose tolerance test (OGTT), fasting plasma lipids, insulin, HbAi_c and fibrinogen were determined. Urinary albumin, protein and creatinine were measured and the glomerular filtration rate was estimated.

[0220] In 2004/05 all surviving participants of AusDiab were invited to attend a 5-year follow-up, at which all the parameters measured at baseline were repeated. Diabetes and obesity status was ascertained using results from blood test and anthropometric data, respectively. CVD outcomes were ascertained by self-report and were adjudicated from medical records. Based on the longitudinal data, the Australian Type II diabetes Risk Assessment Tool (AUSDRISK) was developed (Australian Government DoHaA. The Australian Type II diabetes Risk Assessment Tool (AUSDRISK), An Australian Government, State and Territory health initiative, http://www.health.gov.au/internet/main/publishing.nsf/Content/C73A9D4A2E9C684ACA2574730 002A31B/$File/Risk_Assessment_Tool.pdf). This self-administered test can provide a risk assessment base on a 10 point questionnaire covering age, sex, family history, lifestyle and waist measurement. This enables the classification of an individual into a group with a 5-year risk of developing diabetes between 1 in 100 and 1 in 3.

Sample storage and stability

[0221] All samples in the AusDiab Biobanks used for this study have been stored at - 80°C since collection. The effect of storage time on 329 plasma lipid species in our current lipid profile (including oxidized phosphatidylcholine species) was investigated using 61 control plasma samples collected between 2002 and 2007. Of the 329 lipids examined, a weak correlation between plasma concentration and years of storage was observed (Spearman's Rho -0.256 to -0.324, 0.01<p<0.05) in only 15 lipid species.

Cross sectional and longitudinal study groups

[0222] Two separate plasma lipid profiling studies were performed on the archived samples from the AusDiab cohort. A cross sectional study was performed in which plasma lipid profiles from a sub-group of the AusDiab cohort were determined. This group consisted of 117 individuals with newly diagnosed diabetes who were non-smokers, these were age and sex matched (1 :2) with non-diabetic individuals who were also non-smokers. This group of non-diabetic participants had 64 with pre-diabetes (IGT and IFG combined) and 170 normal glucose tolerant (NGT) participants (see Table 1 for cohort characteristics).

[0223] In a second longitudinal study plasma lipid profiles on a cohort of 223 individuals (115 men and 108 women) who were non-diabetic at baseline were determined (samples collected in 1999-2000) but developed diabetes by follow-up (2005). This group included 165 individuals who were pre-diabetic at baseline and 58 individuals who were NGT at baseline. This group was age and sex matched 1:2 with individuals who were non-diabetic at baseline and remained non-diabetic at follow-up. This group consisted of 430 participants (224 men and 206 women) and contained 94 individuals who were pre-diabetic at baseline but did not develop diabetes by follow-up and 336 individuals who were NGT at baseline (see Table 2 for cohort characteristics).

Sample preparation and lipid extraction

[0224] Plasma samples were randomized prior to lipid extraction and analysis. Plasma samples (200|iL) were thawed and treated with the antioxidant butylhydroxytoluene (BHT) (Ιμί of iOOmM in ethanol). Total lipid extraction from of plasma was performed by a single phase chloroform:methanol extraction. A \0μ1_^ aliquot of plasma was combined with 200 μΐ, CHCl₃ MeOH (2:1) and 15μ1_^ of internal standard mix (Table 3) then briefly vortexed. Samples were mixed (rotary mixer, 10 min), sonicated (water bath, 30 min) then allowed to stand (20 min) at room temperature. Samples were centrifuged (16,000*g, 10 min) and the supernatant was dried under a stream of nitrogen at 40°C. The extracted lipids were resuspended in 50 L ¾0 saturated BuOH with sonication (10 min), followed by 50uL of 10 mM N¾COOH in MeOH. Extracts were centrifuged (3,350xg, 5 min) and the supernatant transferred into 0.2 mL glass vials with teflon insert caps. Mass spectrometric analysis was performed using either 1 or 5 μΐ, injections of the lipid extracts.

Lipid identification and quantification

[0225] Lipid analysis was performed by liquid chromatography, electrospray ionisation-tandem mass spectrometry (LC ESI-MS/MS) using a Agilent 1200 liquid chromatography system combined with an Applied Biosystems API 4000 Q/TRAP mass spectrometer with a turbo-ionspray source (350°C) and Analyst 1.5 data system.

[0226] Liquid chromatography was performed on a Zorbax C18, 1.8 μπι, 50 x 2.1 mm column at 300 μί/πιϊη using the following gradient conditions; 0% B to 100% B over 8.0 min, 2.5 min at 100% B, a return to 0% B over 0.5 min then 3.0 min at 0% B prior to the next injection. DGs and TGs were separated using the same solvent system with an isocratic flow (100 μί/ιηίη) of 85% B. Solvent A and B consisted of tetrahydrofuran:methanol:water in the ratios (30:20:50) and (75:20:5) respectively, both containing 10 mM NH₄COOH.

[0227] Precursor ion scans were performed on plasma extracts from healthy individuals to identify the major lipid species of the following lipid classes: ceramide (Cer), monohexosylceramide (MHC), dihexosylceramide (DHC), trihexosylcermide (THC), G o ganglioside (GM3), sphingomyelin (SM), phosphatidylglycerol (PG), bis(monoacylglycerol)phosphate (BMP), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylinositol (PI), lysophosphatidylcholine (LPC), lysoplatelet activating factor (LPAF), phosphatidylcholine (PC), alkylphosphatidylcholme (APC), cholesterol ester (CE), diacylglycerol (DG) and triaclyglycerol (TG). Modified ceramide (modCer), phosphatidylcholine (modPC) and cholesterol ester (modCE) species were identified using precursor ion scans for mass to charge ratio (m/z) 264.3, m/z 184.1 and m/z 360.3 respectively.

Lipid Nomenclature and examples

^■ [0228] The nomenclature (both systematic and common names) used in this document has come primarily from the two recent publications on this topic from the Lipid Maps Consortium (Fahy et al, JLipidRes 46(5): 839-61, 2005; Fahy et al, J Lipid Res 50(SuppI): S9-14, 2009).

[0229] In addition, a number of terms have been used to define lipid species where the full structure is not known but where characteristic collision induced fragmentation data has provided us with a partial structure of the lipid species. These are as follows .

[0230] modPC xxx.x yy. = modified or undefined phosphocholine containing lipid species with mass/charge ratio of the M+H ion denoted by xxx.x and retention time under our defmed chromatographic conditions defined as yy.y minutes.

[0231] modCer xxx.x/yy.y = modified or undefined sphingosine containing lipid species with mass/charge ratio of the M+H ion denoted by xxx.x and retention time under our defined chromatographic conditions defined as yy.y minutes.

[0232] modCE xxx.x/yy.y = modified or undefined cholesterol containing lipid species with mass/charge ratio of the M+H ion denoted by xxx.x and retention time under our defined chromatographic conditions defined as yy.y minutes.

[0233] Multiple Reaction Monitoring (MRM) experiments were established for each of the lipid species identified from the precursor ion scans (Table 3). A total of 71 DG and TG species and between up to 277 other lipid species were analyzed in two separate experiments. Quantification of individual lipid species was then performed using scheduled multiple-reaction monitoring (MRM) in positive ion mode (Murphy et al, Anal Biochem 366(1): 59-70, 2007; Smyth et al, PLoS Genet 4(9): el000192, 2008; Matthews et al, Diabetologia 53(11): 2431-41, 2010; Tandy et al, Atherosclerosis 2/5(1): 142-7, 2010). Lipid concentrations' were calculated by relating the peak area of each species to the peak area of the corresponding internal standard. CE species were corrected for response factors determined for each species. Total measured lipids of each class were calculated by summing the individual lipid species. Results are expressed as pmol per mL of plasma.

Statistical analysis

[0234] The relative strength of the association of each lipid to obesity, pre-diabetes and diabetes in the cross sectional cohort was determined using binary logistic regression adjusting for the covariates age, sex, education, exercise, systolic blood pressure and waist when analyzing for pre-diabetes and diabetes as the outcomes and adjusting for the covariates age, sex, education, exercise, systolic blood pressure and HbAlc when analyzing for obesity as the outcome.

[0235] The relative strength of the association of each lipid to incident diabetes or prediabetes progression to diabetes in the longitudinal cohort was also determined using binary logistic regression adjusting for the covariates age, sex, education, exercise, systolic blood pressure, smoking and waist.

[0236] Classical statistical analysis requires sample number to be in excess to the number of independent variables. In this context, modern high throughput screens, as described herein, present with an innate difficulty to exclude the possibility that a variable appears to be associated with the outcome, merely for the set of plasmas that have been obtained. This challenge of multiple testing, or the curse of dimensionality is further accentuated by the fact that lipid levels present with fractional changes, rather than high fold change. Fortunately, these variables described herein are not truly independent, which is why most statistical methods of analysis of biological high throughput data involve feature selection algorithms, often using support vector machine learning to define the optimal and most generalizable classifier (Johannes et al, Bioinformatics 26(17): 2136-2144, 2010).

[0237] A recursive feature elimination with three-fold cross-validation (repeated 100 times) was employed using support vector machine learning. This feature selection method recursively removes features based on the absolute magnitude of each hyperplane element (Guyon et al, Machine Learning 46(\): 389-422, 2002). Given lipidomic data with n features per sample, each SVM classifier outputs a hyperplane, w, that can be thought of as a vector with n elements each corresponding to the level of a particular feature. Assuming that the values for each feature have similar ranges, the absolute magnitude of each element in w determines its importance in classifying a sample, because f(x)

Each SVM classifier is first trained with all features, then features corresponding to \w i\ in the bottom 10% are removed, and each classifier is retrained with the smaller feature set. This procedure is repeated iteratively to study prediction accuracy as a function of feature number. Models of varying feature size (e.g., 2, 4, 8, 16, 32....348) were created. Receiver operator characteristic (ROC) analysis was performed for each model. The discriminating power of each model was assessed by the C-statistic and the accuracy. The C-statistic is the area under the ROC curve and has a range of 0.5-1.0, the higher the better the discriminating model.

EXAMPLE 5. Plasma lipid profiling identifies associations with obesity and type II diabetes

[0238] Study Cohort: The pre-diabetes group was older than the NGT group and the diabetes group (Table 15). There was no difference in waist/hip between the diabetes and prediabetes but both were slightly higher than the NGT group. Both diabetes and pre-diabetes showed a higher systolic blood pressure (SBP), total cholesterol and triglyceride levels relative to the NGT group. Fasting plasma glucose (FPG), 2h post load glucose, HbAlc and fasting insulin levels were also different between groups as would be expected (Table 15).

[0239] Lipid identification and quantification: Precursor ion scans and neutral loss scans were used to identify the lipid species present in human plasma (Supplementary Table 15). Quantification of individual lipid species was then performed using scheduled multiple-reaction monitoring (MRM) in positive ion mode (5-7; 13-15). Limit of detection (based on at least three times background signal) was between 2.5-5.0 fmol (injected) which is equivalent to 5-10 nM in plasma. The median coefficient of variation (%CV) for the individual lipid species within the QC samples was 10.6% with 90% below 20.4%.

[0240] Association of lipids with type II diabetes, pre-diabetes and obesity: When the concentrations of the lipid species within each lipid class were summed (total lipid classes) and the logistic regression analysis performed against diabetes (vs. NGT) and pre-diabetes (vs. NGT) adjusting for age, sex, SBP and obesity (as determined by waist/hip) or against obesity (in the combined diabetes and NGT groups) adjusting for age, sex, SBP and diabetes status, the present inventors observed positive associations of each outcome with diacylglycerol (DG), triacylglycerol (TG), phosphatidylglycerol (PG), dihydroceramide (dhCer) and ceramide (Cer) (all p<0.01, Table 16). Phosphatidylethanolamine (PE) and phosphatidylinositol (PI) were positively associated with diabetes and pre-diabetes (relative to NGT)! Free cholesterol was positively associated with prediabetes only while cholesterol ester (CE) was positively associated with both diabetes and obesity. Logistic regression analysis of lipid groups against diabetes (in the diabetes and pre-diabetes groups combined) revealed that only free cholesterol (COH) was significantly (p<0.01), negatively, associated with diabetes.

[0241] Within each lipid class considerable variation of the odds ratio of individual lipid species was observed against each of the outcomes. Odds ratios for the DG species against diabetes ranged from 1.16 (p=0.139, DG 18: 1/20:0) to 3.09 (p= 1.96x1ο^"8, DG 16:0/22:6). The present inventors identified 142 of the 296 lipid species as significantly associated with diabetes (pO.01, Tables 20 and 21). Similarly, 121 lipid species were significantly associated with prediabetes (pO.01). However, there were only 2 lipids that were significantly associated with diabetes relative to pre-diabetes after adjustment for age, sex, SBP and obesity (p<0.01, LPC 20:1 and COH). Logistic regression of lipid species against obesity defined by waist/hip in the combined •NGT and diabetes group identified 84 lipid species that were significantly associated with obesity (p<0.01, Tables 20 and 21). Sub-analysis of the NGT group only showed the same trend for all 84 lipid species although only 49 of these were significant (p<0.01 ).

[0242] The relationship between type Π diabetes and obesity: In order to characterize the degree of relatedness between the plasma lipid profiles associated with diabetes and obesity the odds ratios of all lipids for diabetes was plotted against the odds ratio of the same lipids for obesity (Figure 14). This showed a strong correlation between the lipid profiles associated with diabetes and obesity (R²=0.67). However, there were a number of lipid species that showed a strong and significant association with diabetes but showed a weak or non-significant or even an opposite association with obesity. These lipids may reflect a distinctive aspect of diabetes pathophysiology that is unrelated to obesity. To further explore this, the present inventors identified those lipids with a strong association against diabetes but a weak or non-significant association against obesity by calculating (odds ratio (OR) against diabetes /OR against obesity) then ranked these to identify those lipid species that were specifically associated with either condition (Figure 15). A strong negative association was observed between SM(OH)20:1, PC(0-34:2) and a number of odd chain phosphatidylcholine (oddPC) species with diabetes but no association with obesity. Some PE and PI species showed a strong positive association with diabetes and again no significant association with obesity. There were also a number of lipid species containing n-3 fatty acids (docosapentaenoic acid, C22:5, and docosohexaenioc acid C22:6) that showed a strong positive association with diabetes but a non-significant or weak association with obesity, these included species of PE, PS and PI. This effect was also seen in species of CE DG and TG (Table 21). In contrast, obesity showed a positive association with species of sphingomyelin (SM) which were not associated with diabetes. The present inventors also observed a number of glycolipid species, particularly trihexosylceramide (THC), which showed a negative association with obesity but were not significantly associated with diabetes (Figure 15).

[0243] The relationship between type II diabetes and pre-diabetes: The lipid profile associated with pre-diabetes showed a close homology to the diabetes lipid profile as demonstrated by the plot of odds ratios to diabetes against the odds ratios to pre-diabetes (Figure 14) which had a R² value of 0.76. The present inventors calculated (OR against diabetes/OR against pre-diabetes) then ranked these to identify those lipid species that were specifically associated with either condition (Figure 16). A negative association was observed between diabetes and several lysophosphatidylcholine (LPC) and lysoalkylphosphatidylcholine LPC(O) species which did not show a significant association with pre-diabetes. A stronger association was also observed between multiple CE species and diabetes than with pre-diabetes; this was particularly evident in those species containing monounsaturated and polyunsaturated fatty acids (Figure 16 and Table 21). In contrast, a negative association was observed between THC species and pre-diabetes but no significant association with diabetes and a similar effect with some alkylphosphatidylcholine PC(O) species. Free cholesterol and some LPC species containing saturated fatty acids were positively associated with pre-diabetes but again showed no association with diabetes. A number of species of TG, particularly those containing 14:0, 14: 1 and 16:1 fatty acids, showed stronger positive associations with pre-diabetes than diabetes (Figure 16 and Table 21).

[0244] Multivariate analysis: To assess the global variation in the lipid profiles and the relationship with diabetes, pre-diabetes and obesity the present inventors performed principal component analysis (Figure 17). While the distribution of scores were similar for lean and overweight and obese groups, a difference was observed in the distribution of scores between the NGT, pre-diabetes and diabetes groups, with the latter grouping in the lower right quadrant of the plot.

[0245] Development of classification models: Multivariate models to classify diabetes and pre-diabetes from NGT, incorporating FPG, traditional risk factors (in the form of the AUSDRISK score (Chen L, et al. Med J Aust 2010; 192: 197-202)) and increasing numbers of lipid species, within a three-fold cross validation framework (see Materials and Methods below for detailed description and rationale of model development) were compared with the performance of FPG+AUSDRISK alone (Figure 18). The combination of FPG+AUSDRISK and 5 lipid species gave improved performance over FPG+AUSDRISK alone in the C-statistic (AUC=0.899(CI 0.897- 0.901) compared to 0.838(CI 0.835-0.841)) as well as in accuracy (82.9% (CI 82.6-83.2) compared to 77.4% (CI 77.1-77.7%)). The top ranked features in these models are shown in Table 17. The models were also examined utilizing FPG, AUSDRISK and standard lipid measurements (total cholesterol, HDL cholesterol and triglycerides). These models performed better than the FPG+AUSDRISK model with a C-statistic and % accuracy of 0.884 (CI 0.881-0.887) and 82.0% (CI 81.7-82.3) respectively but were significantly below the performance of the FPG+AUSDRISK+lipids model.

[0246] To further assess how well the FPG+AUSDRISK+lipids model would discriminate diabetes, pre-diabetes and NGT compared to FPG and traditional risk factors alone the present inventors calculated confusion matrices, net reclassification indices and positive predictive values (Table 18). The model to classify diabetes and pre-diabetes from NGT resulted in a net reclassification index (NRJ) of 10.8% (Table 18B) resulting mainly from an increased sensitivity of the FPG+AUSDRISK+lipids model (Table 18A). Further to this, when the threshold of FPG was fixed at >5.5mM, to match current clinical guidelines (Diabetes_Australia: Diabetes management in General Practice: Guidelines for Type 2 Diabetes. 2009 http://www.racgp.org.au/Content NavigationMenu ClinicalResources RACGPGuidelines/Diabetesmanagement/200910diabetesmana gementingeneralpractice.pdf), the sensitivity and specificity of the FPG+AUSDRISK model decreased (Table 18B). Comparison of this model with the FPG+AUSDRISK+lipids model showed a NRI of 29.9%. Stratification of the diabetes and pre-diabetes showed that while the FPG>5.5mM+AUSDRISK model identified similar proportion of diabetes and pre-diabetes (82.9% and 65.5% respectively) to the FPG+AUSDRISK+lipids model (84.6% and 62.5% respectively), the FPG+AUSDRISK model identified 40.6% of NGT as false positive compared with only 10.6% of NGT as false positives in the FPG+AUSDRISK+lipids model (Table 18A). This equated to positive predictive values of 39.4% and 63.3% respectively (Table 18B).

DISCUSSION OF RESULTS PRESENTED IN EXAMPLE 5

[0247] In the study presented in Example 5, the present inventors performed lipidomic analysis of the plasma from 351 individuals and this large cohort was found to provide the ability to dissect out the relative association of obesity, pre-diabetes and diabetes with the plasma lipid profile, independent of each other and other confounding factors.

[0248] Logistic regression analysis was performed on lipid data that had been standardized to the interquartile range, thus the resulting odds ratio for each lipid or lipid class represents the difference in risk of being in the outcome group (diabetes, pre-diabetes or obese) as one moves from the 25^th to the 75^th percentile of the population. This analysis provided a measure of the strength of the association that can be compared across different outcomes and revealed that plasma lipid classes showed similar associations with obesity, pre-diabetes and diabetes, although for most of the lipid classes the strength of the association tended to be lower with obesity than with diabetes or pre-diabetes (Table 16). The associations of both Cer and its biosynthetic precursor dhCer firmly place the Cer biosynthetic pathway as a key metabolic process in obesity as well as diabetes. Here for the first time it can be observed that Cer biosynthesis is positively associated with obesity, even in NGT individuals, and further that Cer biosynthesis is associated with pre-diabetes and diabetes independent of obesity. Thus, it appears that obesity per se is neither sufficient nor required to induce the Cer biosynthesis associated with diabetes. Rather, obesity may up-regulate Cer biosynthesis that in some individuals is further exacerbated (by multiple mechanisms) contributing to the onset of diabetes. This independent association of Cer with diabetes was mirrored with CE, DG and TG, further supporting the association of lipid metabolic pathways with diabetes independent of obesity.

[0249] PE, PI and PG, but not PC, also showed positive associations with pre-diabetes, diabetes and to a lesser extent with obesity. The relationship between PE and obesity has been previously reported by Graessler et al. (PLoS One 2009;4:e6261) who found a significant increase in some plasma PE species in BMI >27.5 individuals relative to B I <27.5. In contrast Pietilainen et al. (PLoS One 2007;2:e218) reported a negative association between some plasma PE species and BMI although in both of these studies there was no adjustment for age sex or other covariates. More recently Fu et al. (Nature 2011;473:528-531) reported on the association between increased PC/PE ratios in mouse liver endoplasmic reticulum in obesity and proposed that this was due to an increased conversion of PE to PC. The present inventors also observed a positive association of PC with obesity, similar to the association with PE, however, this did not hold in the pre-diabetes or diabetes states where the association with PC was not significant despite a strong positive association with PE. This suggests that, in humans, the up-regulation of the conversion of PE to PC may be more important in the setting of obesity than diabetes. The positive associations of PI and PG with pre-diabetes and diabetes may relate to their respective roles as a source of arachidonic acid for the production of prostaglandins and eicosinoids and as a substrate for the production for the mitochondrial specific lipid cardiolipin respectively.

[0250] Of interest was the negative association of oddPC with pre-diabetes and diabetes but not obesity. These lipid species contain the odd chain fatty acids CI 5:0 and CI 7:0, which are products of ruminant digestion and in human diets are derived primarily from dairy fats. This would support a protective role for dairy against diabetes and is in agreement with a mounting body of epidemiological data (Elwood PC, et al. Lipids 2010;45:925-939; Villegas R, et al. Am J Clin Nutr 2009;89: 1059-1067) as well as specific measurement of fatty acid composition that have demonstrated negative associations between C15:0 and C17:0 with incident T2D (Kroger J, et al. Am J Clin Nutr 2011;93:127-142; Patel PS, et al. Am J Clin Nutr 2010;92:1214-1222) and a recent study of Mozaffarian et al. (Ann Intern Med 2010;153:790-799) who reported a negative association between the dairy derived fatty acid, trans-palmitoleate, and incidence of diabetes. Negative associations or trends with pre-diabetes and diabetes were also observed for the ether linked PC species PC(O) and PC(P) which may relate to the their purported role as antioxidants (Ford DA: Clin Lipidol 2010;5:835-852; Skaff O, et al. Biochemistry 2008;47:8237-8245) and degradation by reactive oxygen species in the pre-diabetes and diabetes individuals. Here also the association with obesity was not as strong suggesting this effect predominates in the transition from NGT through pre-diabetes to diabetes. [0251] In addition to differences in the association of lipid classes with obesity and diabetes, the present inventors also observed multiple differences in the associations of individual lipid species; in particular they observed a strong positive association between multiple lipid species containing n-3 PUFA and diabetes, but a weaker or non-significant association with obesity (Figure 15, Table 21). Epidemiological studies have reported a positive association between dietary intake of n-3 PUFA and incident diabetes (Djousse L, et al. Am J Clin Nutr 2011;93: 143-150; Kaushik M, et al. Am J Clin Nutr 2009;90:613-620) although this appears to be specific for marine based n-3 PUFA as the same association was not observed for the plant based cc-linolenic acid (C18:3 n-3) ( Djousse L, et al, supra). When the dietary intake of n-3 PUFA in our cohort was analyzed significant correlations (p values between 1.8 x 10^"2 and 3.1 x 10^"u) with 21 lipid species containing the n-3 PUFA (C22:5 and C22:6, data not shown) were observed, including all those shown in Figure 15: This indicates that dietary intake is likely the major contributor to higher levels of these lipid species in the diabetes group. Comparison of the associations of pre-diabetes and diabetes with lipid species revealed that diabetes was more strongly associated with most CE species; this may relate to higher levels of triglyceridemia and increased VLDL production associated with the progression from pre-diabetes to diabetes. In contrast, pre-diabetes was more strongly associated with multiple TG species, particularly those containing C14:l and C16:l fatty acids suggesting higher lipogenesis in the pre-diabetes state relative to diabetes, and both showing an elevation relative to NGT. This may relate to the hyperinsulinemia in the pre-diabetes state progressing to increased insulin resistance and/or beta cell decompensation in diabetes.

(0252] Although the present inventors have been able to utilize the power of their large population cohort to tease out many significant associations of plasma lipids with obesity in addition to pre-diabetes and diabetes, the PCA analysis clearly shows that pre-diabetes and diabetes are stronger determinants of variance in the plasma lipid profile than obesity. The plasma lipid profile associated with pre-diabetes are also observed as being highly correlated with that of diabetes (R² = 0.758, Figure 14). This implies that most changes in lipid metabolism associated with diabetes have already occurred in the pre-diabetic state, prior to the onset of frank diabetes.

[0253] Baseline data from the AusDiab study shows that FPG alone will identify only 51% of undiagnosed diabetes and that impaired fasting glucose (identified from FPG) represents only 33% of the combined impaired fasting glucose and impaired glucose tolerance (pre-diabetes) group (Dunstan DW, et al. Diabetes Care 2002;25:829-834). Clearly FPG alone is insufficient for the identification of diabetes or indeed of pre-diabetes. Current recommendations for the identification of diabetes are to perform an initial FPG and retest those with a FPG>5.5mM with an OGTT (Diabetes_Australia: Diabetes management in General Practice: Guidelines for Type 2 Diabetes, supra). In practice, other risk factors such as age, sex, waist and family history may also be considered. The present inventors sought to utilize plasma lipids to improve upon the discriminative power of FPG and other risk factors (captured as a single measure in the AUSDRISK score) for the identification of diabetes and pre-diabetes. Models created with FPG and AUSDRISK to classify pre-diabetes and diabetes from NGT showed a significant improvement in both AUC and accuracy with the inclusion of lipids (Figure 18) leading to a NRI of 10.8% (Table 18). These models also showed a higher sensitivity for diabetes (84.6% true positive) than pre-diabetes (62.5% true positive) reflecting the difference in the FPG between these two states with a possible contribution from the subtle difference between plasma lipid profiles.

[0254J In order to assess the diagnostic potential of the FPG+AUSDRISK+lipids model relative to current practice, the present inventors fixed the FPG threshold at 5.5mM in the FPG+AUSDRISK model and compared to the FPG+AUSDRISK+lipids model. While each of these models had the same true positive rates (76.8%), the model containing lipids had a higher positive predictive value of 63.3% compared to the FPG>5.5mM+AUSDRISK only model (39.4%), equating to a NRI of 29.9%. The baseline data from all participants in the AusDiab study (~1 1,000) contained 3.7% known diabetes, 3.7% unknown diabetes, 16.4% pre-diabetes and 79.9% NGT (2). Based on these figures the FPG+AUSDRISK+lipids model could stratify 84.6% T2D and 62.5% pre-diabetes into only 21.8% of the population in comparison to the FPG>5.5 mM+AUSDRISK model which would stratify similar proportions of diabetes and pre-diabetes (82.9% and 65.5% respectively) into 46.2% of the population. Thus, the inclusion of lipids with FPG and AUSDRISK into an initial screen for diabetes could reduce the number required to undergo an OGTT by approximately 53% with no loss in detection rates.

[0255] Many modifications will be apparent to those skilled in the art without departing from the scope of the present invention.

