CA2217264A1 - Calcium supplements and calcium containing beverages comprising vitamin d - Google Patents

Abstract

The present invention relates to calcium supplements and, in particular, to a solid supplement fortified with calcium glycerophosphate, vitamin D, vitamin C, vegetable oil and a gum selected from the group consisting of gum arabic, gum tragacanth and xanthan gum; to a beverage concentrate or additive (liquid or powder) containing calcium, vitamin D, a vegetable oil and a gum; and to a beverage made by reconstituting such beverage concentrates and additives to make a liquid nutritional product fortified with both calcium and vitamin D, and preferably having a low pH. Concentrates may be diluted with water or juices.

Description

W O 96/31130 PCT/US9''01C01 CALCIUM SUPPLEMENTS AND CALCIUM CONTAINING BEVERAGES COMPRISING VITAMIN D

FIELD OF THE INVENTIQN

The p,t:senl invention relates to calcium supplements and, in particular, to a solid sl~FF'Pmer,l fortified with calcium glycerophosphdle, vitamin D and vitamin C; to a beverage concenl, ~te or additive (liquid or powder) containing calcium and vitamin D;
and to a beverage made by reconstituting such beverage concer,l,ales and additives to make a liquid nul,iliol1al product forliried with both calcium and vitamin D, and p,~r~rably having a low pH.

BACKGROUND OF THE INVENTION

Calcium is an esse"lial nutrient; it is a major co",ponent of mineralized tissues and is required for normal growth and dcvelopn,ent of the skeleton and teeth. Over the last decade calcium has enjoyed increased attention due to its potential role in the prevention of osteoporusis. Osteoporosis affects more than 25 million people in the United States and is the major underlying cause of bone fractures in postmenopausal women and the elderly. "Optimal Calcium Intake", JOURNAL OF THEAMERICAN
MEDICAL ASSOCIATION, 272(24): 1942-1948 (1994).
As used herein "osteoporosis" refers to a reduction in the amount of bone mass.
Two important factors influencing the occurrence of osteopol usis are optimal peak bone mass attained in the first two to three clec~des of life and the rate at which bone mass is lost in later years. Adequate calcium intake is critical to achieving optimal peak bone mass and modifies the rate of bone mass loss associated with aging. Wardlaw, "Putting osteoporosis in perspe~;ti~/e", JOURNAL OF THE AMERICAN DIETETIC
ASSOCMTION, 93(9): 1000-1006 (1993).
Several cofactors modify calcium balance and influence bone mass. These include dietary constituents, hormones, drugs, and the level of physical activity. Unique host characteristics may also modify the effects of dietary calcium on bone health.
These include the individual's age and ethnic and genetic background, the presence of gastrointestinal disorders such as malabsorption and the post~a~ ctomy sylldlullle~

W O96/31130 PCTAUS9~1C01 and the plesence of liver and renal rlice~se. Interactions among these diverse cur~-ilur~
may affect calcium balance in either a positive or negative manner and thus alter the optimal levels of calcium intake. "Optimal Calcium Intake", JOURNAL OF THE
AMERICAN MEDICAL ASSOCIATION, 272(24): 1942-1948 (1994).
Calcium requirements vary throughout an individual's lifetime with greater needsoccurring during the period of rapid growth in childhood and adc'os~ence, pr~y"ancy and lactation, and in later adult life. Table 1 pr~senl:~ the optimal calcium requirements which were ~e~ ' shed at a National Institute of Health (NIH) conrelence on optimal calcium intake held June 6-8,1994. "Optimal Calcium Intake", JOURNAL OF THE
AMERICAN MEDICAL ASSOCIATION, 272(24): 1942-1948, at 1943 (1994). The pa~ anl:, at the NIH co"r~rt:"ce considert:d former Recommended Dietary Allowances (RDA) (10th edition,1989) for calcium intake as lererence levels and used them as guideli"es to determine optimal calcium intake in light of new data on calcium-related disorders.

TABLE 1: OPTIMAL CALCIUM INTAKES
OPTIMAL DAILY INTAKE
GROUP (in mg of calcium) Infants Birth-6 months 400 6 months-1 year 600 Children 1-5 years 800 6-10 years 800-1,200 A~lolescenls/Young Adults 11-24 years 1,200-1,500 Men 25-65 years 1,000 Over 65 years 1,500 Women 25-50 years 1,000 Over 50 years (postmenopausal) On esllogens 1,000 Not on estrogens1,500 Over 65 1,500 P, ~yl lal ll and nursing1,200-1,500 CA 022l7264 l997-l0-02 W O96/31130 PCT~U~9"01C01 National consumption data indicate most females over the age of eleven, as well as elderly men, consume amounts of calcium below recommended levels. "NationwideFood Consumption Survey, Continuing Survey of Food Intakes of Individuals", USDANFCS, CFS 11 Report No. 86-93 (1988). Accoldil ,g to the Second National Health and Nutrition Exdr, li. ,alion Survey, the median daily calcium intake for women in the United States was 574 mg. DIETARY INTAKE SOURCE DATA: UNITED STATES, 1976-80, Data From the National Health Survey, Series ll, No. 231, DHHS Pl~lic;31ion No. (PHS), pages 83-1681 (1983).
The preferred approach to attaining optimal calcium intake is through dietary sources. Dairy products are the major contributors of dietary calcium because of their high calcium content (e.g. approki"~alely 250-300 mg/8 oz of cow's milk) and frequency of consumption. As used herein the term "milk" is unde~lood to refer to cow's milk, and the term "dairy products" is understood to refer to food products derived from cow's milk. However, many pe, ~ons, especially women, prefer to limit their intake of dairy products for several r~asOn5: (a) they dislike the taste of milk/milk products; and/or (b) they have a lactose i, ~ lerdnce; and/or (c) they per~eive that some dairy products are too high in fat or protein and may lead to weight gain. Other good food sources of calcium include some green vegetables (e.g. bloccoli, kale, turnip greens, Chinese c~hb~ge), calcium-set tofu, some legumes, canned fish, seeds and nuts. Breads and cereals, while relatively low in calcium, contribute siyl ,iricar,lly to calcium intake because of their frequency of consumption. A number of calcium-fortified food products are currently avaiiable, including fortified juices, fruit drinks, breads and cereals.
Consumption of these foods may be an addilional strategy by persons to achieve their optimal calcium intake.
To maxi",i~e calcium absorption, food selection decisions should include consideration of information on the bioavailabil;ty of the calcium contained in the food.
Bioavailability (absorption) of calcium from food depends on the food's total calcium content and the presence of components which enhance or inhibit calcium absorption.
Bioavailability of minerals in food has been traditionally tested by the balance method, which estimates absorption from the difference between ingested intake and fecaloutput. This approach works well for many nutrients where the difference betweenintake and excretion is large, but is less well suited for an element such as calcium W O96/31130 PCT/US~ ;C01 er,leri- ,g the digestive tract with its sec, e :lions. A decline in fractional absorption from 30% to 20% could have profound nul, ilional significance but would be difficult to detect using the balance method. In conl,dsl, is~t-Fic methods e~li",dle absor~.lion directly from the appearance of the ingested tracer in body fluids. Future clinical evaluations of the bioavailabil;ly of calcium from the liquid nul, ilional product of the presenl invention will use a state-of-the-art isotope tracer method.
Not all calcium salts are created equally. Calcium salts range from 9% elementalcalcium in calcium gluconate to 40% calcium in calcium carbol,dle. Bioavailability depends on solubility. A new calcium delivery system, Calcium Citrate Malate (CCM) claims to be approximately six-times the solubility of either calcium citrate or calcium malate, both of which are themselves suL ~sldnlially more soluble than calcium Cdl Lonale. Smith et al., "Calcium Absorption from a New Calcium Delivery System(CCM)" CALCIFIED TISSUE INTERNATIONAL, 41 :351-352 (1987) relates an expe, i" ,enl in humans wherein calcium from CCM was absorL,ed s4"iricar,l1y better than from either calcium carbonate or milk. 38.3% vs 29.6% and 29.4% respectively. WO91/19692 .I;,closes a process for making a met~ ' ~'e calcium citrate malate.
However, the United States Food and Drug Adllli, l;~ lion (FDA) has advised that, in order for calcium-conl~i. l;. ,9 food i"y~ ed,e. IL~ in convenlional foods or calcium sl~, ~len,ent products to be considered eligible to bear the authorized calcium/osteoporosis health claim, they must meet the requi,~l"e"l~ in 101.14, which include that they have been shown to the FDA's satisfaction to be safe and lawful under the ~pplic- ~'e safety provisions of the act (56 FR at 60699). Safety and lawfulness can be demon:jLI dted in a number of ways, including through a showing that a food is generally recognized as a safe (GRAS), affirmed as GRAS by the FDA, listed in the food additive regulations, or subject to a prior sanction. Of the 36 or more calcium-containing ingredients idenliried by the agency as currently in use the FDA advised that only the following 10 compounds had been demonsl, ~led to be safe and lawful for use in adietary supplement or as a nutrient supplement: calcium Cdl bonale, calcium citrate, calcium glycerophosphate, calcium oxide, calcium pantothenate, calcium phosphate, calcium pyrophosphate, calcium chloride, calcium lactate, and calcium sulfate (56 FR at 60691).
Table 2 summarizes the enhancement and i"hil,ilion factors associated with calcium absorption.