TABLE 1. Baseline characteristics of the cross sectional study cohort

Normal glucose tolerance Diabetes IGT/IFG

(n=170) (n=117) (n=64)

Male(n=81) Female (n=89) Male (n=60) Female (n=57) Male (n=34) Female (n=30)

Age¹ 57.5 (11.8) 62.3(13.7) 59.9(12.5) 63.9(12.6) 64.3(12.0) 68.4(9.6)

Waist' 95.7 (8.1) 85.2 (9.8) 102.3 (9.7) 90.5 (9.7) 97.2 (6.9) 88.8 (9.8)

SBP² 133.5 (21.3) 130(30.5) 143.8(20.3) 141 (28.5) 144.3 (30.5) 136.3 (28.3)

Total cholesterol¹ 5.6 (0.8) 5.9(1.1) 5.9(1.1) 6.1(1.1) 5.9(1.0) 6.3 (1.2)

LDL-C¹ 3.7 (0.8) 3.6(1.0) 3.7 (0.9) 3.7 (0.9) 3.8 (0.9) 3.9(1.0)

Triglycerides² 1.2(0.6) 1.3 (0.8) 1.9(1.8) 2.0 (1.4) 1.7(1.3) 1.9(1.0)

HDL-C² 1.3 (0.4) 1.6(0.5) 1.1 (0.3) 1.4 (0.6) 1.2(0.4) 1.5 (0.5)

Body mass index¹ 26.1 (2.7) 26.2 (3.5) 28.7(3.2) 27.6(3.8) 26.6(3.1) 26.9 (3.42)

¹ Data are mean (standard deviation)

² Data are median (interquartile range)

TABLE 2. Baseline characteristics of the longitudinal study cohort

No diabetes (n=430) Incident diabetes (n=223)

Male (n=224) Female (n=206) Male (n= 115) Female (n= 108)

Age¹ 57.2 (10.8) 54.5 (12.9) 56.8(10.9) 54.2(13.1)

Waist¹ 97.8(10.4) 84.6(11.8) 104.3(11.8) 92.7 (14.8)

Systolic Blood pressure¹ 134.2 (17.8) 129.5 (19.0) 139.4 (16.8) 132.5 (17.4)

Total cholesterol¹ 5.8(1.0) 5.7 (0.9) 5.8(1.1) 6.0(1.0)

LDL-C¹ 3.7(0.9) 3.5 (0.8) 3.7(0.9) 3.7(0.9)

Triglycerides² 1.4(1.1) 1.2(1.0) 1.8(1.1) 1.7(1.4)

HDL-C¹ 1.3 (0.3) 1.6(0.4) 1.2 (0.3) 1.4(0.5)

Body mass index¹ 27.0(3.8) 26.6(4.8) 29.5 (4.4) 29.6(6.7)

Data are mean (standard deviation)

2 Data are median (interquartile range)

TABLE 3. Conditions for precursor ion scan and MRM acquisition methods for lipid identification and quantification

TABLE 4. Logistic regression of total lipid classes in the cross sectional study

Non-Obese vs. Obese NGT vs. Pre-diabetes NGT vs. Diabetes Pre-diabetes vs. Diabetes

Odds Odds Odds Odds

Lipid class ratio p-value 95% CI ratio p-value 95% CI ratio p-value 95% CI ratio p-value 95% CI total Sph 1.22 8.77E-02 0.97-1.54 1.04 8.24E-01 0.75-1.44 0.83 1.56E-01 0.65- .07 0.80 1.67E_T01 0.58-1.10 total dhCer 1.46 2.60E-03 1.14-1.87 2.03 1.79E-04 1.40-2.95 1.85 1.07E-04 1.36-2.52 0.80 2.09E-01 0.56-1.13 total Cer 1.20 1.37E-01 0.94-1.53 1.53 1.91E-02 1.07-2.18 1.42 1.56E-02 1.07-1.89 0.90 5.04E-01 0.65-1.24 total MHC 0.85 1.85E-01 0.68-1.08 1.25 1.73E-01 0.91-1.72 1.00 9.84E-01 0.77-1.29 0.79 1.64E-01 0.57-1.10 total DHC 0.83 1.18E-01 0.65-1.05 0.87 3.96E-01 0.63-1.20 0.74 2.54E-02 0.57-0.96 0.79 1.62E-01 0.57-1.10 total THC 0.58 7.53E-05 0.44-0.76 0.68 4.93E-02 0.47-1.00 0.86 3.35E-01 0.64-1.16 1.09 6.46E-01 0.76-1.57 total GM3 0.73 1.22E-02 0.57-0.93 1.05 7.51E-01 0.76-1.45 0.98 8.95E-01 0.74-1.30 0.84 3.14E-01 0.60-1.18 total SM 1.12 3.68E-01 0.87-1.44 1.12 5.10E-01 0.80-1.57 1.01 9.45E-01 0.76-1.34 0.88 4.64E-01 0.62-1.24 total modCer 1.23 8.26E-02 0.97-1.55 2.11 1.10E-04 1.45-3.09 1.35 3.28E-02 1.02-1.77 0.74 6.06E-02 0.54-1.01 total LPAF 0.69 2.73E-03 0.54-0.88 0.94 7.09E-01 0.69-1.29 0.76 4.59E-02 0.58-0.99 0.77 1.12E-01 0.56-1.06 total LPC 0.75 1.49E-02 0.59-0.94 1.34 7.75E-02 0.97-1.86 0.94 6.66E-01 0.72-1.23 0.72 3.37E-02 0.53-0.97 total PC 1.32 2.62E-02 1.03-1.69 . 1.22 2.44E-01 0.87-1.69 1.17 2.72E-01 0.89-1.54 0.94 7.03E-01 0.68-1.30 total APC 0.72 6.18E-03 0.57-0.91 0.68 2.45E-02 0.48-0.95 0.80 8.97E-02 0.62-1.04 1.05 7.82E-01 0.76-1.45 total oddPC 0.90 3.90E-01 0.70-1.15 0.72 7.17E-02 0.51-1.03 0.67 4.57E-03 0.51-0.88 0.90 5.16E-01 0.65-1.24 total modPC 1.03 8.12E-01 0.81-1.30 0.86 3.46E-01 0.63-1.17 0.82 1.51E-01 0.63-1.07 0.92 6.03E-01 0.67-1.26 total PG 1.46 3.62E-03 1.13-1.88 1.82 1.98E-03 1.24-2.65 1.58 2.35E-03 1.18-2.11 0.90 5.63E-01 0.65-1.27 total PE 1.52 1.95E-03 1.17-1.97 1.77 3.45E-03 1.21-2.61 1.85 9.76E-05 1.36-2.52 1.05 7.94E-01 0.74-1.49 total PS 1.28 4.37E-02 1.01-1.62 1.06 7.32E-01 0.77-1.44 1.27 8.94E-02 0.96-1.67 1.28 1.67E-01 0.90-1.82 total PI 1.16 2.31E-01 0.91-1.48 1.71 3.95E-03 1.19-2.46 1.45 1.18E-02 1.09-1.94 0.84 3.24E-01 0.59-1.19 total CE 1.72 4.47E-05 1.33-2.23 1.79 3.79E-03 1.21-2.66 2.15 2.17E-06 1.56-2.94 1.20 3.06E-01 0.85-1.69 total modCE 1.39 7.50E-03 1.09-1.76 1.58 8.81E-03 1.12-2.24 1.87 2.46E-05 1.40-2.50 1.09 5.77E-01 0.80-1.50 total DG 1.81 6.89E-06 1.40-2.34 2.90 1.55E-06 1.88-4.47 2.11 3.65E-06 1.54-2.90 0.86 3.77E-01 0.62-1.20 total TG 1.83 7.31E-06 1.41-2.39 2.96 1.86E-06 1.90-4.63 1.92 6.23E-05 1.40-2.65 0.79 1.90E-01 0.56-1.12

COH 0.95 6.81E-01 0.76-1.20 1.68 1.87E-03 1.21-2.32 1.10 4.79E-01 0.85-1.43 0.58 3.61E-03 0.40-0.84

TABLE 5. Mean and standard deviation of total lipid classes in the cross sectional study¹

Whole population Diabetes NGT Pre-diabetes Obese Non obese

Lipid Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

SPH 24 36 24 39 24 36 22 27 28 43 21 29

Cer 5592 1368 5874 1499 5252 1121 5982 1509 5766 1413 5470 1325

MHC 4970 1327 4879 1375 4905 1255 5309 1389 4813 1336 5081 1312

DHC 5366 1256 5105 1242 5506 1265 5467 1201 5170 1171 5503 1297

THC 1654 450 1591 480 1705 447 1632 385 1569 411 1713 . 467

GM3 2243 527 2184 493 2240 520 2359 591 2166 536 2298 515

dhCer 746 274 833 322 659 187 823 304 807 307 704 240

modCer 1338. 297 1387 336 1260 242 1454 299 1372 264 1313 316

SM 592279 79684 593266 76072 586969 79577 604579 86028 600858 77708 586240 80686

LPAF 1330 323 1263 325 1368 306 1353 350 1248 277 1388 341

LPC 199545 41568 194788 44982 199212 35815 209126 47929 190513 37460 205903 43198

PC 1562108 215314 1587179 214773 1537647 196704 1581250 256271 1606897 232336 1530583 197018

APC 87869 13479 86123 13758 89813 13782 85898 11470 85205 12377 89744 13930 ON oddPC 84400 14176 81651 15776 86363 13450 84212 12186 83805 14330 84820 14087

modPC 29405 . 4521 28968 4467 29658 4441 29531 4834 29508 4578 29332 4490

PG 225 104 257 121 . 195 79 245 107 249 107 207 98

PE 27716 11867 32023 14035 24270 9490 28998 10482 31144 12904 25304 10455

PS 512 176 543 174 491 166 509 202 535 188 495 166

PI 18780 4728 19736 5072 17763 4311 19732 4642 19483 4966 18285 4499

COH 815797 312843 801769 285027 779481 326246 937907 299247 810777 283454 819331 332608

CE 5270370 1779881 6152137 2092911 4614118 1275268 5401563 1596374 5777241 1734084 4913592 1728247

modCE 200424 111120 241859 129427 166285 79834 215355 118732 219092 101893 187283 115623

DG 46810 27692 58618 34709 . 36053 14903 53796 28730 53862 25442 41846 28189

TG 687272 379552 838199 460193 541637 222163 798200 407374 786062 357886 617735 379767

data expressed as pmol/ml plasma

/ .

TABLE 6. Logistic regression of lipids against obesity, pre-diabetes and diabetes in the cross sectional study

Non-Obese vs. Obese (waist Pre-diabetes (IGT IFG) vs. circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid # Lipid Species ratio 95% CI p- value Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% CI p-value

1 Sph 18:1 1.22 0.97-1.54 8.77E-02 1.04 0.75-1.44 8.24E-01 0.83 0.65-1.07 1.56E-01 0.80 0.58-1.1 1.67E-01

2 dhCer 16:0 1.14 0.91-1.43 2.66E-01 1.23 0.9-1.67 1.96E-01 1.50 1.13-1.99 4.93E-03 1.07 0.77-1.48 6.83E-01

3 dhCer 18:0 1.66 1.28-2.15 1.08E-04 2.08 1.45-3 8.18E-05 2.37 1.7-3.32 4.23E-0? 0.99 0.7-1.41 9.71E-01

4 dhCer l8: l 1.07 0.84-1.36 5.67E-01 1.50 0.97-2.33 6.94E-02 1.54 1.13-2.10 5.91 E-03 1.08 0.74-1.57 7.06E-01

5 dhCer 20:0 1.20 0.96-1.51 1.14E-01 1.72 1.16-2.56 699E-03 1.16 0.9-1.49 2.63E-01 0.82 0.59-1.14 2.29E-01

6 dhCer 22:0 1.45 1.13-1.87 3.I5E-03 2.02 1.39-2.93 234E-04 1.80 1.32-2.44 1.96E-04 0.78 0.55-1.11 1.69E-01

7 dhCer 24:0 1.38 1.08-1.76 9.26E-03 1.80 1.25-2.58 1.40E-03 1.59 1.18-2.14 2.28E-03 0.78 0.55-1.11 1.67E-01

8 dhCer 24: l 1.40 1.1-1.79 6.06E-03 1.97 1.36-2 83 2.81E-04 1.61 1.21-2.15 1.26E-03 0.79 0.56-1.11 1.66E-01

9 Cer 16:0 0.94 0.74-1.19 5.83E-01 1.23 „ 0.88-1.73 2.29E-01 1.20 0.92-1.56 1.74E-01 0.91 0.65-1.27 5.71E-01

10 Cer 18:0 1.61 1.24-2.08 2.99E-04 1.73 1.18-2.53 4.71E-03 1.69 1.26-2.27 4.71E-04 1.04 0.75-1.45 8.23E-01

11 Cer 20:0 1.29 1-1.65 4.83E-02 1.63 1.12-2.38 1.09E-02 1.50 1.13-2 5.66E-03 0.95 0.68-1.33 7.70E-01

12 Cer 22:0 1.34 1.05-1.72 2.09E-02 1.51 1.04-2.19 2.95E-02 1.61 1.19-2.16 1.76E-03 1.01 ' 0.73-1.39 9.66E-01

13 Cer 24:0 1.13 0.89-1.44 2.99E-01 1.37 0.99-1.91 5.99E-02 1.26 0.96-1.67 9.77E-02 0.86 0.63-1.19 3.63E-01

14 Cer 24:1 1.18 0.92-1.52 1.86E-01 1.54 1.06-2.25 2.46E-02 ^• 1.41 1.06-1.88 2.02E-02 0.95 0.68-1.33 7.79E-01

15 MHC 16:0 1.01 0.8-1.28 9.23E-01 1.21 0.89-1.66 ^■ 2.28E-01 0.95 0.73-1.24 7.17E-01 0.76 0.55-1.06 1.05E-01

16 MHC 18:0 1.01 0.8-1.28 9.33E-01 1.25 0.91-1.73 1.68E-01 1.00 0.77-1.3 9.96E-01 0.80 0.57-1.12 1.85E-01

17 MHC 20:0 0.89 0.7-1.12 3.07E-01 1.25 0.91-1.72 1.68E-01 1.12 0.86-1.45 3.94E-01 0.89 0.64-1.24 4.93E-01

18 MHC 22::0 0.89 0.7-1.12 3.26E-01 1.21 0.88-1.66 2.43E-01 1.02 0.78-1.32 9.09E-01 0.82 0.59-1.13 2.26E-01

19 MHC 24:0 0.83 0.66-1.04 l.lOE-01 136 0.98-1.89 6.51E-02 1.05 0.81-1.36 7.36E-01 0.77 0.56-1.07 1.23E-01

20 MHC 24:1 0.80 0.63-1.01 6.55E-02 1.08 0.78-1.49 6.35E-01 0.91 0.7-1.19 4.89E-01 0.86 0.62-1.19 3.65E-01

21 DHC 16:0 0.8S 0.67-1.07 1.72E-01 0.87 0.63-1.21 4.14E-01 0.75 0.58-0.98 3.26E-02 0.79 0.57-1.09 1.52E-01

22 DHC 18:0 1.11 0.87-1.41 3.89E-01 0.91 0.64-1.28 5.78E-01 0.71 0.53-0.94 1.78E-02 0.79 0.57-1.11 1.72E-01

23 DHC 20:0 0.89 0.7-1.14 3.51E-01 1.06 0.76-1.48 7.37E-01 0.82 0.62-1.09 1.66E-01 0.77 0.55-1.09 1.39E-01

24 DHC 22:0 0.83 0.65-1.05 1.19E-01 1.04 0.75-1.44 8.33E-01 0.85 0.65-1.12 2.42E-01 0.83 0.59-1.16 2.77E-01

25 DHC 24:0 0.80 0.64-1.02 7.21E-02 0.97 0.71-1.33 ^• 8.62E-01 1.02 0.78-1.33 8.86E-01 1.01 0.73-1.39 9.65E-01

26 DHC 24:1 0.82 0.64-1.04 9.79E-02 0.83 0.59-1.16 2.79E-01 0.69 0.52-0.91 7.52E-03 0.79 0.56-1.11 1.76E-01

27 THC 16:0 0.66 0.51-0.86 1.83E-03 0.79 0.55-1.13 1.95E-01 0.89 0.66-1.18 4.12E-01 0.99 0.69-1.43 9.62E-01

28 THC 18:0 0.66 0.51-0.85 8E-03 0.81 0.57-1.15 2.41E-01 0.92 0.69-1.21 5.48E-01 1.10 0.78-1.54 5.90E-01

29 THC 20:0 0.67 0.52-0.87 2.45E-03 0.85 0.59-1.21 3.57E-01 0.85 0.65-1.11 2.29E-01 .0.99 0.7-1.41 9.73E-01

30 THC 22:0 0.64 0.49-0.83 6.46E-04 0.60 0.42-0.86 5.77E-03 0.82 0.62-1.1 1.82E-01 1.19 0.84-1.71 3.30E-01

31 THC 24:0 0.57 0.44-0.74 1.72E-05 0.65 0.46-0.93 1.64E-02 0.96 0.72-1.28 7.82E-01 1.27 0.92-1.76 1.53E-01

32 THC 24:1 0.53 0.4-0.7 5.92E-06 0.65 0.44-0.96 3.15E-02 0.87 0.64-1.17 3.50E-01 1.13 0.78-1.64. 5.14E-01

33 GM3 16:0 0.82 0.64-1.05 1.15E-01 0.90 0.64-1.24 5.09E-01 0.84 0.63-1.12 2.44E-01 0.89 0.63-1.24 4.77E-01

34 GM3 18:0 0.7 0.62-1.01 6:05E-02 1.01 0.73-1.41 9.40E-01 0.84 0.64-1.11 2.18E-01 0.78 0.56-1.1 1.55E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid # Lipid Species ratio 95% CI p-value Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% CI p-value

35 GM3 20:0 0.88 0.69-1.12 2.88E-01 1.29 0.93-1.79 1.30E-01 1.09 0.83-1.45 5.33E-01 0.81 0.59-1.13 2.20E-01

36 GM3 22:0 0.89 0.7-1.12 3.15E-01 1.12 0.82-1.53 4.74E-01 1.20 0.91-1.58 1.92E-01 0.91 0.66-1.27 5.89E-01

37 GM3 24:0 0.65 0.5-0.83 5.17E-04 1.09 0.8-1.5 5.81E-01 0.92 0.7-1.21 5.39E-01 0.76 0.54-1.07 1.20E-01

38 GM3 24:1 0.64 0.48-0.85 1.82E-03 1.09 0.79-1.49 6.06E-01 1.10 0.84-1.44 4.80E-01 0.93 0.64-1.36 7.08E-01

39 SM 14:0 1.04 0.82-1.31 7.60E-01 1.12 0.8-1.58 5.09E-01 1.05 0.8-1.38 7.09E-01 0.98 0.65-1.47 9.13E-01

40 SM 15:0 0.95 0.76-1.19 6.44E-01 0.95 0.73-1.23 6.91E- 1 0.99 0.74-1.32 9.27E-01 1.03 0.7-1.51 8.91E-01

41 SM 16:0 1.07 0.85-1.34 5.73E-01 1.15 0.59-2.26 6.75Ε-0Γ 0.86 0.65-1.13 2.79E-01 0.82 0.48-1.4 4.66E-01

42 SM 16:1 1.07 0.82-1.4 5.99E-01 0.95 0.67-1.37 7.99E-01 0.83 0.61-1.12 2.14E-01 0.82 0.56-1.19 2.93E-01

43 SM 18:0 1.21 0.95-1.54 1.21E-01 iji 0.94-1.82 1.06E-01 1.16 0.88-1.52 2.95E-01 0.92 0.67-1.28 6.36E-01

44 SM 18:1 1.34 1.03-1.74 2.86E-02 0.98 0.69-1.41 9.26E-01 0.96 0.72-1.29 7.88E-01 1.00 0.71-1.4 9.91E-01

45 SM 20:0 1.13 0.9-1.43 2.87E-01 1.01 0.74-1.36 9.72E-01 0.92 0.71-1.19 5.28E-01 0.87 0.62-1.21 4.03E-01

46 SM 20:1 1.23 0.96-1.58 l.OlE-01 131 0.93-1.84 1.21E-01 1.24 0.92-1.67 1.50E-01 0.97 0.7-1.33 8.34E-01

47 SM 22:0 1.03 0.81-1.3 8.06E-01 1.07 0.78-1.47 6.87E-01 1.02 0.77-1.34 9.02E-01 0.94 0.68-1.3 7.00E-01

48 SM 22:1 1-36 1.07-1.73 1.17E-02 0.99 0.71-1.37 9.56E-01 1.08 0.82-1.42 6.06E-01 1.04 0.75-1.44 8.37E-01

49 SM 24:0 1.03 0.82-1.29 8.03E-01 1.27 0.93-1.73 1.36E-01 1.29 0.99-1.7 6.08E-02 0.92 0.66-1.29 6.41E-01

50 SM 24:1 1.29 1.01-1.66 4.28E-02 1.05 0.77-1.45 7.50E-01 0.97 0.73-1.29 8.45E-01 0.84 0.59-1.18 3.17E-01

51 SM 24:2 1.17 0.92-1.49 2.11E-01 1.09 0.79-1.5 6.06E-01 1.03 0.77-1.37 8.41E-01 0.94 0.68-1.31 7.24E-01

52 SM26:1 0.77 0.61-0.98 3.28E-02 0.86 0.63-1.18 3.58E-01 0.88 0.68-1.15 3.52E-01 0.98 0.7-1.36 8.87E-01

53 modCer 576.5/7.7 1.08 0.85-1.36 5.24E-01 2.05 1.42-2.96 1.40E-04 1.46 1.11-1.91 6.51E-03 0.74 0.52-1.05 9.48E-02

54 modCer 614.6/5.7 0.70 0.55-0.9 4.43E-03 0.95 0.69-1.32 7.70E-01 0.66 ' 0.51-0.87 3.00E-03 0.70 0.5-0.98 3.56E-02

55 modCer 632.6/9.2 0.77 0.61-0.98 3.25E-02 1.15 0.84-1.58 3.76E-01 1.05 0.81-1.35 7.12E-01 0.86 0.62-1.2 3.82E-01

56 modCer 651.6/7.6 0.82 0.65-1.04 9.76E-02 1.24 0.9-1.73 1.90E-01 0.86 0.66-1.11 2.53E-01 0.67 0.47-0.96 2.94E-02

57 modCer 731.6/6.2 1.05 0.84-1.33 6.64E-01 1.59 1.12-2.27 9.85E-03 1.20 0.92-1.57 1.83E-01 0.78 0.56-1.09 1.43E-01

58 modCer 766.6/7.2 0.81 0.65-1.03 8.27E-02 1.43 1.02-1.99 3.60E-02 0.92 0.71-1.18 4.98E-01 0.65 0.46-0.93 1.78E-02

59 modCer 769.6/8.0 1.03 0.82-1 29 8.01E-01 1.59 0.85-3 1.48E-01 0.92 0.72-1.19 5.42E-01 0.64 0.38-1.09 1.02E-01

60 modCer 798.7 7.3 0.82 0.64-1.04 9.69E-02 1.22 0.89-1.68 2.20E-01 0.80 0.61-1.04 9.64E-02 0.68 0.48-0.96 2.82E-02

61 modCer 875.7/9.2 1.50 1.17-1.93 133E-03 ^• 1.78 1.21-2.61 3-26E-03 IJ8 1.04-1.83 2.67E-02 0.85 0.62-1.16 2.94E-01

62 modCer 883 8/7.8 0.95 0.75-1.21 6.84E-01 1.74 1.22-2.48 2.23E-03 1.14 0.87-1.49 3.37E-01 0.72 0.51-1.01 5.77E-02

63 modCer 886.8/9.1 1 J5 1.06-1.72 1.39E-02 1.90 1.32-2.75 S.74E-04 1.75 1.3-2.34 1.91E-04 0.93 ^• 0.67-1.3 6.65E-01

64 modCer 910.8/9.0 1.13 0.89-1.44 3.07E-01 1.36 0.96-1.93 8.18E-02 1J0 0.99-1.7 5.88E-02 0.94 0.68-1.31 7.31E-01

65 modCer 921.8/9.1 1.06 0.85-1.33 5.93E-01 2.12 1.43-3.15 1.72E-04 1.56 1.18-2.07 1.74E-03 0.90 0.67-1.21 4.78E-01

66 oddPC 29:0 0.97 0.76-1.24 8.23E-01 0.90 0.64-1.27 5.53E-01 0.70 0.53-0.93 1.46E-02 0.81 0.58-1.14 2.24E-01

67 oddPC 31:0 1.26 0.99-1.6 5.98E-02 0.97 0.7-1.35 8.63E-01 0.85 0.65-1.12 2.46E-01 0.94 0.69-1.27 6.71E-01

68 oddPC 31:l 1.18 0.92-1.5 1.93E-01 0.90 0.65-1.26 5.56E-01 0.75 0.57-0.99 3.91E-02 0.85 0.62-1.18 3.35E-01

69 oddPC 33:0 0.62 0.49-0.8 1.50E-04 0.69 0.5-0.96 2.71E-02 0.61 0.47-0.81 5.82E-04 0.86 0.62-1.2 3.75E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid Lipid Species ratio 95% CI p-value Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% CI p-value

70 oddPC 33:l S 0.98-1.59 7.11E-02 0.97 0.7-1.36 8.76E-01 0.82 0.63-1.06 1.35E-01 0.89 0.65-1.21 4.61E-01

71 oddPC 33:2 1.02 0.81-1.29 8.74E-01 1.06 0.75-1.49 7.46E-01 0.98 0.76-1.26 8.66E-01 1.01 0.75-1.37 9.32E-01

72 oddPC 35:0 1.04 0.83-1.31 7.22E-01 0.94 0.68-1.29 6.90E-01 0.83 0.63-1.07 1.52E-01 0.89 0.65-1.23 4.80E-01

73 oddPC 35:l 0.99 0.78-1.25 9.21E-01 0.91 0.65-1.26 5.72E-01 0.80 0.62-1.05 1.08E-01 0.92 0.67-1.26 6.13E-01

74 oddPC 35:2 0.89 0.7-1.13 3.39E-01 0.68 0.49-0.94 1.97E-02 0.60 0.45-0.8 4.89E-04 0.92 0.67-1.28 6.29E-01

75 oddPC 35:3 0.81 0.64-1.02 7.61E-02 0.64 0,45-0.91 1.32E-02 0.76 0.59-0.98 3.25E-02 1.02 0.75-1.38 9.05E-01

76 oddPC 35:4 1.08 0.86-1.36 4.98E-01 1.03 0.76-1.41 8.44E-01 0.90 0.68-1.19 ; 4.54E-01 0.92 0.68-1.25 5.91E-01

77 oddPC 37:2 0.92 0.72-1.18 4.96E-01 0.76 0.54-1.06 1.02E-01 0.68 0.51-0.92 1.08E-02 0.92 0.66-1.27 6.04E-01

78 oddPC 37:3 1.24 0.97-1.59 8.74E-02 0.67 0.47-0.96 2.83E-02 0.67 0.5-0.89 5.13E-03 0.95 0.7-1.3 7.59E-01

79 oddPC 37:4 0.96 0.76-1.21 7.51E-01 0.65 0.47-0.89 7.83E-03 0.65 0.5-0.86 2.49E-03 1.02 0.75-1.4 8.99E-01

80 oddPC 37:5 0.85 0.67-1.08 1.77E-01 1.00 0.71-1.41 9.98E-01 0.86 0.66-1.12 2.71E-01 0.95 0.69-1.3 7.37E-01

81 oddPC 37:6 0.8S 0.66-1.08 1.85E-01 1.24 0.88-1.75 2.26E-01 1.03 0.79-1.35 8.01E-01 0.94 0.68-1.3 7.13E-01

82 PC 28:0 1.18 0.94-1.5 1.53E-01 1.18 0.87-1.6 2.91E-01 0.92 0.7-1.22 5.82E-01 0.83 0.61-1.14 2.42E-01

83 PC 30:2 1.1 1 0.85-1.45 4.45E-01 0.92 0.64-1.32 6.50E-01 0.74 0.54-1.01 5.42E-02 0.76 0.52-1.1 1.45E-01

84 PC 32:0 0.96 0.76-1.22 7.53E-01 1.09 0.79-1.51 6.02E-01 1.18 0.9-1.54 2.28E-01 1.03 0.76-1.39 8.59E-01

85 PC 32:1 1.61 1.24-2.08 3.10E-04 1.63 1.12-2.36 1.02E-02 1.52 1.13-2.05 5.99E-03 0.98 0.71-1.35 8.98E-01

86 PC 32:2 1.47 1.13-1.92 4.73E-03 1.18 0.83-1.68 3.46E-01 0.97 0.72-1.31 8.36E-01 0.84 0.6-1.17 3.07E-01

87 PC 34:0 0.81 0.64-1.03 8.08E-02 1.12 0.82-1.55 4.77E-01 0.89 0.68-1.16 3.77E-01 0.78 0.57-1.08 1.35E-01

88 PC 34:1 1.15 0.91-1.45 2.39E-01 1.09 0.8-1.48 5.92E-01 1.17 0.89-1.52 2.55E-01 1.04 0.76-1.44 7.94E-01

89 PC 34:2 1.18 0.93-1.49 1.64E-01 0.87 0.64-1.19 3.94E-01 0.84 0.64-1.1 2.05E-01 0.93 0.68-1.26 6.32E-01

90 PC 34:3 1.20 0.94-1.55 1.45E-01 1.24 0.87-1.76 2.33E-01 1.05 0.78-1.41 7.58E-01 0.88 0.64-1.21 4.36E-01

91 PC 36:1 1.13 0.9-1.43 2.88E-01 0.98 0.72-1.35 9.07E-01 0.82 0.63-1.07 1.42E-01 0.87 0.65-1.18 3.77E-01

92 PC 36:2 1.09 0.86-1.38 4.57E-01 1.01 0.74-1.38 9.43E-01 0.86 0.65-1.14 2.98E-01 0.83 0.61-1.14 2.55E-01

93 PC 36:3 1.39 1.09-1.78 8.15E-03 0.94 0.67-1.32 7.28E-01 0.93 0.71-1.23 6.29E-01 0.94 0.69-1.29 7.16E-01

94 PC 36:4 1.27 1.01-1.6 4.06E-02 0.96 0.71-1.31 8.17E-01 1.14 0.87-1.5 3.46E-01 1.13 0.83-1.54 4.34E-01

95 PC 36:5 0.99 0.79-1.25 9.48E-01 1J0 0.94-1.8 1.12E-0I 1.29 0.98-1.7 6.60E-02 1.04 0.77-1.4 7.91E-01

9 PC 38:2 1.24 0.97-1.59 8.54E-02 1.14 0.82-1.59 4.23E-01 0.98 0.74-1.29 8.81E-01 0.80 0.58-1.12 1.98E-01

97 PC 38:3 1.33 1.05-1.7 1.94E-02 0.90 0.65-1.24 5.12E-01 0.84 0.63-1.11 2.21E-01 0.95 0.71-1.29 7.62E-01

98 PC 38:4 1J7 1.08-1.74 9.54E-03 1.12 0.82-1.53 4.72E-01 1.15 0.86-1.52 3.40E-01 0.95 0.69-1.32 7.74E-01

99 PC 38:5 1.03 0.82-1.3 7.75E-01 1.14 0.82-1.59 4.22E-01 1.04 0.8-1.37 7.51E-01 0.95 0.7-1.29 7.53E-01

100 PC 38:6 0.84 0.67-1.07 1.55E-01 1J9 0.99-1.95 5.96E-02 1.28 0.98-1.65 6.70E-02 1.04 0.75-1.43 8.27E-01

101 PC 39:7 0.56 0.43-0.73 1.86E-05 1.01 0.72-1.42 9.62E-01 1.06 0.8-1.41 6.90E-01 1.07 0.77-1.49 6.87E-01

102 PC 40:5 1.24 0.98-1.57 7.73E-02 1.56 1.1-2.22 1.32E-02 133 1.01-1.75 4.23E-02 0.89 0.64-1.23 4.66E-01

103 PC 40:6 0.98 0.78-1.24 8.73E-01 1.65 1.17-2.33 4.70E-03 1.44 1.11-1.88 6.68E-03 0.98 0.71-1.37 9.27E-01

104 PC 40:7 0.68 0.53-0.88 3.41E-03 1.16 0.82-1.65 3.93E-01 1.04 0.79-1.37 7.70E-01 0.96 0.67-1.37 8.17E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs, Pre-diabetes (IGT IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid # Lipid Species ratio 95% CI p-valiie Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% CI p-value