W O96131130 PCTrUS9''01C01 FACTORS WHICH ENHANCE OR INHIBIT CALCIUM ABSORPTION
Inhibitors Enl.a.. cer~
Older age (~ 51) Younger age (11-24) Vitamin D deric ncy Healthy vitamin D levels Oxalic acid, fiber & phytates (only if Pregnancy & l~ct~tion achlorhydria present) Esl,ogen (natural & ,~:place",er,l therapy) Caffeine ~ Adequate protein intake Presence of other nutrients in Ca+2 s~FF'e."ent Ca+2: P04 ratio of 1:1 Excess protein intake ~ 2 X RDA Specific disaccha, ides: fructose &
lactose Specific organic acids:
Citric Malic Ascorbic Calcium absor,ulion is directly affected by an individual's vitamin D status.
Vitamin D der: ~nl individuals absorb less calcium than individuals whose vitamin D
stores are adequate. Vitamin D m~GI_b~" . ~ enhance calcium absor~ lion. The major metabolite 1,25-dihydroxyvitamin D, stimulates active transport of calcium in the small intestine and colon. Deri.,;en~y of 1,25-dihydroxyvitamin D, caused by inadequate dietary vitamin D, in~dequ~te exposure to sunlight, impaired activation of vitamin D, or acquired ~si~lance to vitamin D, results in reduced calcium absor,ulion. In the absence of 1,25-dihydroxyvitamin D, less than 10 per~enl of dietary calcium may be absorbed.
Vitamin D deficiency is associated with an increased risk of fractures. Elderly palienl~
are at particular risk for vitamin D d~:ri~,;en.,y because of insuffficient vitamin D intake from their diet, impaired renal s~,llhesis of 1 ,25-dihydroxyvitamin D, and inadequate sunlight exposure, which is normally the major stimulus for endogenous vitamin DS~l ILI ,esis. This is especially evident in homebound or institutionalized individuals.
Supplementation of vitamin D intake to provide 600-800 lU/day has been shown to improve calcium balance and reduce fracture risk in these individuals. Sufficient vitamin CA 022l7264 l997-l0-02 W O96/31130 PCTrUS96/O~Cnl D intake should be ensured for all individuals especi~"y the elderly who are at greater risk for dcvelopment of a deficiency. Sources of vitamin D b~sides supFl-rller)la include sunlight vitamin D-fortified liquid dairy products cod liver oil and fatty fish. Calcium and vitamin D need not be taken together to be effective. FYcessive doses of vitamin D may introduce risks such as h~l en -' llria and hyperc-'o :nia and should be avoided.
Anticonvulsant me- l ~lions may alter both vitamin D and bone mineral met~ho'i~ "
particularly in certain d;sorder:i in the institutionalized and in the elderly. Although s~",pLu"ldlic skeletal .li;eAse is uncommon in noninstitutionalized s~lLi"y:, optimal calcium intake is advised for persons using a"licon~/ulsants. "Optimal Calcium Intake"
JOURNAL OF THEAMERICAN M~EDICAL ASSOClATlON, 272(24): 1942-1948 (1994).
A number of other dietary factors can also affect calcium absorption. Dietary fiber and phytate have been i",plic~led as illhiLililly substances. The binding of calcium by dietary fiber incl eases with increasing pH. The onset of prec;~ ion of calcium phytates occurs in the pH 4-6 range as in achlorl,ydria. At low gastric pH values (2-3) phytate does not bind calcium and calcium binding by dietary fiber would be weak if at all. Thus in normal individuals calcium would reach i"Lesli"al sites as soluble spe~ s.
Dependi"y on the concer,ll ~lions and binding ~ r,ylhs of various food ligands some of the calcium will be absorbed at the i"l~ al sites while the remainder becomes bound as insoluble fiber and phytate cor, F'~xes. Challlpaglle "Low Gastric Hyd,ùcl,'oric Acid Secretion and Mineral Bioavailability" ADVANCES IN EXPERIMENTAL MEDICINE
AND BIOLOGY, 249:
173-184 (1989).
Simple sugars and organic acids also have an effect on bioavailability. Fructosein orange juice and apple juice prur"ùled positive calcium bioavailability from Calcium Citrate Malate (CCM) which is a combination of CaC03 citric acid malic acid: 5:1:1 mol/mol/mol). The lactose in milk forms a soluble compound with calcium. Organicacids such as citric acid malic acid and ascorbic acid may also play a role in the favorable absorption of calcium from CCM. Mehansho et al. "Calcium Bioavailabilily and Iron-Calcium Interaction in Orange Juice" JOURNAL OF THEAMERICAN COLLEGE
OFNUTRITION, 8(1):61-68 (1989).
In addition it is known that high protein intakes specifically of sulfur conlai"i"y amino acids increase urinary calcium excretion. Sulfuric acid radicals are believed to decrease renal tubular resorption. However consumption of high phosphorLJs foods W O96/31130 PCTrUS9"01G01 such as meat, can di~ ish this effect. Spencer et al., "Do Protein and Pl .osphorous Cause Calcium Loss?", JOURNAL OF NUTRI TION, 1 18:657-660 (1988) .
For some individuals, calcium supplements may be the preferred way to obtain optimal calcium intake. Although calcium s~p~!e "ents are available in many salts, calcium calbon~le is usually recon""ended because it cGnl~i.,s more elemental calcium per gram than any of the other salts. The di~il IL~yl dlion and ~issc'ution ul ,ara~ri:,lics of commercial calcium car ondle pleparalions, which vary widely, may produce important differences in calcium absor,~lion. Other problems with using large amounts of calcium carbonate is that it can lead to con~lic,~lion and abdon,i"al distention. When problems arise, calcium lactate or calcium citrate are advised. These sl ~hstitl'tions for calcium carbonate are also i,-cl c~l~d for people with achlorhydria. A popular commercially available calcium s~ !e.--ent is TUMS 500TM which is distributed bySmithKline Beecham, Pittsburgh, Pennsylvania, U.S.A. and is labeled as providing 500 mg of ele."enldl calcium (from calcium carbonale pertablet). I loJJ_vcr, the TUMS
500TM label does not illdicdle that this calcium sul,plen,el,l conldi"s any vitamin D.
U.S. 4,786,510 and U.S. 4,992,282 c lisc,lose the use of calcium citrate malate in a beverage or dietary sucrF!e.,,ent ~lliried with iron, but do not di,c,lose the addition of vitamin D to such a product. WO 92/19251 and WO 92/21355 ~lisulose the use of calcium citrate malate in a low pH beverage, and suggests that vitamin D be added to such a beverage along with oil flavors or v ~igl ,i. ,g oil. However; neither WO 92/19251 or WO 92/21355 disclc!se any other details about how to incorporate vitamin D3 into such a beverage.
EP 0 486 425 A2 discloses a liquid oral nutritional formulation which contains carbohydrates, protein, fat, fiber, calcium, and vitamin D, and has a pH of about 3.5 to 3.9. However, this p~ Ihlic~tion teaches that high amounts of micronutrients such as calcium or magnesium may impair the palatability of the product, and should contain the recommended daily allowance of these nutrients in about one liter or product. In an example in the patent publication this product contains only about 570 mg of calcium per liter and about 211 IU of vitamin D per liter. A commercially available product in accordance with this patent publication is distributed by Sandoz Nutrition under the trade name CITRISOURCE~ and is labeled as providing 570 mg of calcium and 210 IU of vitamin D per liter. By way of comparison, prototypes of a beverage according to the present invention contain about 1,408 mg of calcium per liter and about 338 IU of W O96/31130 PCTAUS~ 1C01 vitamin D3 per liter.
U.S. 4,737,375 teaches beverage concenl,dles and beverages having a pH of 2.5 to 6.5, preferably 3.0 to 4.5, which cor,lai"s calcium. The use of vitamin D3 in this beverage is not ~~isclosed. This patent does not teach the use of calcium glycerophosphale (which is used in prer~r,ed embodiments of the prt:senl invention, as a calcium source. The ~cidu'~nts used in this prior art beverage are chosen frommixtures of citric acid, malic acid and phosphoric acid, and the weight ratio of total acids to calcium is in the range of 4 to 7. The calcium level is 0.06% to 0.15%, pl~ bly 0.10% to 0.15% of the beverage, by weight. By way of cor"pari~on, prototypes of the beverage of the present invention have a weight ratio of total acids to calcium of about 5.1.
Two commercially available beverages which are labeled as being p, uteuled by U.S. 7,737,375 are: (1) Sunny Delight~ With Calcium which is distributed by Procter &
Gamble, Cillc;.lndli, Ohio 45202 U.S.A.; and (2) HAWAIIAN PUNCH~, DOUBLE C
which is distributed by Sundor Brands, Inc., Ci"c;l "-ali, Ohio 45202 U.S.A. Acco,d;. ,g to the "Nutrition Facts" on the labels of these cor"",er-,;ally available products: (a) either product contains vitamin D; (b) neither product conlai. ,s any fat; (c) a 240 mL (8 fluid ounce) serving of Sunny Delight~ With Calcium provides 30% of the recommended daily intake of calcium; (d) a 240 mL (8 fluid ounce) serving of HAWAIIAN PUNCH~, DOUBLE C provides 15% of the recommended daily intake of calcium; and (e) and a 240 mL (8 fluid ounce) serving of each of these products provides 100% of the recommended daily intake of vitamin C. Per the product labels, these percent daily values are based on a 2,000 calorie diet. A review of the ingredient listings on the labels of each of these products i".licdles that both of these beverages are aqueous solutions, and that neither product conldi"s gum arabic. Samples of each of these products were tested regarding their pH values: the pH value of the HAWAIIAN PUNCH~E3) DOUBLE C
was 3.91; and the pH value of the Sunny Delight~ With Calcium was 4.05.
GB 2 196 253 A discloses a beverage conld;l l;l l~ calcium and vitamin D. A
water soluble non-toxic calcium salt is used in a quantity sufficient to provide in the final beverage a calcium ion content of from 1.0 x 1 o-2 to 40 x 1 o-2% wlw. The beverage may contain up to 5 x 106 w/w of vitamin D. However, this published patent application does not teach the use of a gum, su~ch as gum arabic or gum tragacanth, in such a beverage to improve vitamin D3 stability.