105 PC 44:12 0.82 0.65-1.04 9.48E-02 0.71 0.51-0.9 4.29E-02 0.78 0.6-1.01 6.04E-02 1.04 0.75-1.42 8.27E-01

106 APC 30:0 0.84 0.66-1.06 1.35E-01 0.79 0.58-1.08 1.47E-01 0.74 0.56-0.96 2.31E-02 0.95 0.69-1.31 7.69E-01

107 APC 30:1 0.83 0.62-1.11 2.05E-01 0.92 0.6-1.39 6.89E-01 0.73 0.54-0.99 4.52E-02 0.86 0.6-1.21 3.78E-01

108 APC 32:0 0.83 0.66-1.05 1.15E-01 0.71 0.51-0.98 3.85E-02 0.99 0.77-1.28 9.55E-01 1.22 0.9-1.67 2.02E-01

109 APC 32: 1 0.64 0.49-0.82 S-28E-04 0.94 0.67-1.33 7.38E-01 1.05 0.8-1.37 7.19E-01 1.10 0.82-1.48 5.35E-01

110 APC 34:0 · 1.00 0.8-1 26 9.80E-01 0.67 0.48-0.92 1.45E-02 0.80 0.61-1.04 l.OlE-01 1.12 0.81-1.54 4.82E-01

111 APC 34:1 0.62 0.48-0.8 1.90E-04 0.52 0.36-0.76 6.78E-04 0.82 0.62-1.07 1.46E-01 1J0 0.94-1.81 1.13E-01

112 APC 34:2 0.59 0.46-0.76 2.66E-05 0.73 0.52-1.01 5.73E-02 0.67 0.51-0.88 4_»E-03 0.88 0.64-1.22 4.48E-01

113 APC 34:4 0.62 0.48-0.81 3.82E-04 0.58 0.39-0.85 5.05E-O3 0.69 0.52-0.92 1.15E-02 1.06 0.76-1.48 7.27E-01

114 APC 36:0 0.98 0.78-1.23 8.48E-01 0.79 0.58-1.08 1.41E-01 0.90 0.69-1.17 4.25E-01 1.08 0.79-1.48 6.37E-01

115 APC 36:1 0.78 0.61-0.99 4.46E-02 0.62 044-0.88 6.91E-03 0.80 0.61-1.05 l.OlE-01 1.15 0.83-1.59 4.14E-01

116 APC 36:2 0.59 0.46-0.77 6.68E-05 0.53 0.37-0.78 1.07E-03 0.65 0.49-0.86 3.11E-03 1.03 0.74-1.43 8.61E-01

117 APC 36:3 0.59 0.46-0.77 852E-05 0.67 0.47-0.96 2.84E-02 0.61 0.46-0.82 8.36E-04 0.84 0.61-1.17 3.02E-01

118 APC 36:4 0.85 0.68-1.07 1.64E-01 0.72 0.52-1 4.88E-02 0.86 0.67-1.11 2.54E-01 1.12 0.8-1.55 5.19E-01

119 APC 36:5 0.86 0.69-1.08 2.03E-01 0.92 0.67-1.25 5.87E-01 0.92 0.71-1.19 5.41E-01 0.99 0.72-1.37 9.58E-01

120 APC 38:2 0.66 0.51-0.83 5.79E-04 0.77 0.54-1.08 1.34E-01 0.79 0.61-1.02 6.88E-02 0.96 0.7-1.32 7.95E-01

121 APC 38:3 1.08 0.86-1.35 5.17E-01 0.89 0.63-1.26 5.20E-01 0.93 0.72-1.21 6.07E-01 0.99 0.75-1.32 9.65E-01

122 APC 38:4 1.04 0.83-1.31 7.34E-01 0.71 0.51-0.98 3.95E-02 0.85 0.66-1.09 2.01E-01 1.06 0.76-1.47 7.39E-01

123 APC 38:5 0.80 0.63-1.01 5.70E-02 0.72 0.53-0.99 4.63E-02 0.82 0.63-1.07 1.44E-01 1.04 0.75-1.45 8.17E-01

124 APC 38:6 0.66 0.52-0.84 7.68E-04 0.98 0.71-1.36 9.20E-01 1.01 0.77-1.33 9.33E-01 1.05 0.77-1.43 7.55E-01

125 APC 40:7 1.07 0.85-1.34 5.65E-01 1.19 0.86-1.65 298E-01 1.26 0.96-1.64 9.10E-02 0.96 0.69-1.34 8.11E-01

126 LPAF 16:0 0.71 0.56-0.9 4.99E-03 1.05 0.77-1.43 7.67E-01 0.81 0.62-1.05 1.15E-01 0.74 0.54-1.03 7.18E-02

127 LPAF 18:0 0.73 0.57-0.93 9.94E-03 0.81 0 59-1 11 1.84E-01 0.67 0.51-0.89 5.05E-03 0.80 0.57-1.12 1.88E-01

128 LPAF 18:1 0.66 0.52-0.84 7.64&04 0.98 0.71-1.35 8.85E-01 0.83 0.63-1.08 1.67E-01 0.81 0.59-1.11 1.81E-01

129 LPAF 22:0 0.69 0.54-0.89 3.76E-03 1.01 0.74-1.38 9.65E-01 0.68 0.52-0.9 6.53E-03 0.64 0.45-0.92 1.68E-02

130 LPAF 22:1 1.19 0.94-1.49 1.42E-01 0.68 0.49-O.95 2.26E-02 0.84 0.65-1.09 1.91E-01 1.20 0.86-1.67 2.74E-01

131 LPAF 24:1 0.75 0.58-0.96 2.49E-02 0.98 0.75-1.3 9.13E-01 0.71 0.53-0.95 1.99E-02 0.64 0.42-0.97 3.50E-02

132 LPAF 24:2 059 0.46-0.76 3.34E-05 0.78 0.57-1.08 1.34E-01 0.65 0.49-0.85 1.95E-03 0.79 0.55-1.14 2.07E-01

133 LPC 14:0 1.20 0.95-1.51 1.32E-01 1.53 1.09-2.13 1.29E-02 1.01 0.77-1.32 9.57E-01 0.72 0.52-1 5.16E-02

134 LPC 15:0 0.95 0.76-1.2 6.93E-01 0.99 0.72-1.35 9.28E-01 0.68 0.51-0.9 6.76E-03 0.77 0.57-1.03 8.15E-02

135 LPC 16:0 0.90 0.72-1.13 3.56E-01 1.45 1.05-2 2.37E-02 1.11 0.85-1.44 4.50E-01 0.77 0.57-1.03 8.27E-02

136 LPC 16:1 1.04 0.82-1.31 7.56E-01 1.57 1.13-2.19 7J3E-03 1.15 0.88-1.5 3.09E-01 0.79 0.58-1.07 1.21E-01

137 LPC 18:0 0.80 0.63-1 4.86E-02 151 1.08-2.12 1.65E-02 0.96 0.74-1.25 7.73E-01 0.67 0.49-0.92 1.32E-02

138 LPC 18:1 0.60 0.47-0.77 7.07E-05 1.12 0.81-1 56 4.93E-01 0.76 0.57-1.01 5.57E-02 0.70 0.51-0.97 3.11E-02

139 LPC 18:2 0.54 0.42-0.71 6.78E-06 0.79 0.55-1.15 2.17E-01 0.60 0.44-0.82 0E-O3 0.72 0.51-1.03 7.22E-02

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

ipid # Lipid Species ratio 95% CI p-value Odds ratio 95% CI p-value ratio 95% a p-value ratio 95% CI p-value

140 LPC 20:0 0.49 0.38-0.64 1.27E-07 1.11 0.79-1.56 5.54E-01 0.80 0.61-1.06 1.21E-01 0.72 0.52-1 5.11E-02

141 LPC 20:1 0.51 0.4-0.67 7.96E-07 1.15 0.82-1.61 4.28E-01 0.75 0.56-1 4.72E-02 0.67 0.47-0.95 2.36E-02

142 LPC 20:1 0.53 0.41-0.69 1.43E-06 0.96 0.69-1.35 8.31E-01 0.71 0.54-0.95 2.01E-02 0.72 0.51-1 5.15E-02

143 LPC 20:2 0.66 0.51-0.84 6.47E-04 1.22 0.86-1.71 2.63E-01 0.79 0.6-1.04 9.85E-02 0.70 0.51-0.96 2.78E-02

144 LPC 20:3 1.10 0.88-1.39 3.99E-01 1.11 0.8-1.55 5.24E-01 0.94 0.72-1.23 6.41E-01 0.83 0.62-1.12 2.33E-01

145 LPC 20:4 0.86 0.68-1.08 1.98E-01 1.00 0.74-1.35 9.86E-01 0.90 0.68-1.2 4.74E-01 0.87 0.63-1.19 3.82E-01

146 LPC 20:5 0.80 0.63-1.01 5.84E-02 1.25 0.9-1.72 1.79E-01 1.10 0.84-1.45 4.81E-01 0.95 0.71-1.28 7.50E-01

147 LPC 22:0 0.58 0.45-0.75 2.86E-05 0.87 0.62-1.21 3.97E-01 0.74 0.56-0.98 3.59E-02 0.79 0.56-1.1 1.60E-01

148 LPC 22:6 0.66 0.51-0.84 6.95E-04 1.51 1.09-2.11 1.42E-02 133 1.01-1.75 4.04E-02 0.92 . 0.66-1.28 6.08E-01

149 LPC 24:0 0.74 0.53-1.02 6.52E-02 1.08 . 0.81-1.46 5.93E-01 1.03 0.8-1.32 8.24E-01 0.85 0.51-1.41 5.23E-01

150 modPC.506.3/3.4 0.87 0.7-1.1 2.47E-01 1.21 0.87-1.68 2.59E-01 0.96 0.74-1.25 7.84E-01 0.77 0.57-1.06 l.lOE-01

151 modPC.512.3/1.7 0.89 0.71-1.11 3.10E-01 0.81 0.6-1.1 1.82E-01 1.07 0.82-1.39 6.30E-01 1.22 0.89-1.66 2.19E-01

152 modPC 536.3/3.5 0.70 0.53-0.93 1.29E-02 0.83 0.57-1.2 3.12E-01 0.56 0.4-0.8 1.27E-03 0.66 0.44-0.98 4.19E-02

153 modPC.538.3/4.1 0.57 0.45-0.74 2E-05 0.84 0.6-1.18 3.22E-01 0.58 0.44-0.78 2.54E-04 0.69 0.5-0.97 3.07E-02

154 modPC.594.4/3.1 131 1.04-1.66 2.40E-02 0.82 0.6-1.12 2.14E-01 1.23 0.95-1.6 1.17E-01 1.41 1.01-1.99 4.59E-02

155 . modPC.608.4/4.0 1.28 1.01-1.62 3.96E-02 1.13 0.83-1.54 4.40E-01 1.51 1.15-1.98 3.02E-03 1.27 0.91-1.77 1.52E-01

156 modPC.610.4/1.7 1.26 0.99-1.59 5.66E-02 0.91 0.68-1.21 5.07E-01 134 1.02-1.76 3.42E-02 1.41 0.99-2.01 5.91E-02

157 modPC.622.4/4.0 1.13 0.9-1.42 2.87E-01 1.04 0.78-1.38 7.85E-01 1.09 0.83-1.43 5.30E-01 0.96 0.71-1.31 8.07E-01

158 modPC.633.4/4.6 0.96 0.75-1.23 7.46E-01 0.93 0.67-1.29 6.69E-01 0.61 0.45-0.82 1.24E-03 0.70 0.49-1.01 5.37E-02

159 modPC.636.4/3.6 1.59 1.25-2.02 1.88E-04 0.91 0.66-1.25 5.58E-01 135 1.03-1.77 3.14E-02 130 0.94-1.81 1.12E-01

160 modPC.645.4/4.4 1.17 0.9-1:52 2.40E-01 0.97 0.68-1.37 8.59E-01 0.70 0.51-0.96 2.59E-02 0.77 0.53-1.11 1.61E-01

161 modPC.650.4/3.8 1.28 1.02-1.62 3.33E-02 ,0.87 0.64-1:18 3.61E-01 1.24 0.95-1.61 1.13E-01 130 0.93-1.8 1.19E-01

162 modPC.650.4/3.9 1.28 1.01-1.61 3 74E-02 0.81 0.6-1.1 1.78E-01 1.26 0.97-1.64 8.91E-02 1.42 1.02-1.96 3.75E-02

163 modPC.650.4/4.4 1.22 0.97-1.54 9.13E-02 0.72 0.52-0.99 4.05E-02 0.78 0.6-1.02 6.51E-02 1.08 0.78-1.5 6.29E-01

164 modPC.664.4/4.3 1.58 1.24-2.01 2.27E-04 0.99 0.73-1.34 9.27E-01 1.27 0.97-1.66 8.25E-02 1.18 0.85-1.63 3.35E-01

165 modPC.666.4 2.9 1.21 0.96-1.53 1.08E-01 0.98 0.72-1.33 9.04E-01 1.26 0.97-1.64 8.65E-02 1.20 0.86-1.68 2.86E-01

166 modPC.678.4/5.0 1.07 0.84-1.37 5.58E-01 1.21 0.87-1.68 2.64E-01 0.79 0.6-1.04 9.21E-02 0.69 0.49-0.98 3.84E-02

167 modPC.690.4/4.3 1.25 1-1.58 5.33E-02 0.89 0.65-1.22 4.53E-01 1.19 0.92-1.54 1.77E-01 1.20 0.87-1.65 2.58E-01

168 modPC.691.4/5.1 1.11 0.88-1.4 3.88E-01 0.95 0.7-1.29 7.34E-01 0.95 0.73-1.24 6.94E-01 0.97 0.71-1.34 8.71E-01

169 modPC.703.5/4.5 1.18 0.94=1.49 1.57E-01 0.87 0.64-1.18 3.73E-01 1.12 0.86-1.45 3.96E-01 1.19 0.86-1.66 2.87E-01

170 raodPC.704.5/3.8 1.03 0.82-1.29 7.95E-01 0.81 0.61-1.08 1.56E-01 1.09 0.84-1.41 5.31E-01 137 0.98-1.93 6.51E-02

171 modPC.706.5/4.2 1.20 0.96-1.52 1.15E-01 0.75 0.55-1.03 7.17E-02 1.07 0.82-1.39 6.30E-01 1.28 0.93-1.75 1.27E-01

172 modPC.720.5/4.5 1.08 0.86-1.36 4.96E-01 0.69 0.51-0.95 2.23E-02 1.04 0.8-1.36 7.57E-01 1.28 0.95-1.74 l.lOE-01

173 modPC.736.5/5.7 0.82 0.65-1.03 9.53E-02 0.84 0.62-1.15 2.88E-01 0.87 0.68-1.13 2.99E-01 1.01 0.73-1.4 9.55E-01

174 modPC.743.5/6.0 0.62 0.48-0.79 1.07E-04 0.59 0.42-0.83 2.55E-03 0.57 0.43-0.76 1.13E-04 0.91 0.65-1.27 5.81E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) circumfrcnce) NGT vs. Pre-diabetes (IGT IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

175 modPC.752.5/5.7 1.17 0.93-1.48 1.89E-01 1.44 1.04-2 2.87E-02 1.12 0.85-1.48 4.18E-01 0.87 0.64-1.2 4.09E-01

176 modPC.772.5/5.4 1.22 0.97-1.54 9.24E-02 0.65 0.47-0.91 1.14E-02 1.05 0.8-1.39 7.04E-01 1J2 0.97-1.8 · 7.70E-02

177 modPC.773.6/6.5 0.88 0.7-1.12 3.00E-01 0.71 0.51-0.98 3.51E-02 0.56 0.42-0.75 1.30E-04 0.87 0.63-1.2 3.92E-01

178 modPC.788.6/5.2 1.20 095-1.51 1.27E-01 0.68 0.48-0.96 2.79E-02 1.08 0.82-1.41 5.81E-01 1.34 0.97-1.83 7.19E-02

179 modPC.801.6/6.8 1.18 0.93-1.49 1.67E-01 1.02 0.75-1.4 9.00E-01 1.01 0.78-1.32 9.36E-01 0.95 0.69-1.31 7.39E-01

180 modPC.816.6/5.5 1.22 0.96-1.54 l .OOE-01 0.72 0.52-1 4.68E-02 0.99 0.75-1.3 9.28E-01 1.23 0.89-1.71 2.12E-01

181 modPC 827.7/6.8 1.06 0.84-1.33 6.51E-01 0.86 0.62-1.19 3.55E-01 0.70 0.54-0.92 9.96E-03 0.83 0.61-1.13 2.41E-01

182 modPC.828.6/5.8 1.02 0.81-1.29 8.54E-01 1.07 0.79-1.47 6.56E-01 1.04 0.8-1.34 7.77E-01 0.96 0.69-1.34 8.25E-01

183 modPC.843.6/7.2 0.66 0.52-0.84 7.52E-04 1.02 0.74-1.4 9.06E-01 1.02 0.79-1.33 8.60E-01 0.94 0.68-1 29 6.86E-01

184 modPC.866.6/7.2 0.69 0.54-0.89 3.68E-03 0.84 0.61-1.17 3.02E-01 0.99 0.76-1.27 9.10E-01 1.03 0.76-1.39 8.60E-01

18S modPC.877.6/6.0 0.72 0.56-0.92 8.48E-03 1.15 0.84-1.58 3.78E-01 1.12 0.86-1.45 4.09E-01 1.08 0.78-1.51 6.35E-01

186 modPC 879.1/6.1 0.66 0.52-O.85 1.20E-O3 0.98 0.71-1.36 9.25E-01 0.88 0.67-1.15 3.49E-01 0.96 0.69-1.34 8.15E-01

187 PG 16:0 18:1 1.14 0.9-1.45 2.77E-01 1J8 0.97-1.95 7.09E-02 I JI 1-1.72 5.41E-02 0.97 0.7-1.34 8.43E-01

188 PG 16:1 18:1 1.23 0.97-1.55 9.11E-02 1.66 1.15-2.39 6.50E-O3 1.11 0.86-1.45 4.16E-01 0.73 0.51-1.04 7.95E-02

189 PG 18:0 18:1 1.47 1.14-1.89 3.00E- 3 1.80 1.24-2.6 1.94E-03 lil 1.14-2.01 4.51E-03 0.86 0.6-1.22 3.85E-01

190 PG 18:1 18:1 1.66 1.27-2.16 1.86E-04 1.74 1.21-2.5 2.85E-03 1.79 1.32-2.45 2.21E-04 0.98 0.69-1.39 9.16E-01

191 PE 32:0 1 J9 1.09-1.77 8.07E-O3 1-36 0.97-1.9 7.51E-02 1.52 1.14-2.02 4.13E-03 1.06 0.76-1.48 7.36E-01

192 PE 32:1 1.80 1.37-2.37 2.34E-05 1.60 1.1-2.33 1.33E-02 1.77 1.29-2.44 4.44E-04 1.07 0.75-1.52 7.08E-01

193 PE 32:2 1.79 1.35-2.36 4.80E-05 1.50 1.01-2.23 4.33E-02 1.05 0.79-1.4 7.29E-01 0.73 0.51-1.05 8.82E-02

194 PE 34:1 1.82 1.35-2.46 7.94E-05 1.63 1.08-2.45 2.09E-02 1.88 1.34-2.65 2.55E-04 1.14 0.76-1.7 5.25E-01

195 PE 34:2 1.54 1.18-2 1.25E- 3 1.52 1.02-2.25 3.73E-02 I J2 0.98-1.79 6 96E-02 0.96 0.7-1.31 7.77E-01

196 PE 36.0 1.05 0.83-1.32 7.02E-01 1.14 0.84-1.55 4.06E-01 0.93 0.7-1.22 5.82E-01 0.81 0.58-1.13 2.05E-01

197 PE 36: 1 1.62 1.25-2.1 1 3.18E-04 1.89 1.29-2.78 1.19E-03 1.77 1.31-2.38 1.99E-04 0.98 0.69-1.39 9.12E-01

198 PE 36:2 1.54 1.19-1.99 1.05E-03 1.65 1.14-2.39 8.02E-03 1.70 1.25-2.29 5.93E-04 1.00 0.71-1.39 9.85E-01

199 PE 36:3 1.32 1.03-1.69 3.00E-02 1J7 0.96-1.95 8.16E-02 1.42 1.07-1.88 1.66E-02 1.00 0.71-1.39 9.79E-01

200 PE 36:4 1.45 1.11-1.89 6.96E-03 1.40 0.97-2.03 7.59E-02 1.56 1.15-2.13 4.64E-03 1.10 0.76-1.58 6.23E-01

201 PE 36:5 1.21 0.95-1.55 1.28E-01 1.55 1.11-2.18 1.11E-02 1.57 1.17-2.1 2.50E-03 1.03 0.73-1.44 8.82E-01

202 PE 38:1 1_33 1.04-1.7 2.44E-02 1.85 1.29-2.66 8.86E-04 1.87 1.38-2.53 6.12E-05 0.93 0.66-1.31 6.89E-01

203 PE 38:2 1.49 1.14-1.94 3.20E-03 1.71 1.14-2.55 9J24E-03 1.62 1.2-2.19 1.47E-03 1.01 0.71-1.44 9.64E-01

204 PE 38:3 1.73 1.32-2.27 7.40E-05 1.69 1.14-2.5 . 9.41E-03 1.71 1.25-2.33 7.19E-04 1.02 0.72-1.44 9.19E-01

205 PE 38:4 1.62 1.24-2.1 3.68E-04 1.51 1.04-2.18 2.90E-02 1.61 1.19-2.18 1.86E-03 1.03 0.73-1.45 8.63E-01

206 PE 38:S 1J0 1.01-1.68 4.47E-02 1.52 1 06-2.18 2.22E-02 1.56 1.16-2.1 3.44E-03 1.00 0.7-1.42 9.96E-01

207 PE 38:6 1.15 0.9-1.48 2.67E-01 1.83 1.24-2.71 2.47E-03 1.81 1.34-2.46 1.35E-04 1.06 0.76-1.49 7.31E-01

208 PE 40:6 1.43 1.1-1.86 6.86E-03 2.30 1.51-3.51 1.0OE-O4 2.11 1.53-2.92 5.86E-06 1.03 0.74-1.44 8.69E-01

209 PE 0:7 1.02 0.79-1.31 8.87E-01 1.68 1.16-2.43 6.49E-03 1.71 1.27-2.3 4.38E-04 1.05 0.74-1.5 7.71E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid U Lipid Species ratio 95% CI p-value Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% ci p-value

210 PI 32:0 1.25 0.98-1.6 7.01E-02 1.65 1.13-2.41 9.32E-03 1.50 1.12-2.01 6.72E-03 1.00 0.73-1.39 9.86E-01

211 PI 32:1 1.48 1.15-1.9 2.35E-03 1.81 1.24-2.64 2.10E-03 1.83 1.33-2.51 1.84E-04 1.04 0.75-1.44 8.13E-01

212 PI 34:0 1.19 0.94-1.52 1.56E-01 1.60 1.1-2.32 1.35E-02 1.60 1.18-2.16 2.31E-03 1.02 0.73-1,41 9.28E-01

213 PI 34:1 1.23 0.97-1.56 9.48E-02 1.68 1.16-2.43 6.29E-03 1.60 1.19-2.15 1.89E-03 1.01 0.73-1.4 9.47E-01

214 PI 36:0 0.9? 0.78-1.25 9.33E-01 1_35 0.93-1.98 1.18E-01 0.98 0.77-1.26 8.86E-01 0.71 0.47-1.09 1.22E-01

215 PI 36:1 0.98 0.78-1,24 8.93E-01 1.59 0.92-2.73 9.62E-02 0.87 0.68-1.12 2.84E-01 0.76 0.48-1.2 2.35E-01

216 PI 36:2 0.97 0.76-1.24 7.93E-01 1.52 1.06-2.18 2.17E-02 132 0.99-1.78 6.08E-O2 0.86 0.62-1.2 3.83E-01

217 PI 36:3 1.07 0.84-1.37 5.66E-01 1.36 0.95-1.96 9.48E-02 1.17 0.88-1.56 2.80E-01 · 0.84 0.6-1.17 3.01E-01

218 PI 36:4 1.28 1-1.63 5.17E-02 1.50 1.06-2.12 2.26E-02 1.55 1.15-2.08 3.52E-03 0.95 0.67-1.35 7.65E-01

219 PI 38:2 1.19 0.94-1.51 1.57E-01 1.37 0.97-1.93 7.30E-02 1.22 0.93-1.61 1.45E-01 0.90 0.65-1.25 5.30E-01

220 PI 38:3 1.27 1-1.62 4.93E-02 1.40 0.99-1.99 5.50E-02 1.15 0.88-1.52 3.02E-01 0.80 0.57-1.13 2.05E-01

221 PI 38:4 1.17 0.92-1.48 1.94E-01 1.40 1-1.97 5.07E-02 1.16 0.89-1.51 2.71E-01 0.79 0.55-1.12 1.81E-01

222 PI 38:5 1.02 0.8-1.3 8.74E-01 1.48 1.04-2.11 3.12E-02 1.21 0.91-1.61 1.90E-01 · 0.83 0.6-1.16 2.79E-01

223 PI 38:6 0.87 0.69-1.11 2.73E-01 1.83 1.3-2.59 6.24E-04 2.07 1.49-2.88 1.59E-05 1.04 0.76-1.44 7.89E-01

224 PI 40:4 1.41 1.07-1.85 1.55E-02 1.67 1.09-2.56 1.92E-02 1.23 0.92-1.65 1.62E-01 0.91 0.66-1.27 5.90E-01

225 PI 40:5 1.20 0.95-1.51 1.35E-01 2.02 1.38-2.94 2.57E-04 1.58 1.18-2.12 1.92E-03 0.90 0.66-1.23 5.09E-01

226 PI 40:6 0.91 0.72-1.15 4.32E-01 1.81 1.28-2.57 8.96E-04 2.02 1.49-2.74 5.20E-06 1.19 0.86-1.64 2.93E-01

227 PS 32:0 0.93 0.74-1.17 5.43E-01 0.76 0.55-1.04 8.72E-02 0.86 0.67-1.11 2.60E-01 1.06 0.79-1.43 6.99E-01

228 PS 32:1 1.04 0.83-1.31 7.21E-01 1.21 0.91-1.61 1.94E-01 0.93 0.71-1.21 5.84E-01 0.79 0.58-1.08 1.46E-01

229 PS 34:1 1.08 0.86-1.37 5.16E-01 1.34 0.97-1.86 7.42E-02 1J4 1.01-1.78 3.99E-02 1.15 0.79-1.66 4.78E-01

230 PS 34:2 0.97 0.77-1.21 7.76E-01 0.82 0.58-1.14 2.32E-01 1.05 0.82-1.35 6.88E-01 1.25 0.87-1.78 2.26E-01

231 PS 36:0 1.01 0.8-i:26 9.45E-01 0.98 0.71-1.36 9.19E-01 0.99 0.77-1.28 9.67E-01 1.08 0.8-1.45 6.34E-01

232 PS 36:1 1.46 1.14-1.86 2.26E-03 0.92 0.66-1.27 6.05E-01 1.15 0.88-1.5 2.93E-01 1J.8 0.96-1.99 8.42E-02

233 PS 36:2 1.19 0.94-1.5 1.41E-01 1.07 0.79-1.44 6.78E-01 0.97 0.74-1.27 8.01E-01 0.99 0.73-1.36 9.69E-01

234 PS 38:1 0.95 0.75-1.19 6.41E-01 0.79 0.58-1.07 1.34E-01 0.90 0.7-1.16 4.10E-01 1.14 0.82-1.57 4.41E-01

235 PS 38:2 0.91 0.72-1.14 4.04E-01 1.01 0.74-1.39 9.43E-01 1.06 0.82-1.37 6.52E-01 1.10 0.8-1.52 5.50E-01

236 PS 38:3 1.35 1.06-1.7 1.34E-02 1.08 0.8-1.47 6.07E-01 1.19 0.91-1.55 2.08E-01 1.10 0.8-1.51 5.56E-01

237 PS 38:4 1.09 0.85-1.38 4.99E-01 1.22 0.84-1.77 2.90E-01 1.08 0.84-1.37 5.59E-01 0.90 0.63-1.29 5.67E-01

238 PS 38:5 1.10 0.86-1.4 4.48E-01 1.04 0.75-1.44 8.02E-01 1.18 0.89-1.56 2.63E-01 1.24 0.84-1.83 2.83E-01

239 PS 40:4 1.33 1.02-1.72 3.52E-02 1.00 0.73-1.36 9.80E-01 1.17 0.87-1.56 3.01E-01 1.24 0.87-1.78 2.40E-01

240 PS 40:5 1J6 1.06-1.74 1.59E-02 0.98 0.71-1.35 8.87E-01 1.35 1.01-1.8 4.27E-02 1.42 1-2 4.68E-02

241 PS 40:6 1.00 0.79-1.25 9.75E-01 1.1 1 0.81-1.5 5.18E-01 1.66 1.24-2.21 5.71E-04 1.47 1.04-2.06 2.75E-02

242 COH 0.95 0.76-1.2 6.81E-01 1.68 1.21-2.32 1.87E-03 1.10 0.85-1.43 4.79E-01 0.58 0.4-0.84 3.6IE-03

243 CE 14:0 1.62 1.25-2.09 2.47E-04 1.81 1.23-2.67 2.67E-03 1.73 1.27-2.35 5.12E-04 1.05 0.75-1.46 7.85E-01

244 CE 15:0 1.33 1.05-1.68 1.97E-02 1.25 0.89-1.75 2.06E-01 1.20 0.92-1.57 1.80E-01 1.03 0.76-1.41 8.33E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

245 CE 16:0 137 1.07-1.75 1.14E-02 133 0.94-1.87 1.06E-01 1.42 _ 1.08-1.87 1.19E-02 1.10 0.79-1.54 5.69E-01

246 CE 16:1 1.72 1.33-2.23 3.50E-O5 1.96 1.33-2.89 7.07E-04 2.14 1.55-2.95 3.29E- 6 1.12 0.8-1.58 5.00E-01

247 CE 16:2 1.75 1.35-2.27 2.98E-05 1.89 1.28-2.79 1.28E-03 233 1.66-3.28 1.15E-06 1.13 0.81-1.59 4.76E-01

248 CE 17:0 1.15 0.92-1.45 2.25E-01 0.90 ^•0.66-1.25 5.36E-01 0.93 0.71-1.2 5.70E-01 1.05 0.78-1.41 7.52E-01

249 CE 17:1 1.51 1.18-1.93 9.87E-04 1.15 0.81-1.62 4.42E-01 138 1.05-1.82 2.21E-02 1.21 0.88-1.67 2.51E-01

250 CE 18:0 1.26 1-1.59 4.75E-02 1.22 0.89-1.68 2.21E-01 1.09 0.84-1.43 5.12E-01 0.88 0.65-1.2 4.23E-01

251 CE 18:1 1.07 0.84-1.35 5.93E-01 0.97 0.75-1.26 8.20E-01 1.45 0.94-2.25 9.51E-02 1.21 0.91-1.6 1.94E-01

252 CE 18:2 1.75 1.35-2.27 2.47E-05 1.68 1.14-2.48 8.75E-03 1.95 1.44-2.66 2.07E-05 1.12 0.8-1.55 5.18E-01

253 CE 18:3 1.60 1.24-2.07 3.43E-04 1.66 1.14-2.41 8.09E-03 1.92 1.4-2.63 4.64E-05 1.08 0.77-1.52 6.45E-01

254 CE 20:0 1.20 0.95-1.52 1.27E-01 1.12 0.81-1 55 5.06E-01 1.12 0.86-1.46 3.92E-01 0.94 0.66-1.33 7.11E-01

255 CE 20:1 0.82 0.65-1.04 1.03E-01 0.80 0.58-1.1 1 1.80E-01 0.87. 0.67-1.13 3.08E-01 1.06 0.77-1.47 ^' 7.04E-01

256 CE 20:2 ' 130 1.03-1.66 2.97E-02 136 0.97-1.9 7.55E-02 1.45 1.1-1.91 9.23E-03 1.09 0.79-1.5 6.03E-01

257 CE 20:3 1.94 1.48-2.53 1.45E-06 1.68 1.14-2.47 8.75E-03 2.01 1.46-2.76 1.74E-05 1.16 0.82-1.64 3.97E-01

258 CE 20:4 1.76 1.35-2.29 2.48E-05 1.46 1-2.13 4.82E-02 1.94 1.43-2.63 2. 1E-05 1.23 0.87-1.74 2.46E-01

259 CE 20:5 132 1.03-1.68 2.63E-02 1.48 1.03-2.11 3.32E-02 1.82 1.35-2.43 6.74E-0S 1.21 0.88-1.67 2.36E-01

260 CE 22:0 1.42 1.12-1.81 4.30E-03 2.13 1.46-3.12 938E-05 1.62 1.22-2.15 8.79E-04 0.85 0.62-1.16 3.10E-01

261 CE 22:1 0.95 0.76-1.2 6.79E-01 1.28 0.93-1.75 1.29E-01 136 1.04-1.77 2.36E-02 1.10 0.8-1.52 5.48E-01

262 CE 22:2 0.64 0.5-0.82 3.74E-04 0.94 0.68-1.29 6.91E-01 1.11 0.85-1.47 4.39E-01 1.20 0.88-1.63 2.54E-01

263 CE 22:3 1.13 0.89-1.44 3.06E-01 1.05 0.74-1.47 7.93E-01 1.11 0.84-1.46 4.54E-01 1.11 0.81-1.52 5.02E-01

264 CE 22:4 1-38 1.08-1.76 9.07E-03 134 0.94-1.9 1.08E-01 1.54 1.17-2.04 2.34E-03 1.13 0.82-1.58 4.52E-01

265 CE 22:5 1.46 1.14-1.87 3.02E- 3 1.67 1.15-2.41 6.40E-03 1.98 1.47-2.67 8.41E-06 1.20 0.86-1 68 2.82E-01

266 CE 22:6 1.28 1.01-1.63 4.08E-02 1.76 1.22-2.52 2.29E-03 1.92 1.45-2.54 5.56E-06 1.16 0.83-1.6 3.83E-01

267 CE 24:0 1.40 1.09-1.79 7.87E-03 1.74 1.18-2.57 5.14E-03 1.64 1.24-2.18 5.49E-04 1.01 o:75-l .38 9.26E-01

268 CE 24:1 1.19 0.94-1.49 1.45E-01 1.73 1.22-2.46 2.16E-03 1.62 1.24-2.12 3.61E-04 1.02 0.75-1.38 9.13E-01

269 CE 24:2 0.91 0.73-1.15 4.31E-01 1.17 0.84-1.64 3.46E-01 1 4 1.03-1.74 2.99E-02 1.11 0.83-1.49 4.75E-01

270 CE 24:3 0.87 0.69-1.09 2.29E-01 0.70 0.5-0.98 3.55E-02 0.89 0.67-1.18 4.14E-01 1.22 0.92-1.61 1.75E-01

271 CE 24:4 1.01 0.81-1.27 8.98E-01 1.16 0.83-1.62 3.74E-01 136 1.05-1.76 2.12E-02 1.12 0.82-1.53 4.92E-01

272 CE 24:5 1.24 0.97-1.58 7.99E-02 1.59 1.1-2.31 1.41E-02 1.92 1.43-2.57 1.29E-05 1.26 0.92-1.74 1.55E-01

273 CE 24:6 1.18 0.93-1.5 1.73E-01 1.64 1.09-2.46 1.78E-02 1.87 1.34-2.6 2.07E-04 1.27 0.86-1.87 2.32E-01

274 modCE 558.5/7.7 1.25 0.99-1.57 6.43E-02 1.47 1.06-2.05 2.24E-02 1.57 1.19-2.06 1.38E-03 1.03 0.75-1.41 8.74E-01