W O96/31130 PCTrUS~6/01C01 The NIH Consensus Statement recommended that the private sector play an active role in promoting optimal calcium intake by dcvel~p:.,y and l"arl~eli"g a wide variety of calcium-rich foods to meet the needs and tastes of a mulli~ll,ni~ population.
"Optimal Calcium Intake", IQURNAL OF THE AMERICAN MEDICAL ASSOCIATION.
272(24):1942-1948 (1994). Hence, there is provided in acco,dance with one aspect of the present invention a low pH beverage fortified with calcium and vitamin D3 There is provided in accor lance with anolher aspect of the invention a liquid beverage concer,l, dle fG, liried with calcium and vitamin D3. There is provided in accordaoce with yet another aspect of the invention a liquid beverage additive fortified with calcium and vitamin D3.

SUMMARY OF THE INVENTION
Thus, in a first aspect, the invention cGr"priaes a beverage concer,l,al~
c~" ,~,riai"y.
a) a source of calcium; b) vitamin D; c) a vegei ''e oil; and d) a gum.
This concer,l, ~le may be in dry powdered form or it may be in liquid form, in which case it further co",priàesa quantity of an aqueous solution, usually water or juice. Either concentrated form can be reconstitute~/diluted with an aqueous solution to form the desired, final liquid beverage, which forms a second aspect of the invention. Suitable solutions include water, fruit juices and ve~el ''o juices, among others.
The source of calcium prt:~r~bly is calcium glycel uphospl)ale, but may also be calcium citrate malate or calcium carbonate or another food grade calcium salt. The gum may pr~r~rably be selected from gum arabic, gum tragacanth and xa"ll,an gum;whereas the vegelN~le oil may p,~ferdbly be selected from corn oil and partiallyhydrogenated soybean oil.
Rega,.lless of form (dry concenl~le, liquid concelllldle or liquid beverage) thecompositions may further contain supplemental ingredients, such as vitamin C, lactic acid, an ~cid~ nt, a sweetener, a glucose polymer, potassium benzoate or a flavoring agent. Iff desired, the beverage may be carbonated.
In another aspect, the invention provides a calcium supplement in solid form con,pli~i"g calcium glycerophosphate, vitamin D and vitamin C. Preferably, the calcium supplement in solid form comprises calcium glycerophosphaLe, vitamin D3, vegetable oil, vitamin C, and a gum selected from the group consisli"g of gum arabic, gum tragacanth W O96/31130 PCTrU~96/OlC01 and xa~lll,an gum.

BRIEF DESCRIPTION OF THE DRAWINGS
Figs. 1-7 are repr~senl~ e of the methodology used in determining vitamin D3 levels; and Figs. 8-11 are representative of the methodology used in dt:l~nllillilly vitamin C
levels.

DETAILED DESCRIPTION OF THE INVENTION

The levels, half lives and other chara~l~ri~lics and prupe, lies of vitamin D3, calcium and vitamin C It~ d to herein and in the claims were determined, and in the intel~,reL~lion of the claims are to be del~rlllilled, accoldi,lg to the methods set forth in the Appendix A alldched to and made part of this speciric~lion.

SELECTION OF INGREDIENTS USED IN PRACTICING THE INVENTION

The present invention provides high levels of calcium and vitamin D in a carbonated beverage, a noncarbol1ated beverage, a liquid beverage concerllldle, a powdered beverage concerlll dle, a powdered beverage additive, beverages containing a powdered beverage concerll, ~l~ or additive, or a calcium sl ~ 'ament. As used herein and in the claims the terms "liquid nul,ilional product" and "beverage" are under:,lood to be synonymous. As used herein and in the claims a "low pH beverage" is ulldel~lood to refer to a beverage having a pH of less than 4.6. Trial batches of low calorie lemon lime, orange, peach, and wild cherry flavored prototype carbonated beverages have beenmanufactured in accordance with the prt:senl invention. The prototype beverages were manufactured by preparing a beverage concer,l, ~le, then blending the beverage concenl,~le with treated water. The blends where then carbonated and filled intoslandard 12 ounce soda aluminum cans. (Soda aluminum cans are coated in accordance with accepted industry standards to substantially reduce migration of aluminum into the conl~nl~ of the can.) Calcium Source. As used herein and in the claims the term "calcium" used alone refers to elemental calcium, the term "calcium salt" refers to a chemical composition containing elemental calcium, and "calcium source" refers to calcium and/or a calcium salt. The calcium salt used in pr~r~"ed embodi",elll~ of the pr~:sel)l invention is Calcium Glyceruphosph~Le (CaGP) which is generally recognized as safe (GRAS) by the United States Food and Drug Ad"~i"i. l~lion (FDA) (21 CFR 170.3). Another reason for selecting CaGP is that, as already ~ clQsed above in the background section, it is one of the ten calcium compounds ,~c,oy"i~ed by FDA as safe and lawful for use in a dietary sl~FFl_n,ent or as a nutrient supplement for osteoporosis. However; any other sUjtr~le calcium source, such as calcium citrate malate that would be soluble at a pH of about 3.5-4.5 could be employed in the practice of the p,~serll invention.

Calcium glycerophosphdle (CaGP) can be described as a white, odorless, almost ~sl~ ss powder. Its scl ~' ' 'y in water increases in the presence of citric and lactic acids, as stated in the Merck Index. The CaGP used in the trial bdlcl)es was FCC lll grade and was produced by Dr. Paul Lohman GmbH, Emmerthal, Germany and is distributed by Gailard Scl~ ,ger Industries, Inc., Carle Place, New York,11514, USA.
Another reason for selecting CaGP is its excellenl calcium bioavailability. Churella et al., "REL ATIVE CALCIUM (CA) BIOAVAILABILITY OF CA SALTS USED IN INFANT
FORMUL AS", THE FASEB JOURNAL, 4(3):A788 (1990) reports a study which determined the calcium bioav '-~:' 'y of four calcium salts. Rats were fed various diets containing dirr~,~r,l calcium salts for three weeks. At the end of the study, the right femur was removed and tested for calcium. As co",pared to a control, the relative calcium bioavailability was as follows: Il ;~a'c ~m phosphdle 110%, calcium citrate 110% and CaGP 106%. Furthermore, studies r~ ol Led by Hanning et al, "Efficacy of calciumglycerophosphaLe vs conventional mineral salts for total palt:r,ler~l nutrition in low-birth-weight infants: a rarldo",i~ed clinical trial'~3", AMERICAN JOURNAL OF CLINICAL
NUTRITION, 54:903-908 (1991), and Draperet. al., "Calcium Glycerophosphale as a Source of Calcium and Phosphorous in Total Par~nlcr~l Nutrition Solutions", JOURNAL
OF PARENTERAL AND ENTERAL NUTRITION, 15(2):176-180 (1991) showed in low birth weight infants and piglets, respectively, that CaGP is as effective as calcium gluconate as a source of calcium in total parenteral nutrition (TPN) solutions and could be used to prevent under mineralized bones in low birth weight infants.
Yet another reason for selecting CaGP was its high solubility which f~cilit~tes a WO96/31130 PCTrUS9''~C01 larger calcium intake per serving. A number of calcium salts were evaluated for their functionality in the liquid nutritional product of the presenl invention: ~;c~ci~lm phosphale, monocalcium phosphale, calcium chloride, tricalcium phosphdle, calcium citrate, calcium carL,ondle, CaGP, and D-gluconic acid (hemicalcium salt). Aqueous solutions conl~.i.,i"g 500 mg of calcium per 237 mL (8 oz.) serving (2110 ppm) were plepdl~d and the pH was ~fijusted to pH 3.5 and pH 5Ø Results indicated that solubility of calcium salts varied and only calcium Cdl bondle, calcium chlo~ ide, CaGP, and D-Gluconic acid, remained soluble at pH 3.5 for at least one month. In this evaluation sol 1hility was del~""i"ed by a visual examination. At pH 5.0 all samples formed crystals over time. The results of this solubility study are presenled in Table 3.