275 modCE 588.5/7.9 132 1.04-1.67 2.14E-02 1.27 0.91-1.77 1.65E-01 1.58 121-2.07 8.52E-04 1.19 0.85-1.66 3.00E-01

276 modCE 682.7/8.8 1.49 1.16-1.92 2.05E-03 1.28 0.9-1.81 1.69E-01 1.73 , 1.29-2.32 2.27E-04 1.22 0.87-1.7 2.45E-01

277 modCE 790.8/6.6 1.49 1.17-1.9 1.41E-03 1.89 1.3-2.74 8.66E-04 2.26 1.63-3.13 1.08E-06 1.07 0.78-1.46 6.91E-01

278 DG 14:0 14:0 1.70 1.32-2.18 4.14E-05 1.91 1.33-2.75 4.45E-04 1.52 1.14-2.04 4.55E-03 0.82 0.59-1.15 2.55E-01

279 DG 14:0 16:0 1.87 1.44-2.43 2.58E-06 2.17 1.48-3.18 7.78E-05 1.84 1.35-2.51 1-24E-04 0.86 0.61-1.21 3.91E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT/IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

Lipid U Lipid Species ratio 95% CI p- value Odds ratio 95% CI p-value ratio 95% CI p-value ratio 95% CI p-value

280 DG 14:0 18:1 1.87 1.44-2.43 3.18E-06 232 1.56-3.46 3.70E-05 1.62 1.2-2.19 1.50E-03 0.79 0.56-1.12 1.89E-01

281 DG 14:0 18:2 1.75 1.36-2.26 I.56E-05 2.23 1.51-3.3 6.03E-05 1.50 1.12-1.99 5.77E-03 0.77 0.55-1.07 1.16E-01

282 DG 14: 1 16:0 1.86 1.44-2.41 2.46E-06 2.20 1.5-3.23 5.81E-05 1.62 1.2^2.18 1.70E-03 0.79 0.56-1.12 1.86E-01

283 DG 16:0 16:0 1.95 1.49-2.55 9.33E-07 2.63 1.73-3.98 5.66E-06 236 1.69-3.31 5.75E-07 0.91 0.64-1.3 6.01E-01

284 DG 16:0 18:0 1.88 1.44-2.44 2.94E-06 2.80 1.83^.31 2.50E-06 2.18 1.57-3.02 2.68E-06 0.86 0^61-1.21 3.85E-01

285 DG 16:0 18:1 1.88 1.44-2.44 2.48E-06 2.85 1.85-4.38 1.80E-06 2.25 1.62-3.13 1.47E-06 0.86 0.61-1.22 4 04ET01

286 DG 16:0 18:2 1.75 1.36-2.26 1.50E-O5 2.47 1.65-3.71 1.23E-05 2.01 1.48-2.73 6_:99E-06 0.89 0.64-1.22 4.66E-01

287 DG 16:0 20:0 134 1.06-1.69 1.49E-02 1.44 1.04-2 2.91E-02 1.18 0.9-1.55 2.33E-01 0.84 0.62-1.15 2.84E-01

288 DG 16:020:3 1.84 1.41-2.39 5.47E-06 2.11 1.41-3.14 2.72E-04 1.84 1.36-2.5 8.84E-05 0.93 0.66-1.32 6.94E-01

289 DG 16:020:4 1.86 1.43-2.41 4.09E-06 2.22 ^■ 1.5-3.29 7.08E-05 2.00 1.47-2.73 UlE-05 0.91 0.65-1.28 5.75E-01

290 DG 16:022:5 1.53 1.2-1.95 6.85E-04 2.54 1.69-3.8 6.48E-06 2.04 1.5-2.77 5.54E-06 0.90 0.65-1.24 5.18E-01

291 DG 16:022:6 132 1.04-1.66 2.14E-02 2.58 1.75-3.8 1.80E-06 2.42 1.76-3.33 4.63E-08 0.98 0.72-1.34 9.13E-01

292 DG 16:1 18:1 1.85 1.43-2.39 3.40E-06 2.86 1.86-1.39 1.74E-06 2.02 1.47-2.77 1.28E-05 0.84 0.6-1.2 3.39E-01

293 DG 18:0 16:1 1.98 1.51-2.58 5.23E-07 3.01 1.93-4.7 1.17E-06 2.08 1.51-2.88 8.24E-06 0.84 0.59-1.2 3.30E-01

294 DG 18:0 18:0 1.48 1.14-1.91 3.09E-O3 1.82 1.25-2.66 1.94E-03 1.86 1.35-2.55 1.25E-04 0.96 0.7-1.33 8.23E-01

295 DG 18:0 18:1 1.93 1.47-2.52 1.78E-06 3.06 1.95^.81 1.18E-06 2.18 1.57-3.03 3.38E-06 0.85 0.61-1.21 3.69E-01

296 DG 18:0 18:2 1.72 1.33-2.22 3.62E-05 2.55 1.68-3.87 9.93E-06 1.92 1.42-2.6 2.49E-05 0.87 0.64-1.19 3.92E-01

297 DG 18:0 20:4 1.49 1.16-1.91 1.72E-03 1.92 1.33-2.78 5.36E-04 1.62 1.21-2.16 1.24E-03 0.87 0.63-1.21 4.17E-01

298 DG 18:1 18:1 1.67 1.3-2.16 7.92E-05 2.71 1.76-4.19 6.69E-06 1.93 1.42-2.64 3.22E-05 0.87 0.63-1.21 4.13E-01

299 DG 18:1 18:2 1.50 1.18-1.92 1.09E-03 2.08 1.4-3.09 2.93E-04 1.66 1.25-2.2 4.15E-04 0.90 0.67-1.22 5.1 1E-01

300 DG 18:1 18:3 1.46 1.14-1.86 2.44E-03 1.95 1.32-2.86 6.98E-04 1.47 1.12-1.94 634E-03 0.86 0.64-1.15 3.10E-01

301 DG 18:1 20:0 1.37 1.07-1.74 1.09E-02 130 0.94-1.79 1.07E-01 1.26 0.96-1.65 1.02E-01 0.94 0.68-1.31 7.16E-01

302 DG 18:1 20:3 1.45 1.14-1.85 2.43E-03 1.52 1.07-2.17 1.95E-02 139 1.06-1.84 1.95E-02 0.93 0.67-1.28 6.42E-01

^' 303 DG 18: 1 20:4 1.63 1.27-2.09 U5E-04 1.93 1.33-2.79 5.10E-04 1.76 1.31-2.36 1.69E-04 0.90 0.65-1.24 5.15E-01

304 DG 18:2 18:2 133 1.05-1.69 1.89E-02 1.67 1.16-2.41 6.34E-03 1.47 1.13-1.92 4.33E-03 0.92 0.69-1.23 5.91E-01

305 TG 14:0 16:0 18:2 1.91 1.46-2.49 2.02E-06 2.43 1.61-3.67 2.48E-05 1.57 1.16-2.13 3.62E-03 0.74 0.52-1.07 1.07E-01

306 TG 14:0 16:1 18:1 1.72 1.33-2.23 3.93E-05 2.11 1.43-3.13 1.78E-04 1.28 0.96-1.71 9.59E-02 0.69 0.48-0.99 4.40E-02

307 TG 14:0 16:1 18:2 1.62 1.26-2.08 1.52E-04 2.13 1.45-3.14 1.18E-04 1.24 0.94-1.65 1.29E-01 0.67 0.48-0.95 2.54E-02

308 TG 14:0 18:0 18:1 1.85 1.41-2.42 7.17E-06 2.22 1.48-3.34 1.22E-04 1.55 1.14-2.1 5.03E-03 0.79 0.55-1.12 1.89E-01

309 TG 14:0 18:2 18:2 1.45 1.13-1.85 3.05E-O3 1.98 1.35-2.91 4.34E-04 1.23 0.93-1.61 1.46E-01 0.73 0.53-1 5.30E-02

310 TG 14:1 16:0 18:1 1.78 1.37-2.3 1.57E-05 2.21 1.48-3.3 1.09E-04 132 0.98-1.76 6.55E-02 0.69 0.48-1 4.74E-02

311 TG 14: 1 16:1 18:0 1.89 1.45-2.46 2.72E-06 2.65 1.73-4.07 7.52E-06 1.68 1.23-2.29 9.87E-04 0.78 0.54-1.1 1 1.65E-01

312 TG 14: 1 18:0 18:2 1.51 1.17-1.93 1.24E-03 2.46 1.64-3.71 1.59E-05 1.29 0.97-1.71 7.57E-02 0.66 0.46-0.93 1.88E-02

313 TG 14: 1 18:1 18:1 1.65 1.28-2.13 1.07E-O4 2.48 1.64-3.74 1.56E-05 136 1.03-1.81 3.26E-02 0.69 0.49-0.97 3.40E-02

314 TG 15:0 18:1 16:0 1.86 1.42-2.43 5.01E-06 1.97 1.33-2.92 7.44E-04 1.44 1.07-1.94 1.71E-02 0.82 0.58-1.16 2.62E-01

Non-Obese vs. Obese (waist Pre-diabetes (IGT/IFG) vs.

circumfrence) NGT vs. Pre-diabetes (IGT IFG) NGT vs. Diabetes Diabetes

Odds Odds Odds

315 TG 15:0 18:1 18:1 1.53 1.19-1.98 1.11E-03 1.77 1.2-2.61 4-26E-03 1.15 0.87-1.51 3.17E-01 0.77 0.55-1.08 1.30E-01

316 TG 16:0 16:0 16:0 1.97 1.49-2.59 1.52E-06 2.22 1.48-3.33 1.16E-04 2.10 1.5-2.94 1.47E-05 0.96 0.67-1.37 8.11E-01

317 TG 16:0 16:0 18:0 1.88 1.43-2.46 6.48E-06 2.06 1.38-3.07 3.74E-04 1.82 1.32-2.5 2.50E-04 0.92 0.65-1.32 6.61E-01

318 TG 16:0 16:0 18:1 2.05 1.55-2.71 4.78E-07 2.62 1.7-4.02 1.12E-05 234 1.65-3.31 1.94E-06 0.90 0.63-1.31 5.93E-01

319 TG 16:0 16:0 18:2 1.95 1.49-2.55 1-22E-06 2.50 1.65-3.79 1.59E-05 2.18 1.56-3.03 4.68E-06 0.88 0.62-1.25 4.79E-01

320 TG 16:0 16:1 18:1 1.84 1.41-2.4 6.29E-06 2.79 1.8-4.31 4.04E-06 1.68 1.24-2.29 9.46E-04 0.75 0.53-1.08 1.22E-01

321 TG 16:0 18:0 18:1 2.04 1.54-2.7 6.76E-07 239 1.56-3.67 5.97E-05 2.03 1.46-2.83 2.91E-05 0.91 0.64-1.31 6.26E-01

322 TG 16:0 18:1 18:1 1.76 1.35-2.29 3J23E-05 2.75 1.76-4.3 9.16E-06 1.88 1.36-2.59 1.24E-04 0.82 0.58-1.16 2.71E-01

323 TG 16:0 18:1 18:2 1.57 1.23-2.02 3.71E- 4 2.28 1.52-3.42 6.75E-05 1.62 1.21-2.18 5E-03 0.82 0.6-1.13 2.20E-01

324 TG 16:0 18:2 18:2 1.48 1.16-1.89 1.84E-03 2.06 1.4-3.03 2.44E-04 1.54 1.16-2.04 2.85E-03 0.83 0.61-1.13 2.41E-01

325 TG 16:1 16:1 16:1 1.73 1.34-2.24 2.58E-05 2.61 1.73-3.95 5.30E-O6 1.44 1.07-1.93 1.52E-02 0.68 0.48-0.98 3.61E-02

326 TG 16:1 16:1 18:0 1.89 1.45-2.47 3.09E-06 2.50 1.64-3.81 1.94E-05 1.61 1.19-2.19 2.15E-03 0.76 0.53-1.08 1.26E-01

327 TG 16:1 16:1 18:1 1.79 1.38-2.31 1.01E-05 3.02 1.94-4.7 9.02E-07 1.75 1.29-2.39 3.61E-04 0.74 0.51-1.05 9.11E-02

328 TG 16:1 18:1 18:1 1.45 1.13-1.85 3.33E-03 2.61 1.71-4 9.I5E-06 1.48 1.11-1.97 7.84E-03 0.74 0.53-1.04 7.94E-02

329 TG 16: 1 18:1 18:2 1.43 . 1.12-1.82 4.33E-03 234 1.57-3.5 3.10E-05 1.46 1.1-1.94 8.88E-03 0.76 0.55-1.05 9.18E-02

330 TG 17:0 16:0 16:1 1.94 1.48-2.54 1.57E-06 2.01 1.35-3 6.44E-04 1.49 1.1-2.02 1.02E-02 0.83 0.58-1.18 2.99E-01

331 TG 17:0 16:0 18:0 1.97 1.5-2.59 1.19E-06 2.01 1.34-3.01 7.22E-04 1.68 1.23-2.29 1.12E-03 0.92 0.64-1.3 6.28E-01

332 TG 17:0 18:1 14:0 1.81 1.39-2.35 1.18E-05 1.83 1.24-2.69 2_25E-03 1.29 0.96-1.73 8.64E-02 0.79 0.56-1.12 1.85E-01

333 TG 17:0 18:1 16:0 2.04 1.54-2.69 6.62E-07 1.93 1.28-2.9 1.55E-03 1.55 1.14-2.11 5.63E-03 0.88 0.62-1.26 4.83E-01

334 TG 17:0 18:1 16:1 1.72 1.32-2.23 5.68E-05 2.03 1.34-3.06 7.63E-04 1.27 0.96-1.69 9.71E-02 0.77 0.55-1.09 1.40E-01

335 TG 17:0 18:1 18:1 1.64 1.26-2.13 2.46E-04 1.86 1.24-2.8 2.78E-03 135 1.01-1.8 4.19E-02 0.85 0.61-1.18 3.29E-01

336 TG 17:0 18:2 16:0 1.94 1.47-2.54 1.98E-06 231 1.5-3.54 131E-04 1.50 1.11-2.04 8.11E-03 0.79 0.55-1.13 1.99E-01

337 TG 18:0 18:0 18:0 139 1.23-2.05 3.99E-04 1.57 1.1-2.26 1.42E-02 131 0.99-1.74 6.06E-02 0.84 0.6-1.19 3.31E-01

338 TG 18:0 18:0 18:1 1.97 1.49-2.59 1.63E-06 2.25 1.48-3.42 1.44E-04 1.80 1.31-2.47 2.50E-04 0.88 0.61-1.25 4.66E-01

339 TG 18:0 18:1 18: 1 1.84 1.4-2.41 135E-05 2.65 1.7-4.14 1.78E-05 1.98 1.43-2.73 3.25E-05 0.88 0.62-1.24 4.60E-01 .

340 TG 18:0 18:2 18:2 137 1.07-1.74 1.29E-02 2.06 1.4-3.03 2.42E-04 1.56 1.17-2.07 2.18E-03 0.85 0.62-1.15 2.84E-01

■ 341 TG 18: 1 14:0 16:0 1.86 1.42-2.43 5.42E-06 2.17 1.45-3.23 1.51E-04 1.51 1.11-2.05 8.10E-03 0.77 0.54-1.11 1.60E-01

342 TG 18: 1 18: 1 18:1 1.27 1-1.63 5.30E-02 2.21 1.47-3.32 I.49E-04 1.54 1.15-2.06 3.63E43 0.81 0.59-1.11 1.88E-01

343 TG 18:1 18:1 18:2 1.14 0.9-1.45 2.67E-01 1.73 1.19-2.52 4.46E-03 1.29 0.99-1.7 6.37E-02 0.83 0.61-1.11 2.07E-01

344 TG 18:1 18:1 20:4 1.28 1.01-1.62 4,17E-02 1.82 1.27-2.62 1.Q5E-03 1.46 1.1-1.94 9.27E-03 0.79 0.58-1.08 1.44E-01

345 TG 18:1 18:1 22:6 1.03 0.82-1.3 7.75E-01 . 2.23 1.55-3.21 1.46E-05 1.84 1.36-2.47 5.88E-05 0.86 0.64-1.15 3.04E-01

346 TG 18:1 18:2 18:2 1.18 0.93-1.5 1.69E-01 1.81 1.25-2.61 1.52E-03 136 1.04-1.79 2.46E-02 0.82 0.61-1.1 1.82E-01

347 TG 18:2 18:2 18:2 1.15 0.91-1.45 2.42E-01 1.62 1.14-2.3 7.50E-03 131 1-1.7 4.65E-02 0.84 0.63-1.13 2.48E-01

348 TG 18:2 18:220:4 1.26 0.99-1.59 5.89E-02 1.55 1.11-2.16 1.02E-02 131 0.99-1.74 5.57E-02 0.81 0.59-1.1 1.69E-01

TABLE 7. Mean and standard deviation of individual lipid species in the cross sectional study¹

Whole population Diabetes NGT Pre-dia betes Obese Non obese

Lipid U Lipid Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

1 Sph 18:1 24 36 24 39 24 36 22 , 27 28 43 21 29

2 dhCer 16:0 71 21 76 22 67 18 74 25 74 23 70 20

3 dhCer 18:0 78 31 90 35 67 21 86 34 87 34 72 27

4 dhCer l8:l 15 6 18 7 14 5 15 6 16 6 15 6

5 dhCer 20:0 33 12 36 14 31 9 36 11 35 13 32 10

6 dhCer22.0 147 61 166 73 128 41 163 64 161 69 138 52

7 dhCer 24:0 256 119 287 142 225 81 283 134 278 135 241 104

8 dhCer 24:l 160 61 177 66 142 44 180 73 172 63 152 58

9 Cer 16:0 366 85 376 89 354 80 383 84 368 88 365 83

10 Cer 18:0 135 44 150 48 121 37 143 42 148 46 125 40

11 Cer 20:0 123 36 132 39 113 31 131 35 130 35 118 35

12 Cer 22:0 881 250 958 284 810 197 929 261 930 264 846 234

13 Cer 24:0 2900 803 3013 851 2740 662 3120 966 2965 838 2855 779

14 Cer 24:1 1187 320 1245 353 1114 273 1277 332 1223 318 1 161 320

15 MHC 16:0 716 193 702 193 706 182 769 214 715 187 718 197

16 MHC 18:0 183 56 180 55 179 55 197 56 183 51 184 59

17 MHC 20:0 130 38 130 37 128 37 138 41 128 38 133 38

18 MHC 22::0 1203 353 1187 358 1185 344 1276 362 1175 371 1225 337

19 MHC 24:0 1585 449 1567 479 1552 418 1707 460 1529 477 1628 424

20 MHC 24:1 1153 367 1111 383 1155 339 1222 402 1098 330 1193 386

21 DHC 16:0 3753 903 3570 907 3856 905 3813 853 3628 842 3846 934

22 DHC 18:0 89 28 84 25 90 29 92 28 90 27 88 28

23 DHC 20:0 92 30 89 28 92 31 99 31 91 29 94 31

24 DHC 22:0 251 72 242 68 252 73 262 74 244 72 256 71

25 DHC 24:0 264 69 263 71 262 66 269 72 254 65 271 71

26 DHC 24:1 918 269 857 261 954 279 932 240 875 245 949 281

27 THC 16:0 903 237 869 248 926 236 905 216 872 226 927 242

28 THC 18:0 99 34 97 36 100 33 98 31 95 33 102 34

29 THC 20:0 36 15 34 15 37 16 34 11 33 13 37 16

30 THC 22:0 152 56 145 59 161 58 143 45 144 52 159 58

31 THC 24:0 153 49 149 ^■52 157 47 146 47 141 43 161 51

32 THC 24:1 311 101 297 107 323 102 305 85 291 86 327 108

33 GM3 16:0 657 155 633 154 664 148 679 171 645 157 666 152

34 GM3 18:0 278 80 266 79 281 81 293 79 266 81 287 79

35 GM3 20:0 163 44 161 41 161 41 175 52 162 43 164 44

Whole population Diabetes NGT Pre-diabetes Obese Non obese

Lipid # Lipid Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

36 GM3 22:0 351 100 356 92 341 98 368 118 348 101 353 100

37 GM3 24:0 314 96 . 299 90 315 93 339 111 293 97 329 93

38 GM3 24:1 480 146 468 141 ^' 479 145 505 158 455 151 498 141

39 SM 14:0 11584 3997 11502 3962 11361 4049 12326 3894 11896 4081 11422 3862

40 SM 15:0 9409 2768 9133 2843 9529 2820 9592 2479 9444 2811 9395 2746

41 SM 16:0 143217 25731 141482 28953 142868 24136 147313 23434 141987 24465 144206 26596

42 SM 16:1 21798 4928 21312 4499 21884 5110 22458 5169 22263 4964 21494 4893

43 SM 18:0 30249 6661 30942 6615 29306 6401 31484 7153 31274 6680 29521 6582

44 SM 18 1 16212 4041 ^' 16308 3993 16053 4042 16458 4169 16982 4154 15665 3888

45 SM 20:0 19279 3322 19101 3048 19295 3437 19563 3514 19618 3400 19052 3259

46 SM 20:1 10679 3466 11064 3279 10169 3092 11332 4458 11226 3421 10299 3462

47 SM 22:0 90687 13577 91153 12287 89929 13420 91848 16114 91879 13185 89852 13847

48 SM 22: 1 67487 12016 68959 12181 66511 11479 67387 12993 69851 11989 65803 11802

49 SM 24:0 21540 4998 22119 4605 20855 4855 22304 5832 21569 4860 21525 5115

50 S 4: 1 74651 8664 74751 7963 74197 8808 75673 9513 76110 8346 73629 8776

51 SM 24:2 68256 12014 68530 11256 67621 11884 69440 13678 69838 12211 67115 11795

52 SM 26: 1 922 233 893 235 941 235 928 220 890 219 946 240

53 modCer 576.5/7.7 25 8 27 9 23 7 28 8 26 8 25 9

54 modCer 614.6/5.7 22 6 20 6 22 6 22 6 20 5 23 7

55 modCer 632.6 9.2 30 8 30 8 30 8 31 8 29 8 31 . 8

56 modCer 651.6/7.6 3 1 3 1 3 1 3 1 3 1 3 ^' 1

57 modCer 731.6/6.2 51 13 52 14 49 12 54 13 51 12 50 13

58 modCer 766.6/7.2 40 15 38 14 40 15 44 13 38 14 41 15

59 modCer 769.6/8.0 175 61 175 68 168 53 196 67 174 55 ^* 176 66

60 modCer 798.7/7.3 146 49 137 . 45 . 147 51 158 47 138 45 151 50

61 modCer 875.7/9.2 597 192 645 221 546 152 645 199 638 173 567 199

62 modCer 883.8/7.8 108 33 109 35 102 31 120 30 107 30 109 35

63 modCer 886.8/9.1 70 21 76 24 63 17 74 19 74 22 66 19

64 modCer 910.8/9.0 49 14 51 15 47 13 51 13 ^' 51 14 47 13

65 modCer 921.8/9.1 23 14 26 15 19 10 28 18 23 11 23 16

66 oddPC 29:0 67 20 64 23 69 20 70 17 67 21 67 20

67 oddPC 31 :0 1876 720 1878 786 1851 629 1937 822 1958 758 1817 689

68 oddPC 31:l 2164 589 2100 658 2187 550 2219 551 2225 628 2122 559

69 oddPC 33:0 5125 1407 4732 1328 5449 1416 4984 1338 4814 1398 5343 1379

70 oddPC 33:l 6273 1596 6215 1766 6266 1462 6400 1627 6472 . 1634 6131 1560

71 oddPC 33:2 616 188 625 216 607 166 625 187 621 184 613 191

Whole population Diabetes NGT Pre-diabetes Obese Non obese

Lipid U Lipid Mean St Dev Mean St Dev Mean St Dev Mean StDev Mean St Dev Mean St Dev

72 oddPC 35:0 521 138 512 147 525 133 528 137 523 143 519 135

73 oddPC 35:1 12179 2519 11946 2757 · 12288 2402 12312 2371 12155 2488 12197 2552

74 oddPC 35:2 14250 3046 13460 3066 14824 2864 14170 3189 14083 3149 14373 2979

75 oddPC 35:3 6225 1214 6118 1347 6395 1172 5974 997 6091 1175 6316 1236

76 oddPC 35:4 2504 983 2472 ^' 1013 2483 718 2621 1442 2532 954 2483 1006

77 oddPC 37:2 U763 2320 11336 2353 12047 2295 11790 2242 11723 2320 11805 2323

78 oddPC 37:3 5807 1387 5652 1619 5952 1252 5704 1242 5934 1353 5715 1409

79 oddPC 37:4 8436 1785 8079 1813 8772 1728 8193 1741 8375 1767 8472 1803

80 oddPC 37:5 1921 884 1866 897 1955 953 1930 -641 1818 694 1991 992

81 oddPC 37:6 1213 470 1199 533 1207 445 1255 416 1183 476 1234 467

82 PC 28:0 544 498 524 317 510 296 671 969 562 339 533 586

83 PC 30:2 50O3 1172 4821 1055 5059 1179 5186 1321 5135 1231 4910 1125

84 PC 32:0 12988 2352 13189 2561 12763 2021 13217 2727 13037 2599 12942 2169

85 PC 32:1 141263 41505 152620 43430 130915 34462 147989 48342 153587 45856 132434 35773

86 PC 32:2 10059 2813 10085 2715 9848 2553 10572 3538 10686 3021 9616 2582

87 PC 34:0 5445 1282 5324 1267 5427 1232 5715 1417 5322 1303 5528 1266

88 PC 34:1 162602 26647 165527 26261 159961 24543 164268 31965 165970 30416 160125 23447 .

89 PC 34:2 273335 36269 271676 36235 274283 34008 273849 42192 277815 37901 270265 34921

90 PC 34:3 24188 7083 24414 6816 23494 6289 25616 9144 25167 7053 23491 7054

91 PC 36:1 27089 3025 26821 3265 27160 2710 27393 • 3357 27357 3220 26915 2877

92 PC 36.2 227686 35606 225333 32332 227535 35176 232386 42030 231394 35921 225173 35315

93 PC 36:3 140970 23180 142245 24550 140034 21239 141126 25687 145623 22335 137545 23157

94 PC 36:4 120800 29760 124492 29441 118844 26717 119244 37034 125263 32918 117400 26730

95 PC 36:5 3S782 14655 37785 15164 33534 12197 38094 18554 35733 14047 35743 15097

96 PC 38:2 32655 4688 32568 4479 32379 4620 33548 5185 33407 4724 32132 4613

97 PC 38.3 45896 5653 45736 5764 45916 5084 46136 6850 46946 6026 45143 5273

98 PC 38:4 112156 25194 114739 24265 109606 23879 114210 29638 117491 25932 108317 24028

99 PC 38:5 67661 14833 68015 15585 66753 12989 69426 17802 68165 14747 67234 14915

100 PC 38:6 61740 16169 62909 18917 60435 14892 63067 13770 60522 16326 62499 16031

101 PC 39:7 3527 1044 3461 1Γ30 3556 977 3570 1062 3293 941 3688 1085

102 PC 40:5 20464 5662 21441 6033 19300 5065 21772 5923 21279 5658 19865 5604

103 PC 40:6 27425 8822 28756 9953 25960 8258 28885 7469 27576 8331 27268 9157

104 PC 40:7 5842 1856 5665 1924 5913 1909 5979 1564 5562 1775 6031 1892

105 PC 44:12 1803 ; 468 1740 493 1866 464 1752 416 1737 422 1848 494

106 APC 30:0 18 6 18 7 19 6 18 6 18 ^■6 19 7

107 APC 30:1 191 62 181 65 196 59 196 62 179 53 200 67

Whole population Diabetes NGT Pre-diabetes Obese Non obese

Lipid Lipid Mean St Dev Mean St Dev Mean St Dev Mean St ev Mean St Dev Mean St Dev