SOLUBILITY OF CALCIUM SOURCES
(Solutions at 500 mg calcium per 237 mL) Salt At Time of Manufacture 1 MONTH
pH 3.5 pH 5.0 pH 3.5 pH50 Dicalcium insol~ ' 'e insoluble inso' ~' 'e insol~' !e Phosphdle Monocalcium insoluble insoluble insoluble insoluble Phosphal~
Calcium soluble soluble soluble insc' ' !e Chloride Tricalcium inscM' !e illsQ' ~' !e j"s~l l' 'e insol ~' 'e Phosphdle Calcium insoluble insoluble insoll''o insoluble Citrate Calcium soluble partially soluble ills~l l' !e Carbonate Soluble CaGP soluble soluble soluble insoluble D-Gluconic- soluble soluble soluble partially Acid* Soluble * Hemicalcium salt W O96131130 PCTrUS9''01C01 Experiments were repe~ted with calcium carbonate, CaGP, and calcium chloride in a complete liquid nul,ilional product matrix, i.e., in conjunction with aspd,ldl"e, a flavor system and vitamin C. The pH range ev~ ted was 3.54.5. At the lower end of the pH
range all calcium sources were soluble at time of manufacture. After one month it was observed that as the pH increased, calcium Cdl bOndle and CaGP formed crystals, worse in the case of calcium Cdl Londle. In addition, it appeared that the CaGP had a synergistic effect with aspa, lar"e regarding sweetness. Calcium chloride was cor"~let ~ ly soluble throughout the pH range but its bitter flavor made it Ul ,acce,~ lable for the liquid nutritional product of the present invention a~pliG;tlion. Calcium lactate was evaluated in s~ ~hsequent experiments. Although its sol ~hi'ity was excellent it provided a~l, i"ger,l and mineral salt-type notes to the taste of the beverage that made it u"desi, ' 'e.
Still another reason for selecting CaGP is the fact that a beverage matrix containing this calcium salt requires the addition of less acid to acl.i_vc a pH below 4Ø
Acidity is desired in the liquid nul, ilional product of the plesenl invention for several reasons such as: to Illaillldi.l the calcium salt solubility, to cor, r!en,enl flavor, to control ".i obial growth, and to enhance the role of preservatives, specifically potassium ben~ dle or sodium ben~odle. On the other hand, too much acidity can result in increased lal ll ,ess and sourness that make the product u"desi, ' ' = from a sensory point of view. When calcium salts are added to the liquid nutritional product of the pr~senl invention, the solution resists chanyes in pH and more acid is needed to bring down the pH than in commercially available sodas with no calcium fortification. Aqueous solutions of various calcium salts were prepared to deliver 500 mg of elen,enlal calcium per 12 oz.
serving (1408 ppm) and the pH adjusted to pH 3.5 with citric acid. Till ' "e acidity was determined by measuring the amount of 0.1N NaOH needed to raise the pH to 8.3 in a 409 sample co"'- ,i"g 1,409 mg/Kg of a calcium source. The results presented in TABLE 4 indicate that, with the exception of calcium cl-loride, CaGP was the calcium salt that had the lowest lill ' ~'e acidity. Till ' ' l~ acidity is an indication of the total acidity of a beverage.

W O96/31130 PCTrUS~ SC01 TITRABABLE ACIDITY OF CALCIUM SOURCES
Calcium Source Tifratable acidity mL of 0.1N NaOH

Calcium Chloride 0.7 CaGP 43.5 Calcium Lactate 47.1 Tricalcium Pl ,osphale 48.6 Calcium Citrate Malate 53.2 Calcium Citrate 57.5 Calcium Hydroxide 60.6 Calcium Carl,onale 61~4 Calcium glycerophosphate (CaGP) is created by the reaction of glycerophosphate, a weak acid with pKf=6.1, with the strong base calcium hydroxide. GaGP binds calcium with an approximate fu~malio~ con~ nl of 1.7. CaGP, when dissolved in water, dissociates readily to provide "free" calcium ions and p, ulon~led glyu3rophosph~l~
species. Acid-base buffering by monoprulondled glycerophosphale is effective only within the pH range from 4.1 to 8.1 (refer to the Hender~on I l~ssell ~al,k equation), and thus, GP exhibits il ,siy"irlca"l buffering capacity at pH=3.6. On the other hand, anions, such as malate, tartrate, pr~p.onale or succ;"~le, do provide buffer capacity at pH=3.6, and accordi. Igly require more base or acid than GP for final adjustment of pH.
Yet another reason for selecting CaGP is the low aluminum content in commercially available CaGP. It has been theorized that chronic use of calcium supplements which have siyl ~iricanl aluminum conlenls may constitute unnecessary metal exposure. Whiting, "Safety of Some Calcium Supplements Questioned", NUTRITION
REVIEWS. 52(3):95-97 (1994). The aluminum content of some calcium sources is presented in TABLE 5.

W O96/31130 PCTrUS9G/OlG01 ALUMINUM CONTENT OF CALCIUM SOURCES
Calcium Source A/uminum Content in parts per mil/ion (ppm) CaGP 4.55 Calcium Hydroxide 300400 CaCO3 (from fossil shell) 4,4002 CaCO3 (from Dolomite) 171-3152 ' Values deler~ ed by analysis of commercially available compounds.
2 Values from Whiting article.

It has been sugge: ilecl that calcium citrate may play a role in enhal)~i"g aluminum absorption from food, polenlially resulting in toxic serum and urinary aluminum levels.
Sakhaee et al., have successfully demon~l,alt:d however, that the provision of calcium citrate alone without aluminum - conl~i. ,i"g drugs does not pose a risk of aluminum toxicity in subjects with nor",ally functioning kidneys. Sakhee et al., "Calcium citrate without aluminum a nlacids does not cause aluminum retention in p~lienls with fu"~ioni.,g kidneys," BONEAND MINERAL. 20:87-97 (1993).

Vitamin D. As used herein and in the claims the terms "vitamin D" and "various forms of vitamin D" are understood to refer to vitamin D, cholecalciferol (D3), ergocalciferol (D2) and its ki~logiG~Ily active metabolites and precursors such as, 1a, 25-(OH)2 vitamin D; 25 OH vitamin D, its biological precursor; and 1a hydroxyvitamin D, and analogues of the dihydroxy compound. These materials promote i"lesli"al absor~lion of calcium, contribute to plasma calcium regulation by acting on the remodeling processes of accretion and resorption and stimulate reabsorption of calcium by the kidney. While the form of vitamin D3 used in the following exar,,Flos~ prototypes and experiments is cholecalciferol, it is understood that any of the various forms of vitamin D may be used in practicing the present invention, but vitamin D3 is preferred in embodiments which are liquids.

W O96/31130 PCTrUS96/01~01 Dietary calcium and vitamin D are the natural mediators against bone loss.
Vitamin D acts directly on bone cells (osteoblasts, osteoclasts) to alter bone mass. It also promotes gut uptake of calcium. Human skin activates pre-vitamin D molecules when exposed to ultra violet irradiation. In the summer, 15 minutes exposure to sunlight is sufficient to maintain adequate vitamin D levels. On the other hand, during winter, all day exposure to sunlight will produce negligible conversion of vitamin D. The thinner skin associated with aging is a less effective converter than the youthful skin.
The addition of vitamin D to the liquid nul,ilional product of the p,~senl invention was difficult bec~use this is an oil soluble vitamin whereas both the beverage concenl, dle and the beverage of the pr~:se"l i"~ lion are ~queous solutions. A number of po~ !e methods to overcome the immiscibility of these two phases were ev~lu~ter~ The results of these efforts are related below, and batch numbers are sequential throughout the following studies.
There were two major obstacles to overcome regarding the i"Col ~ordLion of vitamin D3 in the preser,l invention: (1) the initial prucessi"g loss of vitamin D3; and (2) the stability of vitamin D3 over the shelf life of the product. To cor"pa, ~: the initial processing loss and stability of vitamin D3 of each variable with successive batches, two criteria were routinely measured: (1) % recovery of vitamin D3 at 0-time; and (2) half life of vitamin D3 (t1~) The % recovery of vitamin D3 of each batch was c~lcl ~l~tPd by dividing the 0-time vitamin D3 result by the theoretical rO, liricdlion of each batch times 100%. (See Table 6).
As used herein "theoretical rO~ liricdlion" refers to amount of vitamin D3 added to the product. As used herein "0-time" refers to the time of initial vitamin D3 analysis of the product. In Table 6, "% Recovery" is the per~;enlage of theoretical fo, liricdlion of vitamin D3 remaining in the product after initial plucessi,lg loss. Only batch 31 did not have the 0-time vitamin D3 determined. Tl)ererurt:, a proj~ct~d result for this batch was extrapolated from the negative exponential regression curve generated from the stability data.

CA 022l7264 l997-l0-02 W O96/31130 PCTrUS9''01C01 O-TI_F VITAMI ~I n~ RF.~UI TCi VF ~SUS THFORFTICAI Ft~RTlFlcATlnN
BATCH 0-TIME THEORETICAL % RECOVERY
440 950 46.3%
- 2 405 950 42.6%

3 450 950 47.4%
Mean for batches 1-3 45.4%

4 249 635 39.1 %
283 635 44.6%
6 294 633 46.4%
Mean for batches 4-6 43.4%
7 371 483 76.8%
8 328 634 51.7%
9 308 618 49.8%
Mean for balches 7-9 59.4%
548 841 65.2%
11 696 844 82.5%
12 680 843 80.7%
13 691 844 81.9%
14 546 842 64.9%
649 845 76.8%
16 679 844 80.5%
17 681 844 80.7%
Mean for batches 10-17 76.7%
18 752 916 82.1%
19 678 915 74.1%
802 916 87.6%
(control batch) 21 784 917 85.5%
22 491 917 53.5%

23 796 916 86.9%

W O96/31130 PCTrUS9''0~GOl TABLE 6 (continued) BATCH 0-TIME THEORETICAL % RECOVERY
24 7998 916 J 87.1%
Mean for balches 18-1978.2%
& 21-24 473 826 57.3%
26 526 825 63.8%
27 539 825 65.3%
28 633 825 76.7%
29 517 793 65.2%
(control batch) 576 823 70.0%
Mean for balcl ,es 25-28 66.6%
&30 31 786*** 840 93.6%
** No Data Available - Ex.,~pGI-'ed From the Exponential Rey~ession Curve To better characterize the stability of vitamin D3 over time in all the bdlcl ,es the Henri-Michaclic Mcnton ex~ oner,lial equation was e" p'oyed. The vitamin D3 results (IU/KG) for each variable was plotted versus time (Day) and a ~ey,~:ssion curve was fitted through the data using the follo.~-;"g equation:
lD] = [Do] e~kt Where: [D] = Vitamin D3 conce"l,dlion (IU/KG) at time (t).
[Do] = Vitamin D3 concenll~lion (IU/KG) at 0-time.
e = Exponential k = Rate conslanl (rate of ioss of vitamin D3 over time) t = Time (days) Stability was defined as the amount of time (days) that would be required for the initial concentration of vitamin D3 to be reduced one half. This was termed half-life (t"2). The more stable the vitamin D3 in a particular formulation the longer it would take for the initial concentration to be reduced by one half. Rean dny;"g the previous equation and making the appropridte sl lhstitutions~ the half-life of vitamin D3 in a W O96/31130 PCT~US96/04601 particular variable could be ex~,essed as:
t"2 = In 2/k t"2 = Time (days) required for vitamin D3 to be reduced by one half of the initial concenl,dlion.
In = Natural log.
k = First order rate constant (rate of loss of vitamin D3 over time):

The various bdlcl1es are described in the r "~J.~i,lg text. For convenience, thebatch numbers are sequential. In addition, the actual vitamin D3 data at each time point for each respective variable are presenled in Tables 8, 10, 12, 13, 14, 16 and 17.
The correlation co~rri onls, initial vitamin D3 concer,l,~lion [Do]~ first order rate consla"l:~ (k), and vitamin D3 half lives (t"2) are also pr~senled in Tables 8, 10, 12, 13, 14, 16 and 17 and should be referred to during the discussion.
A del '_d ~iscuscion of each variable will not be prt:senled since such a presenl~lion would be quite lengthy. Rather an overview of various balcl ,es grouped with respect to the main Vdl ~ '-'~ S that were studied will be ~iscusse~

a. Use of a Water Disper~ 'e Form of Vitamin D3 Early in the dcvelop",ent of the present invention an evaluation was made of a water dispersable vitamin D3 spray dried on a dicalcium phosphale and gum acaciacarrier. The water dispersible vitamin D3 used in this evaluation was a DRY VITAMIN
D3 Type 100-DS pu,chased from Roche Vil~lll ,s and Fine Cher"icals, a division of Hoffman-LaRoche, Inc., Nutley, New Jersey, U.S.A., which conl..i.,s vitamin D3 (chclec~lciferol USP-FCC), 1;~ phosph~le, gum acacia, coconut oil, BHT, lactose, silicon dioxide, sodium ben~u~l~ and sorbic acid. It is a white powder and contains 100,000 IU/g of vitamin D3.
Three batches were manufactured to evaluate the water di~per~ ' le form of vitamin D3. Each batch consisted of an aqueous solution containing pc,lassium benzoate, citric acid, sodium citrate, aspd~ lal"e, calcium glycerophosphate and the water dispersible form of vitamin D3. The resultant product was not homogenized.The final pH of each batch is presented in Table 7. This pH difference, however, did not seem to affect vitamin D3 recovery.