108 APC 32:0 2441 455 2453 507 2459 426 2372 425 2383 431 2481 468

109 APC 32:1 564 161 565 199 566 136 555 146 530 127 588 178

110 APC 34:0 520 139 513 143 " 532 140 502 129 514 129 524 146

111 APC 34:1 4985 931 4889 988 5130 905 4772 841 4774 824 5137 974

112 APC 34:2 5316 1431 4981 1395 5595 1418 5188 1401 4960 1360 5566 1433

113 APC 34:4 1301 373 1225 397 1366 348 1269 367 1219 336 1360 388

114 APC 36:0 68 18 68 19 69 18 67 18 67 18 69 19

115 APC 36:1 745 222 723 226 773 226 711 196 709 183 770 244

116 APC 36:2 4665 1160 4383 1089 4922 1171 4498 1125 4380 990 4868 1230

117 APC 36:3 3830 1301 3469 1131 4096 1340 3784 1339 3477 1015 4081 1420

118 APC 36:4 14707 2264 14641 2393 14940 2293 14209 1853 14466 2235 14870 2279

119 APC 36:5 10039 1998 9949 . 2049 10127 1984 9973 1961 9836 1929 10185 2041

120 APC 38:2 424 145 400 150 444 144 412 132 389 132 448 149

121 APC 38:3 2973 1015 3030 1295 2944 739 2944 1076 3002 1227 2952 841

122 APC 38:4 10849 2165 10812 2289 11014 2218 10481 1727 10786 2043 10888 2254

123 APC 38:5 V 16971 2765 16623 2615 17331 2841 16653 2746 16556 2562 17258 2875

124 APC 38:6 7470 1560 7400 1668 7504 1429 7509 1707 7137 1346 7700 1661

125 APC 40:7 676 172 701 177 655 165 685 179 681 166 673 178

126 LPAF 16:0 629 166 602 168 642 157 645 183 587 140 657 176

127 LPAF 18:0 173 48 161 46 181 49 172 44 160 42 182 51

128 LPAF 18:1 452 132 427 133 464 126 463 143 418 116 475 138

129 LPAF 22:0 11 3 10 2 11 3 11 3 10 3 11 3

130 LPAF 22:1 38 40 38 42 39 38 33 41 43 48 34 32

131 LPAF 24:1 23 7 21 6 24 7 24 6 22 6 24 7

132 LPAF 24.2 5 2 4 2 5 5 2 4 2 5 2

133 LPC 14:0 2057 747 2061 733 1969 647 2281 956 2112 750 2015 744

134 LPC 15:0 1226 369 1149 369 1258 335 1283 - 432 1197 364 1245 372

135 LPC 16:0 95658 21211 95712 23883 93714 17251 100722 24793 93361 20211 97129 21762

136 LPC 16:1 4301 1287 4365 1424 4135 1072 4624 1484 4314 1350 4285 1244

137 LPC 18:0 28736 7677 28033 8418 28395 6666 30930 8466 27547 ^" 7121 29555 7973

138 LPC 18:1 21736 5519 20421 5445 22263 5143 22737 6218 19928 4697 22980 5716

139 LPC 18:2 29264 8520 26739 8011 30999 8659 29275 8002 26030 6541 31529 9033

140 LPC 20:0 118 36 no 36 122 37 123 34 104 28 128 38

141 LPC 20:1 287 87 261 80 298 90 302 80 255 71 309 90

142 LPC 20:1 87 26 80 25 91 27 89 25 78 21 94 28

143 LPC 20:2 316 88 296 91 322 82 335 92 291 77 333 91

Whole population Diabetes NGT Pre-diabetes Obese Non obese

Lipid Lipid Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

144 LPC 20:3 . 2905 811 2927 899 2853 658 3003 992 2885 759 2914 845

145 LPC 20:4 8520 2497 8295 2359 8602 2332 8716 3107 8129 2338 8771 2559

146 LPC 20:5 1536 832 1542 766 1470 793 1697 1020 1383 591 1640 ^'953

147 LPC 22:0 26 9 24 8 27 10 26 8 23 7 28 10

148 LPC 22:6 2799 988 2795 1059 2723 905 3011 1047 2533 826 2981 1050

149 LPC 24:0 59 17 57 15 ^• 61 18 61 16 55 14 62 18

150 modPC.506.3/3.4 6 2 6 2 6 2 6 2 6 2 6 2

151 modPC.512.3/1.7 83 20 84 21 83 19 81 23 82 20 84 20

152 modPC.536.3/3!5 73 22 65 24 77 20 75 21 66 21 77 22

153 modPC.538.3/4.1 82 25 73 25 88 24 83 22 74 22 88 25

154 modPC.594.4/3.1 198 104 216 114 195 104 175 80 212 114 188 95

155 modPC.608.4/4.0 25 16 29 20 23 12 25 14 27 15 24 16

156 modPC.610.4/1.7 16 16 18 19 16 17 13 11 17 17 15 16

157 modPC.622.4/4.0 16 24 15 9 15 9 21 . 53 16 10 16 30

158 modPC.633 4/4.6 18 6 16 5 18 6 18 6 18 6 18 6

159 modPC.636.4/3.6 214 108 242 127 200 92 200 101 239 126 196 90

160 modPC.645.4/4.4 40 20 38 18 41 20 43 25 43 21 38 19

161 modPC.650.4/3.8 990 430 1060 449 963 412 936 431 1046 453 951 410

162 modPC.650.4/3.9 99 47 108 SO 96 44 90 45 105 48 95 46

163 modPC.650.4/4.4 104 8 103 8 105 8 103 7 105 8 103 7

164 modPC.664.4/4.3 98 49 107 53 92 42 96 ^' 54 109 56 90 41

165 modPC.666.4 2.9 168 109 182 118 161 103 159 109 179 121 159 99

166 modPC.678.4/5.0 79 23 76 23 79 22 86 23 81 24 78 22

167 modPC.690.4/4.3 47 22 50 24 46 22 44 22 50 24 45 21

168 modPC.691.4/5.1 18 6 18 6 18 5 1 6 19 6 18 5

169 modPC.703.5/4.5 62 28 64 29 62 27 61 30 65 31 60 26

170 modPC.704.5/3.8 16 11 16 11 16 11 13 9 16 11 16 11

171 modPC.706.5/4.2 28 15 30 15 28 15 25 14 30 15 27 15

172 modPC.720.5/4.5 14 6 15 6 14 5 13 6 14 6 14 6

173 modPC.736.5/5.7 21 7 21 8 22 7 21 7 20 7 22 7

174 modPC.743.5/6.0 3055 777 2798 720 3275 779 2940 708 2855 760 3198 759

175 modPC.752.5/5.7 321 212 329 171 296 194 374 300 334 195 312 223

176 modPC.772.5/5.4 125 80 132 85 122 74 117 87 136 88 117 74

177 modPC.773.6/6.5 6579 1531 6129 1499 6875 1432 6618 1662 6486 1581 . 6648 1497

178 modPC.788.6/5.2 169 98 179 108 165 90 159 100 182 105 160 92

179 modPC.801.6/6.8 13998 2498 14118 2590 13869 2441 14120 2498 14264 2484 13821 2500

Whole population Diabetes NGT Pre-diabetes Obese Non obese

180 modPC 816.6/5.5 80 50 82 55 80 47 76 50 87 55 75 46

181 modPC 827.7/6.8 6310 1 24 6015 1719 6451 1533 6472 1631 6301 1633 6315 1626

182 modPC.828.6/5.8 298 194 302 210 293 189 306 181 301 206 296 ' 187

183 modPC.843.6/7.2 360 76 355 76 360 75 367 76 343 72 372 76

184 modPC.866.6/7.2 80 25 80 26 81 26 79 25 75 21 84 28

185 modPC.877.6/6.0 47 27 46 24 47 30 47 23 42 18 50 31

186 modPC 879.1/6.1 16 8 15 7 16 9 16 7 14 6 17 9

187 PG 16:0 18:1 83 38 90 40 76 35 87 38 87 39 79 36

188 PG 16:1 18:1 10 5 11 6 10 4 11 5 11 5 10 5

189 PG 18:0 18:1 67 41 80 49 55 28 75 45 76 42 60 38

190 PG 18:1 18:1 67 33 78 37 56 25 73 36 76 34 60 31

191 PE 32:0 32 12 36 13 30 10 33 11 34 12 31 11

192 PE 32:1 102 78 128 101 82 56 106 68 125 95 85 58

193 PE 32:2 16 ? 17 11 14 8 17 9 18 11 14 7

194 PE 34:1 1397 717 1681 861 1189 568 1428 601 1608 773 1239 620

195 PE 34:2 1657 857 1889 1048 1465 ^' 675 1739 795 1876 957 1500 744

196 PE 36:0 16 4 15 4 16 4 16 4 16 4 16 4

197 PE 36:1 911 464 1097 576 767 324 954 427 1033 490 818 408

198 PE 36:2 5314 2490 6208 2965 4573 1823 5649 2503 6022 2700 4801 2195

199 , PE 36:3 1577 718 1759 864 1432 588 1628 663 1726 791 1467 641

200 PE 36:4 2142 960 2427 1151 1937 813 2164 805 2370 1052 1974 851

201 PE 36:5 , . 199 100 223 103 176 88 216 112 215 109 187 92

202 PE 38:1 47 15 52 16 42 12 51 16 50 14 45 15

203 PE 38:2 179 58 199 70 164 47 184 49 194 62 168 52

204 PE 38:3 1224 581 1433 724 1063 438 1267 491 1406 633 1091 499

205 PE 38:4 5146 2271 5899 2660 4559 1841 5329 2123 5820 2399 4651 2025

206 PE 38:5 2314 1000 2565 1118 2097 882 2430 954 2517 1092 2161 895

207 PE 38:6 3246 1681 3752 1936 2838 1471 3404 1431 3503 1811 3052 1554

208 PE 40:6 1854 1032 2257 1214 1516 802 2015 917 2096 1074 1675 960

209 PE 40:7 346 191 387 208 309 173 366 188 366 215 329 169

210 PI 32:0 101 75 120 95 87 57 103 71 109 81 ^" 95 71

211 PI 32:1 268 222 346 297 206 124 288 218 314 273 234 171

212 PI 34:0 123 59 141 76 109 41 126 53 130 66 117 52

213 PI 34:1 1295 594 1480 762 1152 422 1336 541 1378 678 1232 517

214 PI 36:0 91 35 92 36 89 35 96 34 90 34 91 35

215 PI 36:1 1119 399 1157 444 1080 373 1154 371 1111 387 1118 395

Whole population Diabetes NGT Pre-diabetes Obese Non obese

216 PI 36:2 2943 896 3059 944 2803 821 3104 950 2968 888 2922 904

217 PI 36:3 814 269 842 296 781 244 852 275 836 276 798 264

218 PI 36:4 1014 396 1115 431 927 337 1063 430 1099 445 954 348

219 PI 38:2 249 78 260 83 237 73 262 81 258 74 243 81

220 PI 38:3 2261 650 2332 690 2163 605 2392 662 2367 648 2186 644

221 PI 38:4 6705 1745 6859 1775 6475 1733 7035 1663 6955 1740 6522 1731

222 PI 38:5 655 212 672 219 625 191 700 242 669 221 644 206

223 PI 38:6 186 75 203 78 168 65 201 82 . 183 70 187 78

224 PI 40:4 113 39 124 47 103 31 118 35 120 40 107 38

225 PI 40:5 412 142 453 163 370 108 447 151 428 147 399. 137

226 PI 40:6 431 174 481 178 388 159 455 180 424 146 ^' 436 191

227 PS 32:0 12 3 11 3 12 3 11 3 12 3 12 3

228 PS 32:1 2 1 2 1 3 1 2 1 2 1 2 1

229 PS 34:1 8 4 8 3 7 4 8 4 8 4 8 3

230 PS 34:2 3 1 3 1 3 1 2 1 3 1 3 1

23 PS 36:0 13 5 13 5 13 4 12 5 13 5 13 5

232 PS 36:1 105 46 111 35 103 56 99 33 114 58 99 35

233 PS 36:2 35 14 35 11 34 15 36 15 36 15 34 13

234 PS 38:1 9 4 9 4 9 4 8 3 9 4 9 3

235 PS 38:2 12 4 13 4 12 4 12 4 12 5 12 4

236 PS 38:3 38 15 40 15 36 13 38 17 40 15 36 14

237 PS 38:4 149 66 158 73 142 57 152 78 156 68 145 66

238 PS 38:5 18 8 19 9 18 7 18 9 18 8 18 9

239 PS 40:4 23 11 25 12 22 9 22 12 24 11 22 10

240 PS 40:5 39 18 43 20 37 15 38 20 42 19 ■37 16 ^'

241 PS 40:6 59 30 66 31 54 24 60 41 60 33 58 29

242 COH 816615 314137 802256 286332 779283 326207 942027 302822 810097 285343 819676 333441

243 CE 14:0 21386 121S8 26407 15114 17237 7896 23225 1 1691 24092 11849 19331 11812

244 CE 15:0 14039 6110 15422 7275 12898 5045 14539 5863 14878 5952 13431 6170

245 CE 16:0 356452 95826 386311 108932 333721 82256 362244 89061 374929 97143 343634 93210

246 CE 16:1 234738 157518 310260 199331 177578 93525 248507 147038 278346 172872 202430 135800

247 CE 16:2 11329 6247 14339 7763 8951 3654 12144 6103 13214 6850 9942 5337

248 CE 17:0 7159 2782 7362 3270 7039 2490 7104 2557 7364 2991 7014 2630

249 CE 17:1 23459 9652 26715 11588 21349 7868 23110 8507 25520 9949 21989 9203

250 CE 18:0 13195 4260 13568 4812 12692 3576 13852 4746 13837 4959 12762 3649

251 CE 18:1 643075 194330 711370 213034 595651 161622 644195 203549 672367 203911 622058 185375

Whole populadon Diabetes NGT Prediabetes Obese Non obese

' 252 CE 18:2- 2072777 706604 2389118 843409 1833142 504916 2130994 660251 2259644 650024 1935076 709761

253 CE 18:3 299878 150289 362899 178725 248191 101288 321960 155132 336536 149734 272793 144226

254 CE 20:0 322 377 351 321 308 453 308 215 324 239 322 450

255 CE 20:1 424 195 410 193 440 198 408 187 402 198 441 192

256 CE 20:2 1368 681 1506 919 1265 545 1391 394 1456 828 1307 551

257 CE 20:3 99103 40931 118671 49133 84885 27904 101096 38896 111979 40202 89540 38200

258 CE 20:4 966285 388799 1147760 463570 841481 292055 966040 334842 1073626 381158 885944 369235

259 CE 20:5 336295 204463 418602 238961 273543 147726 352514 212357 364578 205850 314369 199523

260 CE22.0 129 83 155 106 104 51 148 84 141 83 120 81

261 CE 22:1 222 117 240 138 204 96 235 121 222 118 222 117

262 CE 22:2 31 13 31 13 31 13 31 14 29 12 33 13

263 CE 22:3 181 86 201 101 169 74 179 81 190 86 175 86

264 CE 22:4 1077 396 1223 466 976 325 1079 347 1148 390 1027 393 <

265 CE 22:5 19360 8158 22926 10107 16686 5747 19945 7059 20858 . 7613 18197 8215 -

266 CE 22:6 147833 78355 178941 94362 124343 62726 153357 61986 157542 75385 140292 79228

267 CE 24:0 121 90 155 125 95 - 49 128 74 134 87 112 89

268 CE 24:1 297 131 340 162 · 260 98 319 118 312 130 287 132

269 CE 24:2 34 13 37 15 32 11 35 13 34 14 34 12

270. CE 24:3 14 7 15 7 14 6 13 7 13 6 15 7

271 CE 24:4 51 19 55 22 48 16 52 16 52 18 51 ^l*

272 CE 24:5 137 69 165 97 118 41 138 50 145 65 132 72

273 CE 24:6 230 149 287 198 191 100 232 117 256 156 214 140

274 modCE 558.5/7.7 127623 88638 152748 103467 105209 67294 141230 95254 135965 81946 120287 90321

275 modCE 588.5/7.9 15763 8983 18551 10802 14039 7633 15243 7381 16871 9215 14931 8737

276 modCE 682.7/8.8 44350 29196 55143 37890 36855 19551 44531 26433 49653 29324 40332 28273

277 modCE 790.8/6.6 12687 6792 15417 7948 10182 3852 14351 8157 13969 6242 11733 6985

278 DG 14:0 14:0 48 52 60 57 . 34 31 62 73 57 51 41 51

279 DG 14:0 16:0 S16 506 692 598 348 278 643 637 627 514 431 473

280 DG 14:0 18:1 1346 1008 1690 1226 1010 626 1606 1111 1565 898 · 1172 1013

281 DG 14:0 18:2 635 494 778 611 484 315 772 532 734 436 559 513

282 DG 14:1 16:0 122 112 157 135 86 67 152 131 149 117 101 102

283 DG 16:0 16:0 1414 1352 1973 1646 917 602 1714 1691 1723 1389 1172 1226

284 DG 16:0 18:0 936 746 1240 929 666 353 1100 877 1094 694 805 698

285 DG 16:0 18:1 8528 5732 10995 7062 6320 3031 9883 6306 9828 5090 7481 5637

286 DG 16:0 18:2 4117 2720 5212 3438 3142 1553 4706 2781 4674 2307 3683 2846

287 DG 16:020.0 89 110 103 153 74 62 105 111 98 136 82 84

Whole population Diabetes NGT Pre-dia betes Obese Non obese

288 DG 16:020:3 242 172 310 225 185 94 264 171 281 162 210 161

289 DG 16:020:4 572 454 740 558 425 260 654 523 673 421 492 444

290 DO 16:022:5 206 140 257 163 155 78 245 174 225 119 189 146

291 DG 16:022:6 379 410 486 341 251 186 520 735 390 257 366 487

292 DG 16:1 18:1 339 2178 4301 2715 2567 1310 3949 2117 3996 2039 2932 2069

293 DG 18:0 16:1 247 195 331 247 173 101 289 203 298 187 205 176

294 DG 18:0 18:0 154 110 192 144 122 59 170 117 171 104 138 94

295 DG 18:0 18:1 1635 1154 2136 1504 1200 541 1874 1189 1881 972 1425 1110

296 DG 18:0 18:2 780 509 980 647 597 268 897 544 879 435 698 518

297 DG 18:020:4 248 112 280 136 218 78 268 119 267 97 231 108

298 DG 18:1 18:1 8443 4705 10382 6012 . 6689 2672 9559 4555 9382 3914 7672 4813

299 DG 18:1 18:2 7470 4329 9003 5564 6066 2583 8396 4322 8199 3648 6895 4598

300 DG 18:1 18:3 1554 977 1832 1247 1270 .603 1799 1021 1697 844 1440 1033

301 DG 18:1 20:0 157 169 177 204 136 130 178 185 172 188 146 153

302 DG 18:1 20:3 601 342 696 446 518 220 650 341 657 303 552 333

303 DG 18:1 20:4 1559 973 1881 1207 1264 584 1755 1094 1769 907 1394 954

304 DG 18:2 18:2 1417 1155 1734 1635 1135 619 1585 1020 1527 880 1332 1308

305 TG 14:0 16:0 18:2 13310 11221 16964 13460 9647 . 7070 16361 12704 16007 11000 11276 10800

306 TG 14:0 16:1 18:1 13910 10449 16398 12375 11050 7562 16955 11.321 15992 9652 12296 10517

307 TG 14:0 16:1 18:2 3091 2492 3568 2889 2463 1812 3887 2864 3518 2267 2766 2578

308 TG 14:0 18:0 18:1 1439 1277 1856 1562 1042 759 1732 1478 1668 1149 1243 1234

309 TG 14:0 18:2 18:2 1691 1303 1942 1559 1370 956. 2085 1398 1841 1122 1573 1399

310 TG 14:1 16:0 18:1 3421 2888 4178 3517 2612 1983 4186 3095 4027 2767 2956 2837

311 TG 14:1 16:1 18:0 12022 10403 15889 13038 8379 6083 1462 10887 14883 10996 9868 9225

312 TG 14:1 18:0 18:2 727 436 826 549 605 312 867 406 794 366 671 457

313 TG 14:1 18:1 18:1 6887 3942 7907 4729 5669 2817 8258 4004 7660 3402 6277 4079

314 TG 15:0 18:1 16:0 2782 2151 3453 2516 2127 1436 3293 2501 3239 2075 2420 2061

315 TG 15:0 18:1 18:1 2145 1158 2395 1417 1869 884 2420 1125 2308 996 2005 1194

316 TG 16:0 16:0 16:0 23976 25503 34470 31266 15021 12944 28579 30355 29783 26835 19584 23282

317 TG 16:0 16:0 18:0 6539 6955 9245 8987 4295 3571 7555 7587 7853 6795 5474 6598

318 TG 16:0 16:0 18:1 59523 44691 79315 53445 42127 23733 69552 51763 70329 42227 51112 43136

319 TG 16:0 16:0 18:2 17311 12933 22543 15219 12487 7160 20559 15519 20203 11887 15087 12948

320 TG 16:0 16:1 18:1 62948 37995 76996 47024 49167 24105 73873 37281 72240 34769 55683 37217

321 TG 16:0 18:0 18:1 19394 15956 26344 20513 13666 8224 21904 16542 22991 14801 16425 14864

322 TG 16:0 18:1 18:1 138683 65589 164994 78743 113646 41447 157086 67481 153486 56887 126984 66468

323 TG 16:0 18:1 18:2 66929 33948 78496 40975 55261 21604 76773 36672 73285 29515 61964 35363

Lipid # Lipid Mean St Dev Mean St Pev Mean St Dev Mean St Dev Mean St Dev Mean St Pev

324 TG 16:0 18:2 18:2 19639 12264 23393 14902 15778 7980 23034 13332 21554 10691 18165 12981

325 TG 16:1 16:1 16:1 2938 2257 3537 2649 2239 1598 3702 ^' 2437 3483 2258 2527 2137

326 TG 16:1 16:1 18:0 1125 907 1432 1116 826 542 1360 1009 1318 835 969 877

327 TG 16:1 16:1 18:1 8324 5257 10179 6269 6410 3440 10015 5498 9703 4983 7274 5091

328 TG 16:1 18:1 18:1 14817 7383 16957 9119 12431 5206 17239 6866 16090 6551 13771 7478

329 TG 16:1 18:1 18:2 17017 9272 19483 10994 14079 6245 - 20313 10238 18496 8371 15845 9540

330 TG 17:0 16:0 16:1 6488 5231 8262 6232 4827 3304 7657 6047 7724 5191 5518 4856

331 TG 17:0 16:0 18:0 452 338 587 423 339 202 505 352 536 329 385 308

332 TG 17:0 18:1 14:0 4931 3736 5922 4387 3915 2621 5820 4274 5653 3527 4350 3636

333 TG 17:0 18:1 16:0 4074 3003 5187 3805 3124 1845 4564 3053 4816 2847 3469 2723

334 TG 17:0 18:1 16:1 10081 5619 11670 7070 8498 3987 11382 5250 11168 4810 9185 5680

335 TG 17:0 18:1 18:1 3643 1947 4241 2547 3102 1311 3985 1712 3984 1649 3347 1930

336 TG 17:0 18:2 16:0 6104 3818 7465 4782 4842 2478 6967 3728 7054 3452 5353 3713

337 TG 18:0 18:0 18:0 313 405 417 542 229 237 347 422 382 493 255 285

338 TG 18:0 18:0 18:1 1427 1541 2026 2123 953 724 1591 1477 1727 1503 1160 1283

339 TG 18:0 18:1 18:1 12568 8196 16012 10798 9646 4187 14031 8015 14237 7063 11109 7629

340 TG 18:0 18:2 18:2 2170 1238 2561 1568 1779 758 2492 1285 2334 1057 2026 1276

341 TG 18:1 14:0 16:0 21617 17408 27054 20238 16137 11621 26234 20237 ,25231 16327 18802 17236

342 TG 18:1 18:1 18:1 66533 32372 76365 39892 56359 20503 75586 34957 69994 28676 63368 32890

343 TG 18:1 18:1 18:2 9790 5000 10902 6073 8522 3410 11127 5595 10108 4491 9488 5207

344 TG 18:1 18:1 20:4 2902 1281 3211 1526 2548 861 3276 1474 3050 1112 2766 1301

345 TG 18:1 18:1 22:6 1790 1020 2020 1056 1475 609 2205 1479 1766 736 1789 1155

346 TG 18:1 18:2 18:2 9748 5840 11028 7036 8267 4145 11341 6381 10067 4907 9450 6335

347 TG 18:2 18:2 18:2 2687 2335 3137 3157 2222 1550 3098 2090 2727 1710 2642 2686

348 TG 18:2 18:220:4 444 278 484 304 387 20 . 521 357 471 264 422 283 data expressed as pmol/ml plasma.

TABLE 8. Logistic regression of total lipid classes in the longitudinal study

Diabetes Incidence Pre-diabetes progressors¹

Lipid Class Odds Ratio 95% CI P value Odds Ratio 95% CI P value total Sph 1.03 0.86-1.22 7.67E-01 1.20 0.91-1.58 2.07E-01 total dhCer 1.41 1.17-1.7 2.89E-04 1,14 0.87-1.49 3.40E-01 total Cer 1.30 1.08-1.55 4.93E-03 1.19 0.91-1.57 2.01E-01 total MHC 0.93 0.79-1.11 4.23E-01 0.94 0.72-1.21 6.20E-01 total DHC 0.79 0.67-0.94 7.05E-03 0.86 0.66-1.13 2.83E-01 total THC 0.82 0.69-0.98 3.03E-02 0.82 0.61-1.11 1.96E-01 total GM3 0.92 0.77-1.09 3.17E-01 0.87 0.66-1.15 3.39E-01 total SM 1.00 0.84-1.2 9.74E-01 1.19 0.9-1.56 2.23E-01 total modCer 1.17 0.98-1.39 8.29E-02 1.24 0.94-1.63 1.29E-01 total LPAF 1.09 0.91-1.3 3.49E-01 1.03 0.78-1.34 8.50E-01 total LPC 1.25 1.04-1.49 1.53E-02 1.08 0.83-1.41 5.76E-01 total PC 1.10 0.92-1.31 3.08E-01 1.47 1.08-2 1.45E-02 total APC 0.89 0.75-1.07 2.16E-01 1.20 0.9-1.58 2.12E-01 total oddPC 0.85 0.71-1.01 6.63E-02 1.13 0.87-1.48 3.65E-01 total modPC 0.91 0.76-1.08 2.84E-01 1.11 0.85-1.44 4.51E-01 total PG 1.42 1.17-1.71 2.83E-04 1.13 0.84-1.51 4.08E-01 total LPE 1.23 1.03-1.46 2.44E-02 0.99 0.75-1.3 9.43E-01 total PE 1.43 1.19-1.72 1.55E-04 1.20 0.89-1.61 2.28E-01 total APE 1.09 0.91-1.29 3.44E-01 1.11 0.86-1.44 4.16E-01 total PS 1.16 0.97-1.38 9.64E-02 1.23 0.92-1.66 1.62E-01 total PI 1.27 ^' 1.06-1.52 8.56E-03 1.04 0.79-1.36 7.87E-01 total CE 1.51 1.25-1.83 1.78E-05 1.18 0.88-1.59 2.55E-01 total modCE 1.42 1.18-1.7 2.10E-04 1.15 0.86-1.54 3.32E-01 total DG 1.53 1.26-1.86 1.32E-05 1.36 1.01-1.83 4.21E-02 total TG 1.45 1.19-1.76 1.71E-04 1.31 0.97-1.78 8.20E-02

COH 1.12 0.94-1.33 2.08E-01 1.05 0.8-1.37 7.46E-01 basaed on the analysis of only the pre-diabetes at baseline group.

TABLE 9. Mean and Standard Deviation of total lipid classes in the longitudinal study

No diabetes at baseline Pre-diabetes at baseline Pre-diabetes at baseline Pre-diabetes at baseline and Whole population and follow-up Diabetes incidence and follow-up and diabetes at follow-up diabetes free at follow-up

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev total Sph 131 117 132 120 128 112 131 121 135 121 125 121 total dhCer 674 244 636 230 749 253 749 277 767 255 719 312 total Cer 5426 1358 5278 1300 5712 1423 5653 1442 5753 1498 5480 1329 total MHC 5469 1620 5526 1595 5358 1666 5320 1632 5255 1656 5434 1591 total DHC 5565 1502 5720 1500 5266 1465 5283 1427 5192 1474 5440 1335 total THC 1353 393 1393 394 1277 381 1282 374 1246 373 1345 370 total GM3 1769 455 1790 452 1727 459 1747 464 1706 458 1819 469 total SM 279696 41193 279655 41447 279776 40792 277600 40901 279300 39543 274647 43213 total modCer 1244 261 1220 250 1291 274 1281 260 1306 274 1238 229 total LPAF 1094 302 1095 299 1091 309 1080 302 1078 315 1084 280 total LPC 167466 39640 165727 38259 170820 42060 168848 41407 169181 42711 168269 39254 total PC 1940966 280972 1930233 289624 1961660 262882 1946154 254528 1976512 252471 1893426 250731 total APC 63430 9701 64270 9680 61809 9558 61482 9287 61681 9618 61136 8720 total oddPC 58054 10131 58841 10051 56537 10136 56296 9955 56559 10165 55840 9614 total modPC 37523 5644 37730 5634 37122 5654 36791 5733 36965 5586 36488 5996 total PG 246 120 226 104 284 138 278 133 291 143 257 109 total LPE 990 293 976 288 1018 302 999 293 996 294 1005 293 total PE 27343 11666 25564 10354 30772 13213 30438 12809 _. 31708 13630 28232 10964 total APE 3476 1042 3459 1016 3509 1091 3542 1073 3581 1097 3474 1031 total PS 550 244 534 240 581 249 561 221 579 240 529 179. total PI 27590 7239 26739 6508 29230 8244 28952 7800 29344 8487 28271 6420

COH 1099709 297973 1084938 283871 1128190 322189 1115245 307526 1121968 324701 1103568 276435 total CE 9090559 3271410 8490688 2765451 10247262 3822535 10024432 3424463 10390485 3745820 9388656 2681020 total modCE 83481 48821 76140 41320 97638 58268 90930 51525 96120 56774 81917 39523 total DG 51608 34014 45192 26418 63979 42611 61940 40436 68069 45938 51294 25355 total TG 791153 526179 701010 407504 964972 667941 925622 627473 1014014 732122 772100 334880

¹ data expressed as pmol/ml plasma.

TABLE 10. Logistic regression of lipids against incident diabetes and pre-diabetes progressors

incident diabetes pre-diabetes progressors

# Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value

1 Sph 18:1 1.01 0.84-1.2 8.79E-01 1.20 0.90-1.59 2.06E-01

2 dhCer 16:0 1.10 0.92-1.31 2.86E-01 0.96 0.75-1.23 7.71E-01

3 dhCer l 8:0 1.34 1.11-1.62 2.42E-03 1.11 0.84-1.46 4.65E-01

4 dhCer 18:1 1.09 0.90-1.3 3.87E-01 0.94 0.68-1.27 6.86E-01

5 dhCer 20:0 1.25 1.04-1.49 1.26E-02 1.03 0.80-1.31 8.33E-01

6 dhCer 22:0 1.49 1.23-1.8 3.95E-05 1.20 0.91-1.56 1.85E-01

7 dhCer 24:0 1.44 1.18-1.73 1.69E-04 1.21 0.92-1.57 1.72E-01

8 dhCer 24:l 1.30 1.07-1.56 5.74E-03 1.03 0.77-1.35 8.50E-01

9 Cer 16:0 1.19 0.98-1.42 6.88E-02 1.08 0.82-1.4 5.75E-01

10 Cer 18:0 1.40 1.15-1.69 5.09E-04 1.23 0.90-1.66 1.91E-01

11 Cer 20:0 1.40 1.16-1.69 4.25E-04 1.15 0.86-1.54 3.38E-01

12 Cer 22:0 1.47 1.21-1.77 5.57E-05 1.26 0.96-1.65 9.21E-02

13 Cer 24:0 1.26 1.05-1.51 1.27E-02 1.20 0.92-1.56 1.65E-01

14 Cer 24:1 1.16 0.96-1.4 1.13E-01 0.97 0.72-1.28 8.25E-01

15 MHC 16:0 0.91 0.76-1.07 2.64E-01 0.91 0.69-1.17 4.68E-01

16 MHC 18:0 0.95 0.79-1.13 5.65E-01 0.90 0.68-1.17 4.43E-01

17 . MHC 20:0 0.99 0.83-1.18 9.36E-01 0.97 0.74-1.26 8.34E-01

18 MHC 22:0 1.04 0.87-1.24 6.72E-01 1.03 0.79-1.32 8.42E-01

19 MHC 24:0 1.00 0.84-1.19 9.76E-01 0.95 0.73-1.22 7.10E-01

20 MHC 24: 1 0.83 0.69-0.98 3.43E-02 0.81 0.61-1.07 1.51E-01

21 DHC 16:0 0.78 0.65-0.93 6.14E-03 0.85 0.64-1.11 2.33E-01

22 DHC 18:0 0.94 0.78-1.11 4.80E-01 0.97 0.74-1.25 8.08E-01

23 DHC 20:0 0.96 0.80-1.14 6.41E-01 0.91 0.70-1.19 5.10E-01

24 DHC 22:0 0.94 0.78-1.12 4.82E-01 0.87 0.66-1.13 2.94E-01

25 DHC 24:0 0.91 0.76-1.07 2.64E-01 0.91 0.68-1.21 5.17E-01

26 DHC 24:1 0.81 0.67-0.96 1.82E-02 0.84 0.63-1.1 2.17E-01

27 THC 16:0 0.82 0.68-0.97 2.90E-02 0.80 0.59-1.09 1.64E-01

28 THC 18:0 0.86 0.72-1.03 l .lOE-01 0.79 0.59-1.05 1.16E-01

29 THC 20:0 0.92 0.76-1.11 4.08E-01 1.04 0.77-1.39 7.88E-01

30 THC 22:0 0.95 0.78-1.13 5.50E-01 0.98 0.73-1.28 8.64E-01

31 THC 24:0 0.89 0.74-1.06 2.01E-01 0.84 0.62-1.11 2.29E-01

32 THC 24:1 0.88 0.73-1.03 1.26E-01 0.86 0.64-1.14 2.92E-01

33 GM3 16:0 0.85 0.71-1.01 6.56E-02 0.76 0.55-1.02 7.37E-02

34 GM3 18:0 0.86 0.72-1.02 9.86E-02 0.82 0.62-1.08 1.64E-01

35 GM3 20:0 1.01 0.84-1.2 9.48E-01 0.93 0.69-1.23 6.08E-01

36 GM3 22:0 1.15 0.95-1.37 1.41E-01 1.00 0.76-1.3 9.99E-01

37 GM3 24:0 0.99 0.82-1.18 9.14E-01 0.98 0.74-1.29 8.92E-0I

38 GM3 24: 1 0.87 0.72-1.03 1.20E-01 0.82 0.61-1.1 2.01E-01

39 SM 14:0 0.97 0.80-1.16 7.51E-01 1.16 0.87-1.52 2.98E-01

40 SM 15:0 0.86 0.72-1.03 1.13E-01 1.12 0.85-1.45 4.19E-01

41 SM 16:0 0.96 0.80-1.15 6.95E-01 1.18 0.89-1.56 2.47E-01

42 SM 16: 1 0.83 0.68-1.01 6.88E-02 0.97 0.71-1.33 8.71E-01

43 SM 18:0 1.07 0.89-1.28 4.66E-01 1.07 0.80-1.41 6.32E-01

44 SM 18: 1 0.97 0.80-1.17 7.88E-01 1.11 0.81-1.5 5.07E-01

47 SM 22:0 1.22 1.01-1.45 3.39E-02 1.29 0.99-1.67 5.50E-02

53 modCer 576.5/7.7 1.36 1.12-1.63 1.31E-03 1.17 0.87-1.57 2.89E-01

54 modCer 614.6/5.7 0.67 · 0.55-0.8 2.75E-05 0.74 0.54-1.01 5.86E-02

55 modCer 632.6/9.2 0.92 0.77-1.1 3.83E-01 0.91 0.68-1.2 5.07E-01

56 modCer 651.6/7.6 1.05 0.87-1.25 5.84E-01 1.16 0.90-1.48 2.44E-01

57 modCer 731.6/6.2 1.02 0.84-1.21 8.62E-01 1.04 0.78-1.35 8.04E-01

58 modCer 766.6/7.2 0.95 0.80-1.13 6.03E-01 1.01 0.77-1.32 9.14E-01

59 modCer 769.6/8.0 1.15 0.96-1.37 1.15E-01 1.19 0.91-1.54 1.89E-01

60 modCer 875.7/9.2 1.17 0.97-1.39 9.40E-02 1.22 0.92-1.6 1.60E-01 incident diabetes pre-diabetes progressors

Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value modCer 883.8/7.8 1.05 0.87-1.26 5.73E-01 0.95 0.71-1.26 7.27E-01 modCer 798.7/7.3 0.96 0.80-1.14 6.43E-01 1.02 0.77-1.34 8.83E-01 modCer 886.8/9.1 1.54 1.27-1.86 9.02E-06 1.20 0.90-1.58 2.07E-01 modCer 910.8/9.0 1.43 1.18-1.73 1.89E-04 1.17 0.88-1.54 2.68E-01 modCer 921.8/9.1 1.22 1.02-1.46 2.89E-02 1.2 0.96-1.65 9.15E-02

PC 29:0 0.98 0.82-1.17 8.45E-01 1.01 0.77-1.32 9.21E-01

PC 31 :0 1.02 0.85-1.22 7.95E-01 1.10 0.84-1.42 4.90E-01

PC 31: 1 0.90 0.75-1.07 2.62E-01 1.11 0.84-1.46 4.61E-01

PC 33:0 0.77 0.63-0.93 7.31E-03 0.89 0.66-1.19 4.45E-01

PC 33:1 0.96 0.80-1.14 6.61E-01 0.97 0.74-1.25 7.99E-01

PC 33:2 0.89 0.73-1.06 1.96E-01 1.21 0.91-1.59 1.84E-01

PC 35:0 1.12 0.93-1.32 2.21E-01 1.22 0.91-1.61 1.69E-01

PC 35:1 0.90 0.75-1.07 2.35E-01 1.02 0.77-1.34 8.98E-01

PC 35:2 0.78 0.65-0.93 7.19E-03 1.18 0.90-1.54 2.29E-01

PC 35:3 1.10 0.91-1.32 3.13E-01 1.44 1.07-1.92 1.33E-02

PC 35:4 1.10 0.91-1.32 3.13E-01 1.44 1.07-1.92 1.33E-02

PC 37:3 0.82 0.68-0.97 2.47E-02 1.03 0.79-1.33 8.20E-01

PC 37:4 0.95 0.79-1.13 5.36E-01 1.31 1.00-1.71 4.37E-02

PC 37:5 1.09 0.90-1.31 3.48E-01 1.42 1.05-1.92 2.10E-02

PC 37:6 0.90 0.74-1.08 2.68E-01 1.13 0.84-1.5 4.07E-01

PC 28:0 1.05 0.87-1.25 5.94E-01 0.95 0.72-1.25 7.37E-01

PC 32:0 1.15 0.96-1.37 1.23E-01 1.25 0.94-1.65 1.21E-01

PC 32:1 1.24 1.01-1.49 3.16E-02 0.99 0.74-1.3 9.35E-01

PC 32:2 1.05 0.86-1.27 6.23E-01 1.05 0.77-1.42 7.29E-01

PC 34:0 1.14 0.95-1.36 1.43E-01 1.17 0.88-1.53 2.63E-01

PC 34:1 1.11 0.92-1.32 2.72E-01 1.01 0.75-1.33 9.62E-01

PC 34:2 0.97 0.81-1.15 7.73E-01 1.36 1.01-1.81 3.75E-02

PC 34:3 1.11 0.92-1.33 2.73E-01 1.13 0.84-1.5 4.21E-01

PC 34:5 1.08 0.90-1.29 4.02E-01 1.11 0.83-1.46 4.75E-01

PC 36:1 1.17 0.97-1.39 8.56E-02 1.28 0.98-1.66 6.14E-02

PC 36:2 1.01 0.84-1.2 9.05E-01 1.35 1.01-1.8 4.14E-02

PC 36:3 1.06 0.88-1.27 5.03E-01 1.39 1.01-1.89 3.88E-02

PC 36:4 1.22 1.01-1.46 3.18E-02 1.38 1.04-1.81 2.32E-02

PC 36:5 1.08 0.90-1.29 3.72E-01 1.14 0.86-1.5 3.51E-01

PC 38:4 1.33 1.05-1.67 1.41E-02 1.65 1.15-2.37 6.49E-03

PC 38:5 1.04 0.87-1.24 6.50E-01 1.20 0.91-1.56 1.94E-01

PC 38:6 1.04 0.87-1.24 6.50E-01 1.20 0.91-1.56 1.94E-01

PC 39:7 0.76 0.62-0.92 6.16E-03 0.84 0.63-1.12 2.45E-01

PC 40:5 1.09 0.91-1.31 3.34E-01 1.11 0.83-1.48 4.84E-01

PC 40:6 1.09 0.91-1.31 3.34E-01 1.1 1 0.83-1.48 4.84E-01

PC 40:7 0.84 0.69-1.01 6.88E-02 0.92 0.68-1.23 5.91E-01

PC 44:12 0.75 0.61-0.9 2.76E-03 0.90 0.67-1.18 4.43E-01

APC 30:0 0.78 0.64-0.94 9.39E-03 0.95 0.72-1.24 7.02E-01

APC 30: 1 0.84 0.69-1 5.97E-02 0.95 0.72-1.24 6.97E-01

APC 32:0 0.80 0.66-0.95 1.55E-02 1.00 0.76-1.31 9.93E-01

APC 32:1 0.84 0.69-1 5.11E-02 1.07 0.82-1.38 6.34E-01

APC 34:0 0.91 0.76-1.08 2.85E-01 1.13 0.87-1.46 3.40E-01

APC 34:1 0.77 0.63-0.93 7.31E-03 0.89 0.66-1.19 4.45E-01

APC 34:2a 0.84 0.69-1 5.22E-02 1.01 0.78-1.31 9.27E-01

APC 34:2b 0.70 0.57-0.85 3.90E-04 0.78 0.57-1.04 9.82E-02

APC 36:0 0.93 0.78-1.11 4.51E-01 0.99 0.76-1.28 9.48E-01

APC 36:1 0.80 0.66-0.96 2.09E-02 0.93 0.70-1.22 5.99E-01

APC 36:2 0.75 0.61-0.9 3.23E-03 0.91 0.68-1.2 5.05E-01

APC 36:3a 1.10 0.92-1.31 2.95E-01 1.41 1.06-1.87 1.65E-02

APC 36:3b 0.76 0.62-0.92 5.54E-03 0.86 0.64-1.15 3.15E-01

APC 36:4 1.15 0.96-1.38 1.20E-01 1.38 1.04-1.81 2.27E-02

APC 36:5 1.15 0.96-1.38 1.20E-01 1.38 1.04-1.81 2.27E-02 incident diabetes pre-diabetes progressors'

# Lipid Odds ratio 95% CI . p-value Odds ratio 95% CI p-value

127 APC 38:3 0.88 0.74-1.04 1.33E-01 0.94 0.74-1.19 6.13E-01

128 APC 38:4 1.08 0.90-1.29 4.06E-01 1.41 1.05-1.87 1.97E-02

352 APC 38:5 1.04 0.87-1.24 6.58E-01 1.52 1.12-2.05 5.71E-03

129 APC 38:6 0.98 0.81-1.17 8.33E-01 1.14 0.85-1.51 3.71E-01

130 LPAF 16:0 1.13 0.94-1.35 1.65E-01 1.02 0.77-1.34 8.72E-01

131 LPAF 18:0 1.05 0.87-1.25 6.08E-01 1.06 0.80-1.4 6.89E-01

132 LPAF 18: 1 1.15 0.96-1.37 1.16E-01 1.04 0.78-1.37 7.92E-01

133 LPAF 20:0 0.87 0.72-1.03 1.14E-01 0.84 0.62-1.14 2.67E-01

134 LPAF 22:0 0.78 0.64-0.93 7.72E-03 0.87 0.67-1.1 1 2.65E-01

135 LPAF 22:1 0.79 0.65-0.94 8÷89E-03 0.91 0.70-1.18 5.07E-01

136 LPAF 24:1 0.76 0.62-0.91 3.34E-03 0.85 0.63-1.14 2.94E-01

137 LPAF 24:2 0.83 0.69-0.99 4.81E-02 0.87 0.63-1.19 3.92E-01

138 LPC 14:0 1.28 1.07-1.53 6.76E-03 1.10 0.83-1.45 4.93E-01

139 LPC 15:0 1.08 0.90-1.28 3.87E-01 1.12 0.85-1.45 4.18E-01

140 LPC 16:0 1.33 1.11-1.59 1.73E-03 1.11 0.85-1.44 4.20E-01

141 LPC 16:1 1.18 0.98-1.4 7.14E-02 0.93 0.71-1.2 5.88E-01

143 LPC 18:0 1.27 1.06-1.51 8.92Ε-Ό3 1.10 0.85-1.42 4.63E-01

144 LPC 18:1 1.09 0.90-1.3 3.57E-01 0.92 0.68-1.22 5.62E-01

145 LPC 18:2 0.90 0.73-1.09 3.07E-01 1.02 0.74-1.38 9.22E-01

146 LPC 20:0 1.10 0.91-1.31 3.22E-01 1.10 0.84-1.44 4.86E-01

147 LPC 20: 1 1.00 0.83-1.2 9.64E-01 0.97 0.73-1.26 8.06E-01

148 LPC 20:2 1.01 0.84-1.21 8.88E-01 0.92 0.69-1.21 5.72E-01

149 LPC 20:3 1.17 0.97-1.39 9.75E-02 0.99 0.75-1.3 9.49E-01

150 LPC 20:4 1.26 1.04-1.51 1.39E-02 1.22 0.92-1.59 1.57E-01

151 LPC 20:5 1.08 0.89-1.29 4.24E-01 1.05 0.79-1.4 7.15E-01

353 LPC 22:0 1.06 0.88-1.26 5.32E-01 1.19 0.91-1.56 1.99E-01

152 LPC 22:6 1.05 0.88-1.25 5.79E-01 0.99 0.75-1.3 9.46E-01

153 LPC 24:0 1.04 0.86-1.24 6.64E-01 1.22 0.92-1.61 1.64E-01

154 modPC.506.3/3.4 1.23 1.02-1.47 2.27E-02 1.17 0.90-1.5 2.29E-01

155 modPC.512.3/1.7 0.95 0.79-1.14 5.97E-01 0.89 0.66-1.19 4.29E-01

156 modPC 620.4/2.6 1.10 0.91-1.3 3.06E-01 . 1.28 0.96-1.7 8.87E-02

158 modPC.536.3/3.5 1.00 0.83-1.19 9.78E-01 1.09 0.82-1.43 5.61E-01

159 modPC.538.3/4.1 1.04 0.86-1.25 6.75E-01 1.11 0.84-1.45 4.52E-01

160 modPC.594.4/3.1 1.15 0.96-1.37 l .l lE-01 1.24 0.93-1.64 1.39E-01

161 modPC.608.4/4.0 1.00 0.83-1.2 9.89E-01 1.14 0.87-1.48 3.30E-01

162 modPC.610.4/1.7 1.29 1.07-1.54 5.69E-03 1.36 1.04-1.76 2.27E-02

383 modPC.633.4/4.6 0.81 0.67-0.98 3.63E-02 0.91 0.67-1.23 5.39E-01

354 modPC.636.4/3.6 1.21 1.00-1.44 4.13E-02 1.15 0.86-1.54 3.42E-01

165 modPC.645.4/4.4 0.87 0.71-1.05 1.51E-01 0.88 0.64-1.2 4.26E-01

166 modPC.650.4/3.8 1.13 0.94-1.34 1.81E-01 1.09 0.81-1.46 5.46E-01

82 modPC.650.4/3.9 1.12 0.93-1.33 2.32E-01 1.13 0.87-1.45 3.66E-01

167 modPC.666.4/2.9 1.03 0.86-1.22 7.38E-01 0.99 0.74-1.31 9.52E-01

355 modPC 678.4/4.37 1.14 0.94-1.35 1.64E-01 1.20 0.90-1.59 2.1 1E-01

107 modPC 678.4/4.9 0.94 0.77-1.12 4.81E-01 1.13 0.84-1.49 4.14E-01

168 modPC.678.4/5.0 1.17 0.97-1.39 9.07E-02 1.31 0.98-1.74 6.79E-02

66 modPC.690.4/4.3 1.04 0.87-1.23 6.90E-01 1.13 0.85-1.49 3.95E-01

106 modPC.690.4/5.4 0.86 0.71-1.03 1.04E-01 1.08 0.81-1.41 5.98E-01

173 modPC.692.4/5.1 0.86 0.71-1.03 1.04E-01 0.90 0.68-1.18 4.37E-01

174 modPC.736.5/5.7 0.85 0.71-1.01 6.67E-02 1.00 0.75-1.32 9.87E-01

349 modPC.743.5/6.0 0.81 0.67-0.98 3.05E-02 1.05 0.77-1.41 7.72E-01

177 modPC.773.6/6.5 0.89 0.74-1.06 2.20E-01 1.03 0.78-1.34 8.36E-01

356 modPC 798.6/5.7 1.15 0.95-1.39 1.34E-01 1.18 0.90-1.53 2.34E-01

179 modPC.801.6/6.8 1.00 0.83-1.19 9.93E-01 1.17 0.89-1.52 2.41E-01

180 modPC.818.6/6.6 0.91 0.75-1.09 3.32E-01 1.12 0.83-1.51 4.38E-01

183 modPC.843.6/7.2 0.93 0.77-1.1 4.10E-01 1.05 0.80-1.35 7.32E-01

184 modPC.866.6/7.2 0.82 0.68-0.98 3.41E-02 0.94 0.70-1.24 6.70E-01

185 modPC.877.6/6.0 0.99 0.83-1.18 9.52E-01 0.88 0.66-1.14 3.42E-01 incident diabetes pre-diabetes progressors

# Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value

186 modPC 879.1/6.1 0.84 0.69-1 5.52E-02 0.85 0.63-1.12 2.47E-01

357 PE Unk 9b 1.05 0.87-1.25 6.13E-01 1.17 0.83-1.62 3.62E-01

187 PG 16:0 18:1 1.32 1.09-1.58 4.07E-03 1.06 0.78-1.41 7.18E-01

188 PG 16:1 18:1 1.37 1.13-1.64 9.60E-04 T.T3 0.85-1.49 3.91E-01

189 PG 18:0 18:1 1.38 1.13-1.68 1-20E-03 1.09 0.80-1.48 5.77E-01

190 PG 18:1 18:1 1.44 1.18-1.74 2.09E-04 1.25 0.93-1.68 1.33E-01

191 PE 32:0 1.37 1.14-1.64 5.38E-04 1.19 0.90-1.55 2.14E-01

192 PE 32:1 1.30 1.07-1.57 7.41E-03 1.00 0.74-1.32 9.78E-01

194 PE 34:1 1.33 1.10-1.6 2.53E-03 0.99 0.74-1.31 9.42E-01

195 PE 34:2 1.34 1.11-1.61 2.10E-03 1.12 0.84-1.49 4.32E-01

358 PE 35:2 1.17 0.97-1.39 8.38E-02 1.07 0.80-1.41 6.27E-01

359 PE 35:3 1.08 0.90-1.29 3.81E-01 0.98 0.74-1.29 8.97E-01

196 PE 36:0 1.01 0.84-1.19 9.50E-01 1.03 0.77-1.36 8.61E-01

197 PE 36:1 1.44 1.19-1.74 1.69E-04 1.19 0.89-1.59 2.34E-01

198 PE 36:2 1.37 1.13-1.64 1.03E-03 1.16 0.87-1.53 3.08E-01

199 PE 36:3 1.30 1.08-1.55 5.07E-03 1.12 0.84-1.48 4.19E-01

200 PE 36:4 1.46 1.20-1.76 1.01E-04 1.25 0.92-1.68 1.51E-01

201 PE 36:5 1.30 1.08-1.55 5.26E-03 1.08 0.81-1.42 6.08E-01

202 PE 38:1 1.39 1.14-1.68 8.92E-04 1.13 0.86-1.47 3.78E-01

203 PE 38:2 1.39 1.14-1.67 7.32E-04 1.14 0.85-1.52 3.76E-01

204 PE 38:3 1.43 1.17-1.73 4.08E-04 1.11 0.81-1.52 5.02E-01

205 PE 38:4 1.37 1.12-1.66 1.35E-03 1.28 0.94-1.72 1.05E-01

206 PE 38:5 1.30 1.08-1.55 5.21E-03 1.31 0.96-1.77 7.83E-02

207 PE 38:6 .1.31 1.08-1.59 5.48E-03 1.14 0.84-1.53 3.78E-01

360 PE 39:4 0.98 0.81-1.16 7.95E-01 0.91 0.70-1.17 4.68E-01

361 PE 40:4 1.59 1.30-1.93 4.59E-06 1.31 0.95-1.78 9.50E-02

208 PE 40:6 1.44 1.18-1.74 2.86E-04 1.22 0.90-1.64 2.00E-01

209 PE 40:7 1.23 1.01-1.47 . 3.17E-02 1.06 0.79-1.4 7.12E-01

362 PE Unk 9a 1.01 0.83-1.2 9.51E-01 1.01 0.77-1.31 9.38E-01

363 PE 40:5 1.29 1.07-1.55 5.52E-03 1.18 0.87-1.59 2.68E-01

364 APE 36:2a 0.94 0.78-1.12 4.94E-01 1.04 0.80-1.34 7.87E-01

365 APE 36:2b 1.18 0.98-1.41 6.56E-02 1.07 0.83-1.38 5.80E-01

366 APE 36:3a 1.01 0.84-1.2 9.42E-01 0.95 0.72-1.24 6.97E-01

367 APE 36:3b 1.05 0.87-1.24 6.05E-01 1.13 0.88-1.44 3.31E-01

368 APE 36:4 1.07 0.89-1.27 4.69E-01 1.14 0.86-1.49 3.59E-01

369 APE 38:4a 1.10 0.92-1.31 2.97E-01 1.07 0.81-1.41 6.30E-01

370 APE 38:4b 1.08 0.90-1.29 3.72E-01 1.09 0.84-1.41 5.05E-01

371 APE 38:5a 1.10 0.91-1.3 3.15E-01 1.04 0.79-1.35 7.91E-01

372 APE 38:5b 1.24 1.03-1.48 1.85E-02 1.28 0.99-1.66 5.79E-02

373 APE 40:5 1.01 0.84-1.2 8.87E-01 1.07 0.82-1.39 5.98E-01

374 APE 40:6 1.04 0.87-1.24 6.54E-01 1.07 0.82-1.38 6.15E-01

375 lyso PE 16:0 1.29 1.07-1.54 6.33E-03 0.95 0.71-1.26 7.38E-01

376 lyso PE 18:0 ,1.41 1.17-1.68 2.14E-04 0.90 0.67-1.19 4.64E-01

377 lyso PE 18:1 1.14 0.95-1.35 1.55E-01 1.03 0.78-1.34 8.37E-01

378 lyso PE 18:2 1.03 0.86-1.23 7.14E-01 1.15 0.88-1.51 2.95E-01

210 PI 32:0 1.18 0.98-1.41 7.78E-02 1.04 0.80-1.36 7.49E-01

211 PI 32:1 1.25 1.03-1.51 2.00E-02 0.92 0.70-1.2 5.58E-01

212 PI 34:0 1.15 0.95-1.37 1.48E-01 1.11 0.83-1.46 4.72E-01

213 PI 34:1 1.15 0.95-1.37 1.30E-01 0.88 0.68-1.14 3.43E-01

214 PI 36:0 1.03 0.85-1.24 7.28E-01 0.96 0.73-1.26 7.89E-01

215 PI 36:1 1.09 0.91-1.29 3.55E-01 0.91 0.70-1.17 4.61E-01

216 PI 36:2 1.14 0.95-1.36 1.55E-01 1.08 0.83-1.4 5.59E-01

217 PI 36:3 1.14 0.95-1.36 1.60E-01 0.97 0.74-1.28 8.50E-01

218 PI 36:4 1.34 1.11-1.61 1.84E-03 1.09 0.82-1.44 5.29E-01

219 PI 38:2 1.06 0.88-1.26 5.10E-01 0.86 0.66-1.13 2.87E-01

220 PI 38:3 1.10 0.92-1.31 2.89E-01 0.90 0.67-1.19 4.56E-01

221 PI 38:4 1.22 1.01-1.45 2.96E-02 1.15 0.87-1.52 3.08E-01 incident diabetes pre-diabetes progressors

# Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value

222 PI 38:5 1.17 0.98-1.4 7.50E-02 1.05 0.79-1.38 7.40E-01

223 PI 38:6 1.09 0.90-1.3 3.56E-01 0.88 0.67-1.15 3.65E-01

224 PI 40:4 1.24 1.03-1.48 2.04E-02 0.97 0.73-1.26 8.06E-O1

225 PI 40:5 1.24 1.03-1.48 1.94E-02 1.08 0.82-1.41 5.87E-01

226 PI 40:6 1.14 0.94-1.36 1.72E-01 1.00 0.76-1.3 9.85E-01

232 PS 36: 1 1.17 0.97-1.39 9.39E-02 1.16 0.85-1.56 3.40E-01

233 PS 36:2 1.13 0.94-1.34 1.91E-01 1.19 0.88-1.59 2.58E-01

236 PS 38:3 1.08 0.90-1.29 3.95E-01 1.12 0.82-1.5 4.72E-01

237 PS 38:4 1.14 0.95-1.36 1.38E-01 1.18 0.89-1.56 2.35E-01

238 PS 38:5 1.08 0.90-1.28 4.02E-01 1.15 0.87-1.52 3.22E-01

240 PS 40:5 1.16 0.96-1.39 1.18E-01 1.21 0.89-1.62 2.09E-01

241 PS 40:6 1.16 0.96-1.38 1.04E-01 1.21 0.90-1.6 1.94E-01

242 COH 1.12 0.93-1.33 2.18E-01 1.07 0.81-1.4 6.31E-01

243 CE 14:0 1.42 1.17-1.72 3.24E-04 1.13 0.84-1.49 4.11E-01

244 CE 15:0 1.17 0.97-1.4 8.38E-02 1.06 0.80-1.39 6.93E-01

245 CE 16:0 1.27 1.05-1.52 1.10E-02 1.07 0.80-1.41 6.54E-01

246 CE 16: 1 1.40 1.15-1.71 7.42E-04 1.02 0.76-1.35 9.05E-01

247 CE 16:2 1.44 1.18-1.74 2.48E-04 1.09 0.81-1.46 5.45E-01

248 CE 17:0 1.08 0.90-1.29 3.71E-01 1.04 0.78-1.37 8.02E-01

249 CE 17:1 1.26 1.04-1.52 1.74E-02 0.99 0.74-1.31 9.50E-01

250 CE 18:0 1.22 1.01-1.45 3.46E-02 1.20 0.91-1.56 1.97E-01

251 . CE 18: 1 1.38 1.14-1.66 7.63E-04 1.01 0.74-1.37 9.30E-01

252 CE 18:2 1.43 1.18-1.72 1.81E-04 1.19 0.90-1.57 2.04E-01

253 CE 18:3 1.47 1.21-1.78 8.81E-05 1.17 0.87-L57 2.88E-01

254 CE 20:0 1.21 1.01-1.45 3.44E-02 1.00 0.75-1.32 9.74E-01

255 CE 20:1 1.10 0.91-1.31 2.98E-01 1.13 0.86-1.47 3.66E-01

256 CE 20:2 1.19 0.99-1.43 5.42E-02 0.99 0.74-1.3 9.38E-01

257 CE 20:3 1.46 1.19-1.79 2.63E-04 1.06 0.78-1.44 6.98E-01

258 CE 20:4 1.53 1.25-1.85 1.95ΕΌ5 1.26 0.92-1.69 1.41E-01

259 CE 20:5 1.32 1.09-1.59 3.36E-03 1.1 1 0.82-1.48 4.94E-01

260 CE 22:0 1.38 1.14-1.65 6.00E-04 1.21 0.93-1.56 1.53E-01

261 CE 22: 1 1.11 0.92-1.32 2.62E-01 1.13 0.86-1.49 3.71E-01

262 CE 22:2 1.00 0.83-1.19 9.80E-01 1.02 0.78-1.33 8.67E-01

263 CE 22:3 0.97 0.81-1.15 7.51E-01 0.91 0.70-1.17 4.61E-01

264 CE 22:4 1.43 1.17-1.73 2.98E-04 1.19 0.88-1.61 2.57E-01

265 CE 22:5 1.40 1.16-1.69 4.70E-04 1.12 0.82-1.5 4.78E-01

266 CE 22:6 1.29 1.07-1.55 7.34E-03 1.05 0.78-1.39 7.46E-01

267 CE 24:0 1.39 1.15-1.67 4.63E-04 1.26 0.96-1.65 8.69E-02

268 CE 24:1 1.13 0.94-1.35 1.89E-01 1.04 0.78-1.39 7.70E-01

269 CE 24:2 0.98 0.82-1.17 8.57E-01 1.03 0.78-1.35 8.41E-01

270 CE 24:3 1.02 0.85-1.21 8.33E-01 0.96 0.73-1.25 7.69E-01

271 CE 24:4 1.05 0.87-1.25 6.03E-01 1.01 0.76-1.32 9.34E-01

272 CE 24:5 1.21 1.01-1.45 3.48E-02 1.03 0.77-1.36 8.40E-01

273 CE 24:6 1.22 1.01-1.46 3.87E-02 0.92 0.68-1.25 6.14E-01

274 modCE 558.5/7.74 1.40 1.16-1.69 4.15E-04 1.16 0.86-1.54 3.27E-01

275 modCE 588.5/7.94 1.35 1.1 1-1.62 1.62E-03 1.19 0.88-1.59 2.62E-01

276 modCE 682.7/8.76 1.30 1.08-1.55 4.82E-03 1.08 0.81-1.43 5.82E-01

277 modCE 790.8/6.57 1.43 1.18-1.73 2.06E-04 1.22 0.91-1.62 1.81E-01

278 DG 14:0 14:0 1.41 1.16-1.7 4.29E-04 1.24 0.92-1.66 1.44E-01

279 DG 14:0 16:0 1.54 1.26-1.87 1.67E-05 1.32 0.97-1.77 6.97E-02

280 DG 14:0 18:1 1.46 1.20-1.77 1.40E-04 1.24 0.91-1.67 1.66E-01

281 DG 14:0 18:2 1.41 1.16-1.69 3.62E-04 1.31 0.98-1.74 5.95E-02

282 DG 14:1 16:0 1.51 1.24-1.83 3.37E-05 1.32 0.98-1.77 6.21E-02

283 DG 16:0 16:0 1.67 1.36-2.05 6.83E-07 1.39 1.02-1.89 3.60E-02

284 DG 16:0 18:0 1.68 1.36-2.05 5.63E-07 1.44 1.05-1.97 2.11E-02

285 DG 16:0 18: 1 1.65 1.34-2.02 1.43E-06 1.39 1.01-1.9 4.13E-02

286 DG 16:0 18:2 1.57 1.29-1.9 5.09E-06 1.42 1.06-1.89 1.72E-02 incident diabetes pre-diabetes progressors

# Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value

287 DG 16:0 20:0 1.24 1.03-1.48 1.90E-02 1.36 1.04-1.76 2.21E-02

288 DG 16:0 20:3 1.53 1.25-1.85 1.82E-05 1.35 1.00-1.81 4.74E-02

289 DG 16:0 20:4 1.71 1.40-2.08 7.90E-08 1.59 1.16-2.16 3.60E-03

290 DG 16:0 22:5 1.50 1.23-1.81 3.56E-05 1.43 1.05-1.93 1.94E-02

291 DG 16:022:6 1.42 1.17-1.71 3.21E-04 1.24 0.93-1.65 1.35E-01

292 DG 16: 1 18:1 1.40 1.15-1.7 6.39E-04 1.16 0.86-1.55 3.35E-01

293 DG 18:0 16:1 1.55 1.26-1.9 2.74E-05 I.3I 0.96-1.76 8.60E-02

294 DG 18:0 18:0 1.43 1.17-1.72 2.88E-04 1.30 0.96-1.74 8.56E-02

295 DG 18:0 18: 1 1.61 1.31-1.96 3.98E-06 ^* 1.42 1.03-1.94 2.76E-02

296 DG 18:0 18:2 1.52 1.25-1.84 1.96E-05 1.44 1.07-1.92 1.33E-02

297 DG 18:0 20:4 1.47 1.22-1.77 4.92E-05 1.67 1.22-2.27 1.09E-03

298 DG 18:1 18:1 1.39 1.15-1.68 6.46E-04 1.23 0.91-1.66 1.71E-01

299 DG 18: 1 18:2 1.31 1.09-1.57 3.48E-03 1.28 0.96-1.68 8.15E-02

300 DG 18: 1 18:3 1.36 1.13-1.63 1.00E-03 1.32 0.99-1.73 5.29E-02

301 DG 18: 1 20:0 1.13 0.94-1.34 I .78E-01 1.20 0.92-1.56 1.74E-01

302 DG 18: 1 20:3 1.29 1.07-1.54 6.72E-03 1.21 0.91-1.58 1.85E-01

303 DG 18: 1 20:4 1.53 1.27-1.85 8.29E-06 1.50 1.11-2.01 7.73E-03

304 DG 18:2 18:2 1.23 1.02-1.47 2.32E-02 1.29 0.98-1.68 6.28E-02

305 TG 14:0 16:0 18:2 1.37 1.12-1.67 1.52E-03 1.20 0.89-1.6 2.25E-01

306 TG 14:0 16:1 18:1 1.24 1.02-1.49 2.80E-02 1.07 0.80-1.44 6.32E-01

307 TG 14:0 16:1 18:2 1.21 1.00-1.45 4.53E-02 1.15 0.86-1.51 3.40E-01

308 TG 14:0 18:0 18: 1 1.43 1.17-1.75 4.83E-04 1.28 0.94-1.74 1.13E-01

309 TG 14:0 18:2 18:2 1.19 0:99-1.42 5.33E-02 1.24 0.94-1.63 1.20E-01

310 TG 14:1 16:0 18:1 1.34 1.09-1.62 3.52E-03 1.18 0.88-1.58 2.62E-01

311 TG 14: 1 16:1 18:0 1.34 1.10-1.63 3.65E-03 1.10 0.82-1.47 5.12E-01

312 TG 14:1 18:0 18:2 1.26 1.04-1.51 1.61E-02 1.17 0.87-1.56 2.78E-01

313 TG 14:1 18: 1 18: 1 1.21 1.00-1.45 4.84E-02 1.11 0.83-1.47 4.82E-01

314 TG 15:0 18:0 18:1 1.19 0.99-1.42 6.35E-02 1.21 0.90-1.61 1.94E-01

315 TG 15:0 18: 1 16:0 1.39 1.14-1.69 9.90E-04 1.29 0.95-1.74 9.37E-02

316 TG 16:0 16:0 16:0 1.55 1.26-1.9 3.10E-05 1.31 0.95-1.78 9.15E-02

317 TG 16:0 16:0 18:0 1.54 1.25-1.89 3.96E-05 1.34 0.98-1.82 6.23E-02

318 TG 16:0 16:0 18: 1 1.56 1.26-1.91 2.73E-05 1.33 0.96-1.82 8.26E-02

319 TG 16:0 16:0 18:2 1.53 1.25-1.86 2.44E-05 1.43 1.05-1.93 2.18E-02

320 TG 16:0 16: 1 18: 1 1.34 1.09-1.62 3.97E-03 1.14 0.83-1.53 4.10E-01

321 TG 16:0 18:0 18:1 1.58 1.28-1.95 1.65E-05 1.37 0.99-1.88 5.18E-02

322 TG 16:0 18:1 18: 1 1.39 1.14-1.69 9.64E-04 1.27 0.92-1.74 1.37E-01

323 TG 16:0 18: 1 18:2 1.32 1.09-1.58 3.88E-03 1.34 1.00-1.78 4.60E-02

324 TG 16:0 18:2 18:2 1.29 .1.07-1.55 5.87E-03 1.36 1.03-1.8 2.88E-02

325 TG 16:1 16: 1 18:1 1.32 1.08-1.59 5.50E-03 1.15 0.86-1.54 3.30E-01

326 TG 16: 1 16: 1 16:1 1.23 1.01-1.48 3.27E-02 1.08 0.81-1.42 6.07E-01

327 TG 16:1 16: 1 18:0 1.41 1.16-1.72 6.06E-04 1.29 0.96-1.73 9.02E-02

328 TG 16: 1 18: 1 18:1 1.18 0.98-1.42 7.51E-02 1.05 0.78-1.4 7.33E-01

329 TG 16:1 18: 1 18:2 1.24 1.03-1.49 2.05E-02 1.23 0.93-1.62 1.45E-01

330 TG 17:0 16:0 16: 1 1.42 1.16-1.74 5.20E-04 1.31 0.96-1.77 8.23E-02

331 TG 17:0 16:0 18:0 1.48 1.20-1.81 1.59E-04 1.33 0.97-1.8 6.72E-02

332 TG 17:0 18: 1 14:0 1.34 1.10-1.62 3.40E-03 1.29 0.95-1.73 9.69E-02

333 TG 17:0 18: 1 16:0 1.44 1.17-1.76 3.68E-04 1.35 0.98-1.84 5.90E-02

334 TG 17:0 18:1 16:1 1.28 1.05-1.54 1.22E-02 1.23 0.91-1.66 1.68E-01

335 TG 17:0 18: 1 18:1 1.30 1.07-1.56 6.91E-03 1.27 0.94-1.72 1.18E-01

336 TG 17:0 18:2 16:0 1.40 1.14-1.71 9.51E-04 1.31 0.96-1.79 8.54E-02

337 TG 18:0 18:0 18:0 1.34 1.10-1.63 2.92E-03 1.35 1.00-1.8 4.28E-02

338 TG 18:0 18:0 18:1 1.55 1.26-1.9 3.34E45 1.42 1.04-1.93 2.70E-02

339 TG 18:0 18: 1 18:1 1.51 1.23-1.84 6.63E-05 1.35 0.99-1.84 5.69E-02

340 TG 18:0 18:2 18:2 1.31 1.08-1.57 4.66E-03 1.39 1.05-1.83 2.09E-02

341 TG 18:1 14:0 16:0 1.39 1.14-1.69 1.12E-03 1.21 0.89-1.64 2.16E-01

342 TG 18: 1 18: 1 18:1 1.27 1.05-1.52 1.23E-02 1.19 0.88-1.58 2.49E-01

343 TG 18: 1 18: 1 18:2 1.19 0.99-1.41 6.35E-02 1.24 0.94-1.63 1.15E-01 . incident diabetes pre-diabetes progressors

# Lipid Odds ratio 95% CI p-value Odds ratio 95% CI p-value

344 TG 18:1 18:1 20:4 1.39 1.14-1.67 6.22E-04 1.56 1.15-2.11 3.84E-03

345 TG 18:1 18:1 22:6 1.14 0.94-1.36 1.62E-01 1.13 0.85-1.49 3.80E-01

346 TG 18:1 18:2 18:2 1.22 1.01-1.46 3.12E-02 1.30 0.98-1.72 6.13E-02

347 TG 18:2 18:2 18:2 1.18 0.98-1.4 7.59Ετ02 1.30 0.99-1.71 5.69E-02

348 TG 18:2 18:2 20:4 1.40 1.15-1.68 4.58E-04 1.68 1.24-2.27 7.81E-04 basaed on the analysis of only the pre-diabetes at baseline group.

TABLE 11. Mean and standard deviation of individual lipid species in the longitudinal study'

Incident non- Pre-diabetes non

Whole population Incident diabetes diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev n Lipid 653 653 223 223 430 430 260 260 165 165 95 95

1 Sph 18:1 131 117 128 112 132 120 131 121 135 121 125 121

2 dhCer 16:0 59 18 61 19 57 17 62 20 62 20 62 21

3 dhCer 18:0 73 32 82 33 69 30 83 36 85 33 79 40

4 dhCer l8:l 14 6 15 7 14 6 15 7 15 7 15 6

5 dhCer20:0 31 11 33 11 30 11 34 13 34 11 34 17

6 dhCer22:0 123 52 140 55 114 47 139 61 144 57 130 66

7 dhCer 24:0 254 102 285 111 238 93 283 117 293 112 266 123

8 dhCer24:l 135 53 147 52 128 52 149 58 150 52 147 66

9 Cer 16:0 294 66 303 72 289 63 302 72 304 76 299 66

10 Cer 18:0 155 56 171 58 146 53 169 57 175 59 158 52

11 Cer 20:0 135 42 146 43 129 40 143 41 146 44 138 36

12 Cer 22:0 894 278 974 302 852 255 956 305 987 319 902 274

13 Cer 24:0 2781 742 2911 796 2714 705 2872 820 2931 841 2770 774

14 Cer 24:1 1168 317 1207 316 1147 316 . 1210 309 1210 326 1212 280

15 MHC 16:0 672 191 658 197 679 188 655 188 648 193 667 179

16 MHC 18:0 156 53 153 53 158 53 152 52 149 49 158 56

17 MHC 20:0 149 50 148 52 150 49 146 51 144 51 148 52

18 MHC 22:0 1289 414 1286 425 1290 408 1268 430 1265 426 1273 439

19 MHC 24:0 1853 579 1839 611 1860 562 1819 603 1802 609 1848 596

20 MHC 24:1 1349 460 1274 446 1388 463 1280 434 1246 442 1339 415

21 DHC 16:0 3890 1071 3672 1043 4004 1070 3690 1013 3634 1043 3786 957

22 DHC 18:0 96 32 95 35 96 31 94 34 94 36 94 30

23 DHC 20:0 79 28 78 27 80 28 77 27 76 . 27 80 27

24 DHC 22:0 235 72 230 73 238 71 230 74 224 75 239 71

25 DHC 24.0 296 89 288 94 300 85 286 92 282 95 294 85

26 DHC 24:1 968 316 903 298 1001 320 906 293 881 299 948 280

27 THC 16:0 716 201 676 193 737 202 681 190 661³ 187 714 190

28 THC 18:0 86 31 83 35 87 29 84 32 ^■82 35 87 26

29 THC 20:0 30 13 29 13 30 13 28 13 29 13 28 11

30 THC 22:0 117 44 112 46 119 42 111 45 109 46 114 42

31 THC 24:0 135 47 127 48 139 46 127 48 123 49 134 46

32 THC 24:1 270 96 . 250 84 281 100 251 88 242 83 268 94

Incident non- Pre-diabetes ηαη- Whole population Incident diabetes diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 16S 95 95

33 G 3 16:0 526 133 504 127 538 135 510 128 493 121 539 136

34 GM3 18:0 216 66 207 66 221 65 212 67 206 65 222 69

35 GM3 20:0 126 ^' 36 127 38 126 35 127 37 125 37 130 35

36 GM3 22:0 274 86 283 93 269 82 283 94 282 95 285 92

37 GM3 24:0 237 77 233 82 239 74 234 82 230 83 241 82

38 GNU 24: 1 389 125 374 120 398 126 381 124 369 121 402 126

39 SM 14:0 13513 3147 13517 3205 13511 3121 13435 3144 13525 3185 13279 3084

40 SM 15:0 9305 2119 9033 2048 9447 2143 9054 2145 9044 2079 9072 2265

41 SM 16:0 144046 21848 142563 21140 144815 22192 141371 21291 141703 20123 140792 23281

42 SM 16:1 21159 3996 20956 3987 21264 4001 20897 . 3817' 20816 3805 21037 3854

43 SM 18:0 32622 6652 33271 6721 32285 6598 33173 6749 33520 6699 32569 6828

44 SM 18:1 16347 3863 16475 3758 16281 3920 16418 3720 16555^■- 3650 16179 3847

47 SM 22:0 42703 7743 43961 8270 42051 7381 43252 8078 44136 8023 41718 7984

53 modCer 576.5/7.7 14 5 16 5 14 5 15 5 16 5 15 4

54 modCer 614.6/5.7 35 9 33 8 36 9 . 33 8 32 8 34 7

55 modCer 632.6/9.2 2 1 2 1 2 1 2 1 2 1 2 1

56 modCer 651.6/7.6 2 1 2 1 2 1 2 . 1 2 1 2 1

57 modCer 731.6/6.2 4 2 5 2 4 1 4 2 4 2 4 2

58 modCer 766.6/7.2 38 12 37 12 38 12 36 12 36 12 36 11

59 modCer 769.6/8.0 145 40 149 41 142 40 147 42 149 42 142 42

60 modCer 875.7/9.2 554 167 587 182 537 156 585 173 604 181 552 153

61 modCer 883.8/7.8 150 42 153 40 149 43 153 39 152 40 154 38

62 modCer 798.7/7.3 166 49 162 47 169 50 159 47 158 48 159 45

63 modCer 886.8/9.1 65 19 72 21 62 17 71 21 73 22 68 18

64 modCer 910.8/9.0 46 13 49 14· 44 11 48 13 49 14 47 11

65 modCer 921.8/9.1 23 13 25 14 22 13 26 14 27 15 23 13

67 PC 29:0 111 51 111 55 111 49 110 52 110 55 109 46

68 PC 31:0 1602 567 1623 615 1591 540 1598 593 1620 632 1560 520

69 PC 31: 1 2100 458 2070 443 2115 465 2064 449 2068 445 2056 458

70 PC 33:0 399 75 382 76 409 73 379 71 373 7⁴ 388 67

71 PC 33: 1 5436 1294 5445 1335 . 5431 1274 5426 1303 5406 1320 5461 1281

72 PC 33:2 4792 1262 4607 1227 4889 1271 4561 1218 4594 1210 4504 1237

73 PC 35:0 493 155 500 154 489 156 497 149 506 152 481 144

Incident non- Pre-diabetes non- Whole population Incident diabetes . diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean S Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

74 PC 35:1 12155 2221 11947 2267 12264 2192 11866 2155 ^■ 11839 2195 11913 2094

75 PC 35:2 12964 2742 12289 2556 13314 2772 12214 2613 12267 2554 12120 2724

76 PC 35:3 224 67 224 67 224 67 223 73 227 69 216 80

78 PC 35:4 2212 661 2210 664 2213 ; 660 2200 725 2240 682 2131 792

79 PC 37:3 4968 1162 4794 1144 5058 1163 4852 1191 4838 1157 4877 1253

80 PC 37:4 7663 1840 7504 1956 7746 1773 7466 1934 7613 2033 7210 1730

81 PC 37:5 1778 655 1750 575 1793 694 1737 601 1763 588 1693 623

84 PC 37:6 1132 419 075 392 1161 429 1089 400 1088 395 1092 412

85 PC 28:0 370 289 381 283 365 292 358 236 356 238 361 233

86 PC 32:0 12925 2368 13057 2289 12857 2408 12938 2317 13047 2277 12748 2384

87 PC 32:1 1791 526 1916 537 1726 509 1905 551 1916 537 1886 578

88 PC 32:2 663 161 672 156 658 164 667 . 158 668 148 665 174

89 PC 34:0 4755 866 4810 922 4726 835 4761 889 4798 945 4697 784

90 PC 34:1 150439 25449 152803 25363 149213 25437 152688 25278 152719 24312 152635 27008

91 PC 34:2 276063 47810 273699 45322 277289 49057 270553 43905 274655 43795 263428 43407

92 PC 34:3 24459 6882 24944 6854 24208 6891 24503 6624 24635 6348 24275 7106

93. PC 34:5 279 151 293 159 271 146 287 150 293 158 276 136

94 PC 36:1 11234 2112 11483 2168 11105 2073 11331 2171 11528 2060 10990 2322

95 PC 36:2 212315 37330 212225 35897 212362 38092 209210 35064 212620 35835 203286 33036

98 PC 36:3 868561 144656 877591 133944 863878 149847 869325 129239 883147 127960 845319 128595

99 PC 36:4 118174 29083 122383 29593 115991 28606 121881 30347 125399 30591 115772 29075

100 PC 36:5 36769 13665 37707 12872 36283 14048 37882 13516 38284 13389 37185 13777

101 PC 38:4 104891 26869 110162 28399 102157 25650 108755 27930 112706 28932 101891 24779

102 PC 38:5 22865 6456 22704 6077 22949 6650 22979 6277 23100 5964 22768 6813

103 PC 38:6 48399 13665 48058 12862 48576 14075 48639 13286 48896 12624 48193 14421

104 PC 39:7 135 74 119 50 144 83 125 67 118 50 139 88

105 PC 40:5 12207 3813 12570 3792 12018 3814 12764 3738 12888 3725, 12550 3770

108 PC 40:6 26370 8236 27156 8192 25962 8239 27575 8075 27842 8047 27111 8145

109 PC 40:7 • 5625 1644 5345 1531 5771 1683 5446 1532 5332 1457 5645 1644

110 PC 44: 12 1811 528 1701 525 1868 521 1704 499 1681 525 1743 450

111 APC 30:0 195 59 184 61 201 57 183 55 181 57 185 51

112 APC 30:1 137 42 131 42 140 41 131 41 129 41 134 40

350 APC 32:0 2183 409 2108 409 2222 404 2097 394 ^'2097 415 2097 358

Incident non- Pre-diabetes non< Whole population Incident diabetes diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

114 APC 32:1 1609 326 1548 330 1640 319 1539 325 1536 329 1544 320

115 APC 34:0 545 119 535 121 551 117 526 120 528 119 522 122

116 APC 34:1 4283 806 4094 811 4382 . 786 4061 766 4002 790 4163 717

117 APC 34:2a 4401 1218 4139 1111 4537 1250 4165 1229 4117 1118 4250 1403

351 APC 34:2b 2719 637 2531 . 628 2816 621 2530 604 2455 606 2661 579

118 APC 36:0 75 19 74 19 76 19 73 19 72 19 73 19

119 APC 36:1 1098 258 1049 259 1123 254 1034 245 1019 242 1061 251

120 APC 36:2 3598 867 3365 805 3719 874 3352 855 3277 775. 3483 970

121 APC 36:3a 1333 227 1334 228 1333 227 1325 220 1349 221 1283 212

122 APC 36:3b 3139 889 2888 835 3269 890 2900 867 2813 832 3051 909

123 APC 36:4 919 192 923 202 916 186 915 186 932 194 885 168

126 APC 36:5 9009 1881 9056 1986 8985 1827 8971 1821 9136 1901 8684 1644

127 APC 38:3 1460 392 1428 409 1476 383 1412 398 1403 392 1427 410

128 APC 38:4 8091 1625 8042 1600 8117 1639 7923 1526 8063 1536 7679 1485

352 APC 38:5 12313 2086 12174 2054 12385 2102 12075 1933 12273 2009 11732 1752

129 APC 38:6 6653 1563 6521 1453 6721 1614 6584 1433 6609 1405 6542 1487

130 LPAF 16:0 521 152 525 159 520 149 521 155 522 162 520 144

131 LPAF 18:0 148 45 146 45 150 46 144 44 144 45 143 41

132 LPAF 18:1 372 111 373 112 372 111 368 110 366 111 371 108

133 LPAF 20:0 8 3 7 3 8 3 7 3 7 2 8 3

134 LPAF 22:0 8 2 8 2 8 2 8 2 8 2 8 2

135 LPAF 22:1 13 7 12 7 14 8 12 7 12 7 13 7

136 LPAF 24:1 19 6 17 6 19 6 17 5 17 5 18 5

137 LPAF 24:2 4 2 3 1 4 2 3 1 3 1 3 1

138 LPC 14:0 1894 689 2020 724 1829 661 1979 715 2010 739 1926 672

139 LPC 15:0 1071 328 1069 354 1071 314 1058 344 1067 366 1043 304

140 LPC 16:0 82600 20467 85912 21786 80883 19555 84923 21641 85622 22271 83708 20561

141 LPC 16:1 3837 1320 4007 1463 3749 1232 4010 1449 3967 1485 4086 1390

143 LPC 18:0 26211 7633 27403 8667 25593 6968 26870 8385 27218 8934 26265 7341

144 LPC 18:1 20198 5901 20002 6138 20300 5780 19741 5863 19404 6082 20327 5445

145 LPC 18:2 20297 6462 18863 5448 21041 6818 18813 5977 18397 5300 19536 6972

146 LPC 20:0 118 37 117 41 119 35 114 40 113 40 115 39

147 LPC 20:1 273 87 264 88 277 87 263 92 257 89 274 97

Incident non- Pre-diabetes non- Whole population Incident diabetes diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

148 LPC 20:2 279 80 271 77 282 82 270 81 265 76 281 88

149 LPC 20:3 2733 749 2802 738 2697 752 2782 740 2770 701 2804 807

150 LPC 20:4 6808 2049 6937 1999 6740 2073 6865 2054 6938 1966 6737 2204

151 LPC 20:5 1130 535 1136 512 1127 547 1142 528 1136 525 1152 537

353 LPC 22:0 16 6 16 6 17 5 16 6 16 6 15 6

152 LPC 22:6 1 1 1 1 1 1 1 1 1 1 1 1

153 . LPC 24:0 4 2 Ά 2 4 2 4 2 4 2 4 2

154 modPC.506.3/3.4 3 1 4 1 3 1 3 1 4 1 3 1

155 modPC.512.3/1.7 61 16 60 15 62 16 60 15 59^' 15 61 15

156 modPC 620.4/2.6 37 22 38 23 37 21 36 21 37 22 33 17

158 modPC.536.3/3.5 48 15 47 15 49 15 46 15 45 15 46 15

159 modPC.538.3/4.1 63 21 61 23 64 20 60 22 60 23 59 19

160 modPC.594.4/3.1 212 134 222 137 207 132 208 126 217 135 191 108

161 modPC.608.4/4.0 39 U 38 11 39 10 37 11 37 12 37 11

162 modPC.610.4/1.7 0 0 0 0 0 0 0 0 0 0 0 0

383 modPC.633.4/4.6 15 5 14 5 15 5 14 4 14 4 14 5

354 modPC.636.4/3 6 151 88 164 95 144 83 152 80 159 87 139 63

165 modPC.645.4/4.4 31 17 31 19 31 16 30 15 29 13 31 17

166 modPC.6504/3 8 710 380 741 404 694 366 685 339 706 365 649 286

82 modPC.650.4/3.9 71 40 73 40 70 40 68 36 70 38 65 32

167 modPC.666.4 2.9 122 70 124 73 120 68 116 62 118 66 113 55

355 modPC 678.4/4.37 250 ' 137 264 151 243 128 240 126 251 135 221 106

107 modPC 678.4/4.9 77 18 77 19 77 18 76 18 76 19 75 18

168 modPC.678.4/5.0 113 74 119 78 110 71 108 68 114 74 .96 55

66 modPC.690.4/4.3 46 29 46 30 46 29 42 26 44 27 39 22

106 modPC.690.4/5.4 1651 371 1601 359 1677 374 1605 367 1598 365 1618 372

173 modPC.692.4/5.1 13 4 13 4 13 4 13 4 13 4 13 4

174 modPC.736.5/5.7 18 6 17 6 19 7 17 6 17 6 17 5

349 modPC.743.5/6.0 1048 262 1006 241 1070 271 1008 240 1005 230 1013 258

177 modPC.773.6/6.5 15392 2751 15102 2796 15543 2719 15043 2804 15024 2803 15077 2821

356 modPC 798.6/5.7 252 97 266 97 244 96 261 95 271 99 243 86

179 modPC.801.6/6.8 14190 2424 14216 2503 14177 2384 14052 2476 14199 2434 13797 2540

180 modPC.818.6/6.6 2226 651 2121 567 2281 684 2155 585 2147 563 2170 625

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev .

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

183 modPC.843.6/7.2 159 38 157 38 161 37 155 38 155 38 155 38

184 modPC.866.6/7.2 26 7 25 7 27 7 25 6 24 7 25 6

185 modPC.877.6/6.0 . 35 26 33 22 36 28 34 25 33 22 37 29

186 modPC 879.1/6.1 14 7 13 6 15 7 13 6 13 6 15 6

357 PE Unk 9b 238 84 239 89 238 81 242 85 246 88 237 79

187 PG 16:0 18:1 85 42 94 45 81 40 93 43 96 46 89 . 38

188 PG 16:1 18:1 18 11 21 12 16 10 20 11 21 12 18 10

189 PG 18:0 18:1 79 47 95 56 71 40 94 55 98 59 , 86 45

190 PG 18:1 18:1 64 33 74 37 58 29 72 36 76 39 64 29

^' 191 PE 32:0 32 13 36 15 30 11 35 14 36 15 33 12

192 PE 32:1 96 72 114 85 86 63 113 82 116 89 108 69

194 PE 34:1 1449 805 1641 892 1349 737 1660 ' 933 1691 957 1606 892

195 PE 34:2 1856 941 2071 1011 1745 884 2039 994 2100 1017 1934 949

358 PE 35:2 115 53 122 57 111 50 121 54 124 ^' 57 116 49

359 PE 35:3 12 18 14 30 12 7 14 28 15 34 12 6

196 PE 36:0 20 5 20 5 20 5 20 . 5 20 5 20 5

197 PE 36:1 1084 587 1293 727 976 465 1245 684 1334 757 1092 502

198 PE 36:2 5051 2462 5772 2862 4678 2137 5614 2784 5875 2959 5162 2401

199 PE 36:3 1416 660 1555 756 1344 592 1525 719 1575 780 1437 592

200 PE 36:4 2289 1052 2570 1137 2143 975 2525 1134 2631 1181 2342 1029

201 PE 36:5 200 123 226 135 186 114 222 130 228 140 211 112

202 PE 38 1 56 19 61 20 53 17 61 20 63 21 58 18

203 PE 38:2 114 51 127 53 108 48 126 54 129 54 121 53

204 PE 38:3 698 414 817 461 637 373 809 464 844 478 749 435

205 PE 38:4 4690 2087 5334 2464 4356 1776 5271 2322 5561 2469 4767 1955

206 PE 38:5 1351 713 1446 813 1303 652 1429 776 1475 843 1351 641

207 PE 38:6 3264 1738 3576 1810 3102 1680 3619 1832 3736 1892 3414 1712

360 PE 39:4 133 45 131 49 134 43 132 50 130 51 135 46

361 PE 40:4 144 100 178 120 126 82 170 112 184 123 147 88

208 PE 40:6 1892 1061 2199 1186 1732 953 2214 1191 2334 1251 2007 1055

209 PE 40:7 306 172 327 186 295 163 335 185 341 197 325 161

362 PE Unk 9a 269 92 266 98 271 89 270 97 270 103 270 85

363 PE 40:5 567 320 639 374 529 281 626 339 653 372 579 268

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

364 APE 36:2a 123 43 120 41 125 43 120 43 120 41 121 46

365 APE 36:2b 53 19 54 19 52 19 55 22 54 19 55 25

366 APE 36:3a 141 59 139 57 143 59 143 59 142 59 144 60

367 APE 36:3b 199 65 199 69 200 63 199 74 201 72 197 77

368 APE 36:4 444 192 445 191 443 193 453 191 462 195 436 185

369 APE 38:4a 194 67 200 71 191 65 200 71 203 72 196 70

370 APE 38:4b 495 204 501 209 491 201 506 202 ^' 511 206 497 197

371 APE 38:5a 754 265 755 265 753 265 767 267 768 268 767 267

372 APE 38:5b 554 202 580 222 541 190 579 210 597 213 548 200

373 APE 40:5 154 54 152 56 155 53 153 55 154 58 152 51

374 APE 40:6 366 111 365 123 366 105 367 118 370 127 361 101

375 lyso PE 16:0 242 80 253 86 236 77 249 83 249 84 249 82

376 lyso PE 18:0 247 82 266 96 237 72 260 89 266 94 250 79

377 lyso PE 18:1 174 64 177 63 1 2 65 173 60 172 61 175 57

378 lyso PE 18:2 328 116 322 105 331 121 317 110 309 100 332 126

210 PI 32:0 171 145 193 160 159 136 187 158 196 169 172 136

21 1 PI 32:1 209 172 254 209 186 145 255 206 258 216 251 186

212 PI 34:0 65 54 69 57 63 52 66 54 70 61 60 38

213 PI 34:1 2259 1130 2500 1329 2134 991 2492 1294 2506 1369 2469 1158

214 PI 36:0 9 6 10 7 8 6 10 7 10 7 9 6

215 PI 36:1 1726 667 1831 792 ' 1671 586 1799 759 1814 818 1772 646

216 PI 36:2 4539 1434 4731 1631 4439 1311 4653 1512 4723 1567 4532 1411

217 PI 36:3 1265 450 1330 516 1232 407 - 1305 472 1312 500 1294 422

218 PI 36:4 1624 655 1797 726 1534 596 1755 687 1798 729 1679 605

219 PI 38:2 201 81 206 81 198 81 207 83 203 83 212 84

220 PI 38:3 2389 792 2503 847 2330 757 2489 805 2489 864 2488 695

221 PI 38:4 10494 2667 11007 2854 10228 2528 10939 , 2764 11141 2941 10588 2402

222 PI 38:5 1018 350 1077 406 988 314 1056 367 1069 403 1033 294

223 PI 38:6 290 129 303 136 283 125 308 141 305 140 313 144

224 PI 40:4 150 57 163 64 144 52 163 63 165 67 160 54

225 PI 40:5 531 ^' 213 577 252 507 186 575 241 587 263 555 198

226 PI 40:6 650 282 680 295 634 275 692 309 697 312 684 307

232 PS 36:1 146 62 154 54 142 66 148 49 153 51 140 45

Incident nan- Pre-diabetes non- Whole population Incident diabetes diabetes Pre-diabetes population Pre-diabetes progressors progressors

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

233 PS 36:2 33 18 35 20 33 16 34 18 36 21 31 12

236 PS 38:3 53 26 56 26 52 25 54 23 - 55 25 51 18

237 PS 38:4 183 95 194 104 177 90 186 92 192 101 175 73

238 PS 38:5 21 13 21 13 20 13 20 12 21 12 19 11

240 PS 40:5 50 31 53 29 49 31 51 26 53 29 48 22

241 PS 40:6 64 36 68 39 61 35 67 37 69 39 64 32

242 COH 1099700 297973 1128200 322189 1084900 283871 1115200 307526 1122000 324701 1103600 276435

243 CE 14:0 38410 22444 45541 26726 34711 18863 44389 25294 46602 27646 40545 20143

244 CE 15:0 27518 11516 29452 12759 26515 10694 28756 11740 29343 12480 27738 10314

245 CE 16:0 599225 150406 632993 154248 581713 145495 628663 154207 634363 150993 618763 159963

246 CE 16:1 472081 341480 581978 415017 415088 280284 573550 390526 597791 423486 531448 323266

247 CE 16:2 18212 10611 21745 12778 16379 8762 21320 12053 22341 13217 19546 9514

248 CE 17:0 16138 5579 16665 6183 15865 5226 16384 5520 16571 5762 16059 5086

249 CE 17:1 39647 19793 44248 21955 37261 18146 43572 20237 44203 21135 42476 18632

250 CE 18:0 24312 7665 25711 8313 23586 7211 25200 7871 25849 8109 24072 7345

251 CE 18:1 1142000 345632 1238100 377511 1092200 317068 1217100 330067 1230800 343293 1193500 306063

252 CE 18:2 3698700 1315230 4126700 1548270 3476800 1115340 4033600 1433280 4192000 1567590 3758500 1119130

253 CE 18:3 465199 258046 556791 321788 417699 202471 529531 276613 560598 312180 475573 190146

254 CE 20:0 495 335 555 378 464 306 529 346 535 352 518 337

255 CE 20:1 846 271 846 269 845 272 833 269 837 268 826 271

256 CE 20:2 2083 581 2190 589 2027 569 2168 578 2178 567 .2151 600

257 CE 20:3 95966 40995 110717 46396 88316 35616 109070 45252 112732 47583 102708 40343

258 CE 20:4 1702800 719181 1957500 840963 1570700 607375 1906100 752554 1995000 814024 1751800 605347

259 CE 20:5 484718 302885 562892 352129 444176 265400 550378 318097 577936 357679 502513 227918

260 CE 22:0 242 182 292 232 216 142 280 222 302 246 241 166

261 CE 22:1 451 229 458 199 447 243 462 236 460 204 465 285

262 CE 22:2 48 19 47 1? 49 19 47 18 46 19^' 47 17

263 CE 22:3 111 56 111 56 110 56 111 58 110 59 113 57

264 CE 22:4 1780 612 1946 638 1693 581 1907 594 1952 584 1828 608

265 CE 22:5 19986 8404 22451 9448 18708 7506 21862 8656 22551 9299 20665 7300

266 CE 22:6 237880 119588 265612 132600 223497 109679 266854 131676 273748 137712 254880 120240

267 CE 24:0 252 187 309 240 ' 222 - 145 294 236 321 260 248 181

268 CE 24:1 686 308 728 321 664 299 713 289 722 289 . 699 291

Mean St Dev Mean StDev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 653 223 223 430 430 260 260 165 165 95 95

269 CE24:2 60 22 59 22 60 22 59 22 59 21 59 22

270 CE 24:3 18 8 17 7 18 8 17 8 17 8 18 10

271 CE 24:4 65 26 66 26 65 26 64 23 64 24 63 21

272 CE 24:5 192 85 203 80 187 87 198 79 199 76 195 83

273 CE 24:6 355 196 392 191 336 197 384 186 388 191 378 177

274 modCE 558.5/7.74 41998 27156 48959 31109 38387 24119 45442 . 28021 47776 30014 41388 23782

275 modCE 588.5/7.94 12921 9864 15220 11680 11729 8550 13534 9672 14387 10113 12052 8707

276 modCE 682.7/8.76 12452 8225 14216 9801 11537 7116 13680 9145 14321 ^■ 10162 12567 6948

277 modCE 790.8/6.57 16111 10378 19243 13902 14487 - 7478 18275 12874 19636 15061 15911 7210

278 DG 14:0 14:0 46 55 61 66 39 46 58 68 64 72 46 59

279 DG 14:0 16:0 570 617 770 761 465 498 728 757 821 826 568 591

280 DG 14:0 18:1 1354 1115 1708 1330 1171 935 1640 1315 1798 1444 1367 1004

281 DG 14:0 18:2 587 499 730 596 513 423 707 600 781 651 579 477

282 DG 14: 1 16:0 127 137 171 175 105 106 160 169 180 189 124 122

283 DG 16:0 16:0 1593 1618 2205 2025 1275 1249 2079 1952 2367 2166 1580 1386

284 DG 16:0 18:0 1183 1039 1582 1291 976 807 1493 1233 1698 1388 1137 794

285 DG 16:0 18:1 9941 6968 12677 8542 8522 5489 12207 8138 13465 9061 10022 5627

286 DG 16:0 18:2 4412 3096 5525 3867 3835 2419 5377 3759 5966 4203 . 4353 2538

287 DG 16:020:0 100 110 122 155 89 76 115 147 132 175 85 69

288 DG 16:020:3 321 242 407 310 276 182 396 293 438 334 322 180

289 DG 16:020:4 639 545 854 673 527 425 807 643 925 721 602 406

290 DG 16:022:5 219 149 268 181 193 122 264 174 290 196 218 115

291 DG 16:022:6 347 319 424 331 306 306 432 338 465 · 359 375 291

292 DG 16:1 18:1 3607 2656 4448 3262 3171 2157 4357 3086 4672 3471 3808 2175

293 DG 18:0 16:1 318 291 424 372 262 220 404 354 454 402 317 223

294 DG 18:0 18:0 202 148 250 182 178 120 236 164 264 189 189 93

295 DG 18:0 18:1 1835 1450 2388 1841 1549 1095 2266 1730 2555 1982 1763 997

296 DG 18:0 18:2 887 675 1119 848 767 527 1076 815 1210 925 842 501

297 DG 18:020:4 277 140 326 171 251 114 312 156 344 175 257 94

298 DG 18:1 18:1 9264 5894 11132 7350 8296 4699 10827 6747 11689 7777 9329 4042

299 DG 18:1 18:2 8538 5732 10081 7355 7738 4479 9882 6909 10757 7968 8362 4135

300 DG 18:1 18:3 1481 1040 1777 1348 1327 797 1709 1265 1877 1465 1418 726

301 DG 18:1 20:0 118 150 144 208 105 107 131 195 151 232 96 94

Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev Mean St Dev

# Lipid 653 -653 223 223 430 430 260 260 165 165 95 95

302 DG 18:1 20:3 651 415 757 516 596 338 756 485 804 550 672 328

303 DG 18:1 20:4 1657 1148 2060 1430 1448 905 1983 1343 2206 , 1523 1597 827

304 DG 18:2 18:2 1333 1083 1568 1419 1212 835 1539 1338 1695 1541 1268 818

305 TG 14:0 16:0 18:2 18100 17591 23364 22386 15369 13751 22116 21302 24660 24584 17696 12835

306 TG 14:0 16:1 18:1 18796 16376 22782 ¹ 20036 16729 13684 21561 18757 23424 21810 18325 11080

307 TG 14:0 16:1 18:2 4065 4133 4917 5027 3623 3508 4683 4786 5194 5577 3796 2755

308 TG 14:0 18:0 18:1 1968 2120 2584 2571 1649 1763 2414 2447 2767 2854 1799 1297

309 TG 14:0 18:2 18:2 1975 1872 2348 2459 1781 1446 2239 2290 2492 2692 1798 1221

310 TG 14:1 16:0 18:1 4679 4717 5963 6170 4013 3580 5603 5799 6230 6778 4515 3262

311 TG 14:1 16:1 18:0 . 16731 17067 21665 22175 14172 13001 20829 20995 22754 24197 17485 13235

312 TG 14:1 18:0 18:2 889 662 1061 862 800 509 1011 797 1105 948 848 369

313 TG 14:1 18:1 18:1 8265 5754 9649 7282 7546 4623 . 9388 6798 10073 7961 8199 3810

314 TG 15:0 18:0 18:1 3741 3703 4789 4939 3197 2712 4523 4692 5119 5529 3489 2371

315 TG 15:0 18:1 16:0 2739 1897 3135 2426 2534 1516 3023 2242 3276 2646 2582 1150

316 TG 16:0 16:0 16:0 9614 11246 13301 13178 7702 9571 12414 12626 14149 14074 9400 8901

317 TG 16:0 16:0 18:0 8904 10764 12435 13185 7073 8735 11583 12675 13453 14378 8335 8076

318 TG 16:0 16:0 18:1 77479 65860· 100747 82448 65411 51451 95587 78440 106832 90173 76056 46468

319 TG 16:0 16:0 18:2 „ 21800 18914 28224 23840 18469 14732 26822 22757 30413 26210 20584 12901

320 TG 16:0 16:1 18:1 80937 56346 98012 70844 72082 44697 94540 66448 101701 77130 82103 39244

321 TG 16:0 18:0 18:1 24643 24623 33256 31072 20177 19060 31197 29529 35768 34144 23259 16374

322 TG 16:0 18:1 18:1 169189 90180 198509 108378 153984 74852 192633 100979 206236 115353 169008 63175

323 TG 16:0 18:1 18:2 82174 . 49128 96026 61528 74990 39467 93228 57563 101619 66361 78654 33291

324 TG 16:0 18:2 18:2 24705 17418 29274 22332 22335 13657 28152 20938 31190 24221 22875 11832

325 TG 16:1 16:1 18:1 1299 1221 1595 1551 1146 976 1563 1512 1684 1713 1354 ' 1054

326 TG 16:1 16:1 16:1 1511 1548 1970 1901 1273 1266 1855 1803 2121 2113 1392 910

327 TG 16:1 16:1 18:0 10504 8128 .12827 10628 9299 6140 12525 10054 13560 11723 ^' 10726 5794

328 TG 16:1 18:1 18:1 18791 11229 21481 14046 17395 9155 21114 12887 22166 15016 19286 7673

329 TG 16:1 18:1 18:2 21679 14051 25092 18169 19909 10956 24377 16843 26329 19754 20988 9082

330 TG 17:0 16:0 16:1 8804 8961 11467 11994 7424 6482 10788 11358 12267 13380 8218 5710

331 TG 17:0 16:0 18:0 448 587 621 747 358 459 575 709 678 829 395 364

332 TG 17:0 18:1 14:0 6529 6227 8161 8237 5682 4659 7705 7822 8689 9223 5996 3937

333 TG 17:0 18:1 16:0 5232 4900 6718 6429 4462 3655 6328 6064 7185 7144 4841 2958

334 TG 17:0 18:1 16:1 13040 9204 15430 11821 11800 7207 14787 10920 16133 12878 12448 5527

Mean St Dev Mean St Dev Mean St Dev Mean StDev Mean St Dev Mean StDev u^' Lipid 653 6S3 223 223 430 430 260 260 165 165 95 95

335 TG 17:0 18:1 18:1 5022 3181 5883 3971 4575 2575 5644 3638 6129 4247 4801 1966

336 TG 17:0 18:2 16:0 8162 6484 10126 - 8338 7143 4988 9645 7766 10709 9101 7797 4031

337 TG 18:0 18:0 18:0 128 195 172 258 106 149 160 245 188 288 111 129

338 TG 18:0 18:0 18:1 1860 2644 2625 3428 1463 2020 2423 3229 2897 3851 1600 1334

339 TG 18:0 18:1 18:1 15412 12819 19709 16582 13184 9638 18630 15481 20986 18273 14538 7154

340 TG 18:0 18:2 18:2 2467 1890 2963 2494 2210 1419 2828 2319 3170 2734 2234 1100

341 TG 18:1 14:0 16:0 27635 24780 34998 30171 23817 20479 32950 28804 36573 33055 26656 17792

342 TG 18:1 18:1 18:1 27191 15860 31682 20484 24862 12217 30410 18474 32696 21814 26440 9223

343 TG 18:1 18:1 18:2 15775 10429 18143 13839 14548 7860 17435 12568 19048 14857 14634 6132

344 TG 18:1 18:1 20:4 3666 1910 4266 2496 3355 1428 4108 2288 4495 2654 3438 1193

345 TG 18:1 18:1 22:6 1004 869 1070 737 970 929 1130 885 1145 774 1106 1054

346 TG 18:1 18:2 18:2 11940 8763 13943 12066 10902 6181 13227 10861 14593 12953 10853 4830

347 TG 18:2 18:2 18:2 1088 1073 1298 1532 979 707 1223 1377 1380 1655 949 569

348 TG 18:2 18:220:4 572 410 690 569 510 277 648 525 737 620 494 225 values expressed as pmol/ml plasma.

TABLE 12. Ranked features in the recursive feature elimination models for prediction of diabetes in the cross sectional and longitudinal studies ·

Cross sectional study RFE Cross sectional study RFE selection of top 64 features

Cross sectional Longitudinal study, RFE selection (validated on the (RFE of longitudinal for feature study, RFE selection selection longitudinal study) election)

Lipid rank AUC = 0.714 AUC = 0.703 AUC = 0.695 AUC = 0.684

1 dhCer 18:1 DG 16:0 20:4 CE 16:1 modCer 614.6

2 CE 24:5 modCer 614.6 CE 20:4 LPC 18:2

3 modPC 536.3 dhCer 22:0 CE 16:2 DG 16:0 18:1

4 PI 32:1 LPC 18:2 CE 24:5 DG 16:0 16:0

5 Cer 22:0 PC 39:7 dhCer l8:0 PE.36.1

6 LPAF 24:1 modPC 610.4 DG 16:0 18:2 PE.38.1

7 CE 20:5 TG 18:2 18:220:4 LPC 15:0 CE 20:4

8 CE 16:2 PC 35:2 LPC 18:2 dhCer 18:0

9 APC 32:0 dhCer 24:0 CE 18:2 DG 18:0 18:1

10 CE 16:1 modCer 886.8 modCer 614.6 DG 16:0 18:2

11 PC 30:2 APC 36:3b dhCer l8:l DG 18:0 20:4

12 dhCer 18:0 PE 38:1 LPC 18:1 Cer l8:0

. 13 DG 18:2 18:2 modCer 910.8 modCE 588.5 Cer 22:0

14 modPC 773.6 APC 34:2b PI 32:1 CE 20:3

15 modPC 538.3 LPAF 22:0 CE 20:5 SM 16:1

16 oddPC 35:4 DG 16:0 18:0 SM 16:1 CE 24:0

TABLE 13. ROC analysis of multivariate models and other risk scores for the prediction of incident diabetes in the longitudinal study

Asymptotic 95%

Std. Asymptotic Confidence Interval

Test Result Variable(s) Area

Error Sig. Lower

Upper Bound Bound

Model based on longitudinal

.700 .694 .706 study (32 lipids)

Model based on cross sectional

.695

study (32 lipids)

Model based on lipids selected

from the cross sectional stud .683 .678 .689

(16 lipids)

AUSDRISK¹ .682 .024 .000 .635 .729

FINRISK² .680 .024 .000 .633 .727

San Antonio³ .739 .023 .000 .695 .784

Framingham

.698 .023 .000 .654 .743 Offspring Study⁴

ARIC⁵ .760 .022 .000 .716 .803

AUSDRISK calculates risk of incident diabetes over 5 years, variables used were age, sex, ethnicity, parental history of diabetes, history of blood glucose, use of antihypertensive medications, smoking, physical activity and waist circumference (Chen et al, 2010 (supra)).

² FINRISK predicts drug treated diabetes over 10 years, variables used were age, body mass index, waist circumference, antihypertensive drug therapy, history of elevated blood glucose level (Lindstrom and Tuomilehto, 2003 (supra)).

³ San Antonio score is the probability of developing diabetes over 7.5 years, variables used were age sex, ethnicity, fasting plasma glucose, systolic blood pressure, HDL-C body mass index, family history (Mann et al., 20l0(supra)).

⁴ Framingham offspring study is a score which is translated to a risk (%) over 8 years, variables used were fasting plasma glucose, body mass index, HDL-C family history of diabetes triglycerides, hypertension (Mann et al., 20l0(supra)).

⁵ ARIC score calculates risk of incident diabetes over 9 years, variables used were age, family history, fasting plasma glucose, systolic blood pressure, waist circumference, height, HDL-C, triglycerides (Mann et al, 2010(supra)).

TABLE 14. Abbreviations

Abbreviation Lipid class

acCer acylceramide

APC alkylphosphatidylcholine

BMP bis(monoacylglycero)phosphate

CE cholesterol ester

Cer ceramide

COH cholesterol

DG diacylglycerol

DHC dihexosylceramide dhCer dihydroceramide

GM3 GM₃ ganglioside

IGT impaired glucose tolerance

IFG impaired fasting glucose

LPAF lysoplatelet activating factor

LPC lysophosphatidylcholine lysoPE lysophosphatidylethanolamine

MHC monohexosylceramide modCE modified cholesterol ester modCer modified ceramide modPC modified phosphatidylcholine

NGT normal glucose tolerance oddPC odd chain phosphatidylcholine

PC phosphatidylcholine

PE phosphatidylethanolamine

PG phosphatidylglycerol

PI phosphatidylinositol

PS phosphatidylserine

SM sphingomyelin

TG ^; triaclyglycerol

THC trihexosylcermide

BIBLIOGRAPHY

Australian Government DoHaA. The Australian Type II diabetes Risk Assessment Tool (AUSDRISK), An Australian Government, State and Teritory health initiative, ht^://vvTvw.health.gov.au/mtemet mairi/publishmg.nsf7Conten

002A31B/$File/Risk_Assessment_Tool.pdf

Cao et al, Cell 134(6): 933-44, 2008

Chen et al, MedicalJoumal of Australia 792(4): 197-202, 2010

Dunstan et al, Diabetes Res Clin Pract 57(2): 119-129, 2002

Fahy et at, J Lipid Res 46(5): 839-61, 2005

Fahy et al, J Lipid Res 50(Suppl): S9-14, 2009

Guyon et al., Machine Learning 46(1): 389-422, 2002

Johannes et al., Bioinformatics 26(17): 2136-2144, 2010

Lindstrom and Tuomilehto, Diabetes Care 26(3): 725-31, 2003

Mann et al, Am J Epidemiol 777(9): 980-8, 2010

Matthews e. a/., .9/afofo/og/tf 53(11): 2431-41, 2010

Murphy et al, Anal Biochem 366( 1 ): 59-70, 2007

Schmidt et al, Diabetes Care 25(8): 2013-8, 2005

Smyth et al., PLoS Genet 4(9): el000192, 2008

Tandy et al, Atherosclerosis 213(1): 142-7, 2010

Claims

CLAIMS:

1. An assay to stratify a test subject as susceptible or non-susceptible with respect to developing type II diabetes, the assay comprising, consisting or consisting essentially of determining in a test subject the levels of at least two lipid analytes selected from the list consisting of:

(i) one or more modified lipid analytes listed in Table 10;

(ii) two or more non-modified lipid analytes listed in Table 10; and/or

wherein the level of an individual lipid analyte listed in Table 10 is different between susceptible control subjects and non-susceptible control subjects and wherein the level of the lipid analytes in the test subject relative to a control identifies the test subject as being susceptible (with a high risk of developing diabetes) or non-susceptible (a low risk of developing diabetes).

2. The assay of claim 1, comprising, consisting or consisting essentially of comparing the level of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject. selected from a susceptible control subject and a non-susceptible control subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the non-susceptible control subject identifies the test subject as being non-susceptible, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the susceptible control subject identifies the test subject as being susceptible.

3. An assay to stratify a test subject as a susceptible or non-susceptible with respect to developing type II diabetes, the assay comprising, consisting or consisting essentially of determining in a test subject the levels of at least 8 to 32 or at least 5 to 40 or at least 10 to 20 or at least 8 to 16 lipid analytes selected from the group consisting of dhCer 18: 1, CE 24:5, modPC 536.3, PI 32: 1, Cer 22:0, LPAF 24: 1, CE 20:5, CE 16:2, APC 32:0, CE 16: 1, PC 30:2, dhCer 18:0, DG 18:2 18:2, modPC 773.6, modPC 538.3, oddPC 35:4., DG 16:0 20:4, modCer 614.6, dhCer 22:0, LPC 18:2, PC 39:7, modPC 610.4, TG 18:2 18:2 20:4, PC 35:2, dhCer 24:0, modCer 886.8, APC 36:3b, PE 38: 1, modCer 910.8, APC 34:2b, LPAF 22:0 and DG 16:0 18:0., CE 16: 1, CE 20:4, CE 16:2, CE 24:5, dhCer 18:0, DG 16:0 18:2, LPC 15:0, LPC 18:2, CE 18:2, modCer 614.6, dhCer 18: 1, LPC 18:1, modCE 588.5, PI 32:1, CE 20:5, SM 16:1., modCer 614.6, LPC 18:2, DG 16:0 18: 1, DG 16:0 16:0, PE.36.1, PE.38.1, CE 20:4, dhCer 18:0, DG 18:0 18: 1, DG 16:0 18:2, DG 18:0 20:4, Cer 18:0, Cer 22:0, CE 20:3, SM 16:1, and CE 24:0 wherein the level of the individual lipid analytes are different between susceptible control subjects and non-susceptible control subjects, and comparing individual levels of the selected lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a susceptible control subject or a non-susceptible control subject, and wherein the level of the lipid analytes in the test subject relative to the control subject identifies the test subject as being susceptible or non-susceptible.

4. The assay of claim 1 or 2 wherein the or each modified lipid analyte in (i) is selected from a modified cholesterol ester (modCE), a modified ceramide (modCER) and a modified phosphatidylcholine (modPC).

5. The assay of claim 1 or 2, wherein the non-modified lipid analytes in (ii) are selected from a ceramide (dhCer), a ceramide (Cer), a dihexosylceramide (DHC), a phosphatidylglycerol (PG), a phosphatidylethanolamine (PE), a phosphatidylinositol (PI), a cholesterol ester (CE), a diacylglycerol (DG) a triacylglycerol (TG), a (LPAF), an alkylphosphatidylcholine (APC), a phosphatidylcholine (PC), a lysophosphatidylethanolamine (lysoPE), and an odd chain phosphatidylcholine (oddPC).

6. The assay of claim 1 or 2, wherein the or each modified lipid analyte in (Hi) is selected from a modified cholesterol ester (modCE), a modified ceramide (modCER) and a modified phosphatidylcholine (modPC) and the or each non-modified lipid in (iii) is selected from a ceramide (dhCer), a ceramide (Cer), a dihexosylceramide (DHC), a phosphatidylglycerol (PG), a phosphatidylethanolamine (PE), a phosphatidylinositol (PI), a cholesterol ester (CE), a diacylglycerol (DG) a triacylglycerol (TG), a (LPAF), an alkylphosphatidylcholine (APC), a phosphatidylcholine (PC), a lysophosphatidylethanolamine (lysoPE), and an odd chain phosphatidylcholine (oddPC).

7. The assay of claim 1 or 2 wherein the at least two lipid analytes are selected from dhCer 18:0, dhCer 22:0, dhCer 24:0, dhCer 24:1, Cer 18:0, Cer 20:0, Cer 22:0, DHC 16:0,modCer 576.5/7.7, modCer 614.6/5.7, modCer 886.8/9.1, modCer 910.8/9.0, PC 33:0, PC 35:2, PC 39:7, PC 44: 12, APC 30:0, APC 36:2, APC 34: 1, APC 34:2b, APC 36:36, LPAF 22:0, LPAF 24:1, LPAF 14:0, LPAF 22: 1., LPAF 16:0, LPC 18:0, modPC.610.4/1.7, PG 16:0 18:1, PG 16: 1 18: 1, PG 18:0 18:1, PG 18: 1 18:1, PE 32:0, PE 32: 1, PE 34: 1, PE 34:2, PE 36:1, PE 36:2, PE 36:3, PE 36:4, PE 36:5, PE 38: 1, PE 38:2, PE 38:3, PE 38:4, PE 38:5, PE 38:6, PE 40:4, PE 40:6, PE 40:5, lyso PE 16:0, lyso PE 18:0, PI 36:4, CE 14:0, CE 16:1, CE 16:2, CE 18:1, CE 18:2, CE 18:3, CE 20:3, CE 20:4, CE 20:5, CE 22:0, CE 22:4, CE 22:5, CE 22:6, CE 24:0, modCE 558.5/7.74, modCE 588.5/7.94, modCE 682.7/8.76, modCE 790.8/6.57, G 14:0 14:0, DG 14:0 16:0, DG 14:0 18: 1, DG 14:0 18:2, DG 14: 1 16:0, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18:1, DG 16:0 18:2, DG 16:0 20:3, DG 16:0 20:4, DG 16:0 22:5, DG 16:0 22:6, DG 16: 1 18: 1, DG 18:0 16:1, DG 18:0 18:0, DG 18:0 18: 1, DG 18:0 18:2, DG 18:0 20:4, DG 18:1 18:1, DG 18:1 18:2, DG 18:1 18:3, DG 18: 1 20:3, DG 18: 1 20:4, TG 14:0 16:0 18:2, TG 14:0 18:0 18:1, TG 14:1 16:0 18:1, TG 14:1 16:1 18:0, TG 15:0 18:1 16:0, TG 16:0 16:0 16:0, TG 16:0 16:0 18:0; TG 16:0 16:0 18:1, TG 16:0 16:0 18:2, TG 16:0 16:1 18:1, TG 16:0 18:0 18: 1, TG 16:0 18:1 18: 1, TG 16:0 18:1 18:2, TG 16:0 18:2 18:2, TG 16: 1 16:1 18:1, TG 16:1 16: 1 18:0, TG 17:0 16:0 16:1, TG 17:0 16:0 18:0, TG 17:0 18: 1 14:0, TG 17:0 18:1 16:0, TG 17:0 18:1 18:1, TG 17:0 18:2 16:0, TG 18:0 18:0 18:0, TG 18:0 18:0 18: 1, TG 18:0 18:1 18: 1, TG 18:0 18:2 18:2, TG 18:1 14:0 16:0, TG 18:1 18: 1 20:4, and TG 18:2 18:2 20:4.

8. The assay of claim 1, comprising, consisting or consisting essentially of determining the individual levels of at least two lipid analytes selected from modCer 614.6/5.7, PC 33:0, PC 39:7, PC 44:12, APC 34:1, APC 34:2b, APC 36:2, APC 36:3b, LPAF 22:0, LPAF 24: 1, PE 40:4, DG 16:0 16:0, DG 16:0 18:0, DG 16:0 18:1, DG 16:0 18:2, DG 16:0 20:4, DG 18:0 16: 1, DG 18:0 18: 1, TG 16:0 16:0 18:1, TG 16:0 18:0 18:1, OddPC 35.2, OddPC 37.3 and OddPC 37.4. (Figure 10A)

9 The assay of claim 1, comprising, consisting or consisting essentially of determining the individual levels of at least two lipid analytes selected from PC 35:3, PC 35:4, PC 37:5, PC 38:4, APC 36:3a, APC 36:4, APC 38:4, APC 38:5, DG 16:0 18:0, DG 16:0 18:2, DG 16:0 20:0, DG 16:0 20:4, DG 16:0 22:5, DG 18:0 18:2, DG 18:0 20:4, DG 18: 1 20:4, TG 16:0 16:0 18:2, TG 18:0 18:2 18:2, TG 18: 1 18: 1 20:4, and TG 18:2 18:2 20:4. (Figure 10B)

10. An assay to stratify a test subject with respect to diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from dhCer 16:0, DHC 24: 1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.608.4/4.0, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 wherein the individual levels of the lipid analytes are different between diabetic and pre-diabetic control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a diabetic control subject and a NGT control subject, and wherein the level of the lipid analytes in the test subject relative to a control indicates that the subject is diabetic and not pre-diabetic. (Figure 7B)

11. An assay to stratify a test subject with respect to diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from modCer 883.8/7.8, APC 34:1, COH, TG 14:0 16:0 18:2, TG 14:0 16:1 18:1, TG 14:0 16:1 18:2, TG 14:0 18:2 18:2, TG 14:1 16:0 18:1, TG 14:1 16:1 18:0, TG 14:1 18:0 18:2, TG 14:1 18:1 18:1, TG 15:0 18:1 18:1, TG 16:0 16:1 18:1, TG 16:1 16:1 16:1, TG 16:1 16:1 18:0, TG 16:1 16:1 18:1, TG 16:1 18:1 18:1, TG 16:1 18:1 18:2, TG 17:0 18:1 16:1, and TG 17:0 18:2 16:0 wherein the individual levels of the lipid analytes are different between diabetic control and pre-diabetic control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic control subject or an NGT control subject, and wherein the levels of the individual lipid analytes in the test subject relative to a control provides an indication that the subject is pre-diabetic and not diabetic. (Figure 7A)

12. An assay to stratify a test subject with respect to obesity or diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from THC 18:0, THC 24:0, THC 24:1, GM3 24:0, GM3 24:1, PC 32:2, PC 36:3, PC 39:7 PC 40:7, APC 32:1, APC 38:6, LPC 20:0, LPC 20:1, LPC 22:6, modPC.843.6/7.2, modPC.866.6/7.2, modPC.877.6/6.0, PE 32:2, CE 22:2, and TG 17:0 1_.8:1 14:0 wherein the individual levels of the lipid analytes are different between obese control subjects and diabetic control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from an obese control subject or a non-obese control subject, and wherein the levels of the lipid analytes in the test subject relative to a control provides an indication that the test subject is obese independent of diabetes. (Figure 5B)

13. An assay to stratify a test subject with respect to obesity or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from GM3 24:0, GM3 24:1, PC 39:7, PC 40:7, APC 32:1, APC 38:6, LPAF 18:1, LPC 18:1, LPC 20:0, LPC 20:1, LPC 20:1, LPC 20:2, LPC 22:0, LPC 22:6, modPC.636.4/3.6, modPC.664.4/4.3, modPC.843.6/7.2, modPC.877.6/6.0, modPC 879.1/6.1, and PS 36:1 wherein the individual levels of the lipid analytes are different between obese control subjects and pre-diabetic control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from with an obese control subject or non-obese control subject, and wherein the individual level of the lipid analytes in the test subject relative to a control provides an indication that. the subject is obese independent of pre-diabetes. (Figure 6B)

14. An assay to stratify a test subject with respect to obesity or diabetes, the assay comprising, consisting or consisting essentially of determining in the test subject the levels of at least two lipid analytes in Table 6 selected from modCer 921.8/9.1, oddPC 35:2, oddPC 37:3, oddPC 37:4, PC 40:6, modPC.633.4/4.6, modPC.773.6/6.5, modPC 827.7/6.8, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:6, PS 40:6, CE 22:6, CE 24:5, CE 24:6, modCE 790.8/6.6, DG 16:0 22:6, and TG 18:1 18: 1 22:6 wherein the individual levels of the lipid analytes are different between diabetic control subjects and obese control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a NGT control subject or diabetic control subject, and wherein the levels of the lipid analytes in the test subject relative to a control provides an indication that the test subject is diabetic independent of obesity. (Figure 5A)

15. An assay to stratify a test subject with respect to obesity or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from modCer 576.5/7.7, modCer 883.8/7.8, modCer 921.8/9.1, PC 40:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, DG 16:0 22:5, DG 16:0 22:6, DG 18:0 18: 1, DG 18: 1 18:1, TG 14:1 18:0 18:2, TG 16: 1 16: 1 18: 1, TG 16: 1 18:1 18: 1, TG 16:1 18:1 18:2, TG 18: 1 18:1 18:1, and TG 18:1 18: 1 22:6 wherein the individual levels of the lipid analytes are different between obese control subjects and pre-diabetic control subjects and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic control subject or NGT control subject, and wherein the levels of the individual lipid analytes in the test subject relative to a control identifies the test subject as pre-diabetic independent of obesity. (Figure 6A) - I l l -

16. An assay to stratify a test subject with respect to obesity or diabetes or pre-diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two polyunsaturated lipid analytes selected from PC 38:6, PC 40:5, PC 40:6, PC 40:7, PC 44:12, APC 38:6, LPC 22:6, PE 38:6, PE 40:6, PE 40:7, PI 38:6, PI 40:5, PI 40:6, PS 40:6, CE 22:5, CE 22:6, DG 16:0 22:5, DG 16:0 22:6, and TG 18: 1 18:1 22:6 wherein the individual levels of the lipid analytes are different between obese control subjects and diabetic control subjects or pre-diabetic control subjects, and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a pre-diabetic control subject or diabetic control subject or NGT control subject, and wherein the level of the lipid analytes in the test subject relative to a control provides an indication that the test subject is diabetic or pre-diabetic independent of obesity. (Figure 8)

17. An assay to stratify a test subject as normal (NGT) with respect to diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from dhCer 16:0, DHC 24:1, modCer 614.6/5.7, LPAF 22:0, LPAF 24:2, LPC 15:0, LPC 18:2, mod PC.536.3/3.5, mod PC.538.3/4.1, mod PC.608.4/4.0, mod PC.633.4/4.6, mod PC.773.6/6.5, mod PC 827.7/6.8, PS 40:6, CE 16:2, CE 20:4, CE 20:5, CE 24:5, modCE 588.5/7.9, and modCE 682.7/8.8 and comparing the individual levels of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a normal control subject and a diabetic control subject or pre-diabetic control subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the diabetic control subject or pre-diabetic control subject identifies the test subject having diabetes or pre-diabetes, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the normal (NGT) subject identifies the test subject as a normal (NGT) subject with respect to diabetes.

18. An assay to stratify a test subject as normal (NGT) with respect to diabetes, the assay comprising, consisting or consisting essentially of determining the levels of at least two lipid analytes in Table 6 selected from modCer 883.8/7.8, APC 34:1, TG 14:0 16:0 18:2, TG 14:0 16:1 18:1, TG 14:0 16: 1 18:2, TG 14:0 18:2 18:2, TG 14: 1 16:0 18: 1, TG 14:1 16:1 18:0, TG 14:1 18:0 18:2, TG 14: 1 18:1 18:1, TG 15:0 18: 1 18:1, TG 16:0 16: 1 18:1, TG 16:1 16: 1 16: 1, TG 16:1 16: 1 18:0, TG 16: 1 16: 1 18:1, TG 16: 1 18:1 18:1, TG 16:1 18:1 18:2, TG 17:0 18:1 16:1, and TG 17:0 18:2 16:0 and comparing the level of the at least two lipid analytes in the test subject to the respective levels of the same lipid analytes in at least one control subject selected from a normal control subject and a diabetic or pre- diabetic control subject, wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the diabetic control subject or pre-diabetic control subject identifies the test subject having diabetes or pre-diabetes, and wherein a similarity in the respective levels of the at least two lipid analytes between the test subject and the normal control subject identifies the test subject as a normal (NGT) subject with respect to diabetes.