WO96/31130 PCTrUS~6/O~C01 BATCH pH
3.50 2 4.19 3 4.97 The initial prucessil ,g losses for bal~,hes 1-3 was severe (mean = 45.4%
Recovery - Table 6). The loss of vitamin D3 was primarily due to: (a) the fact that the vitamin D3 was not homogeni~ed into the product matrix; and (b) there was no emulsifier pr~se"l that would assist in maintaining the vitamin D3 in solution.
Therefore, the vitamin D3 was lost by the coating of the manufacturing equipment with vitamin D3. The stability of Vitamin D3 in these three balcl-es was not ~cceFt~~'e over the shelf life of the product. As shown in Table 8, one half of the initial vitamin D3 was lost in approxil"dlely 12.6 days.

VITAMIN D~ (IU/KG OF PRQDUCT) VERSUS DAYS

Days1 o2 440 405 450 Corr. Coef. 0.955 0.968 0.965 [Dol 462 418 466 k 0.0636 0.0501 0.0529 t.,2 10.9 13.8 13.1 Average Half Life (t 1e) of Vitamin D3 for Batches 1-3 is 12.6 Days Days after initial vitamin D3 testing. 0-time testing occurred 7 days after the product was manufactured.
2 Results for batches 1-3 were corrected via control value on day 0 and day 7.
b. Use of Polysorbate 80 as an Emulsifier W O96131130 PCTrUS_~'01C01 A series of ex~.eri",ents were conducted using vitamin D3 in Polysorbate 80 manufactured to selected speciricdlions by Vildlllills Inc., Chicago, lllinois, U.S.A.
Polysorbate 80 is a water soluble, non-ionic emulsifier used for various ap~ ions in the food industry. It is a polyoxyethylene derivative of sorbitan monooleate which i"laracl:j with the oil and aqueous phases in an emulsion to form a barrier at the interface that causes a reduction in Van der Waals forces and an improvement in emulsion stability. It was e,~,uel,led that the use of Polysorbate 80 to incorporate the vitamin D3 would improve its recovery and stability by causing di~ er~iGn of the oil phase in the continuous ~queous phase.
The effect of Polysorbate 80 was evaluated in three ex,ueri,,,ental bdlcl-es of a low pH beverage. Liquid beverage concer,l,dles were pr~:part:d as desc,ibed above, i.e., adding to water sodium ben~oale (instead of polassium benzoate as a preservative), citric acid, polassium citrate, aspa,ldl"e, calcium glycerophosphale, and vitamin D3 in a premix cor,l~i.,;"g Polysorbate 80 and propylene glycol. The resultant liquid beverage concer,l,dles were not homogenized and were diluted with five parts of water before carbonation. The vitamin D3 fo, lificdlion level for each batch was 635 IU/KG of product. All bdlches contained vitamin C. The vdr ~les in bdlches 4-6 are presented in Table 9. These Vdli~ s were added in an all~:lll,ul to improve vitamin C
stability, since it has been found that cysteine, when added in a carefully conl,~"ed amount can overcome vitamin C deleriordlion in packaged beverages (U.S. Patent 3,958,017, May 18, 1976).

BATCHVARIABLE
4Cysteine,1.5% of Vit. C

5 No Cysteine 6Cysteine + 250 PPM WPC

The overall mean % Recovery for batches 4-6 was comparable to the previous - batches conldi"i"g the water dispersible form of vitamin D3. The mean % Recovery was 43.4% (Table 6). However, as shown in Table 10, the stability of these bdlches improved significantly. The half life of vitamin D3 in these batches ranged from 257 CA 022l7264 l997-l0-02 W O96/31130 PCTrUS~'VlC01 days to 1,160 days. Cysteine addilion did not affect vitamin D3 recovery, but batch 6 with whey protein concenl~le (WPC) showed minimal loss of vitamin D3 during 60 days of shelf life.
In addilion, batch 6 also had slightly better initial vitamin D3 Recovery than those bdlches in this series without protein. This suygesled that a more rugged emulsion and some sort of matrix was needed as shown in Table 10. The use of WPCis not i"~ I;c-.led if the product of the invention is desired to be low in calories or free of calories, but otherwise may be used in the practice of the invention. In an dll~ L to make a low calorie or calorie free product, the use of mechal-ical means such ashomogeni~lion was inve~ !~d VITAMIN D3 (IU/KG QF PRODUCT) VERSUS DAYS

Days' Corr. Coef. 0.450 0.783 0.512 [Do] 236 275 288 k 0.0015 0.0027 0.0006 t"2 462 257 1160 Average Half Life (t1,2) of Vitamin D3 for Batches 4-6 is 626 Days.
' Days after initial vitamin D3 testing. 0-time testing occurred 2 days after the product was manufactured.

W O96/31130 PCTrUS9GI'~1C

C. Use of Homogeni~dlion In the next series of studies, the vitamin D3/Poly~orL,dle premix was combined with the aqueous phase and the blend was emulsified by passing it through a two-stage Gaulin-L-100 homogenizer at a given pressure. The purpose of this homogeni~dlion step is to break up, or evenly disperse the oil phase into the aqueous phase so that the particle size of the emulsion is Surr enlly small to retard coalescence of the oil phase and prevent sepa~ dlion. A two-stage homogenization is needed since the fine pa, licles formed during the first stage can clump. The second stage, set at a lower pressure, is needed to break up the clumps, thereby making a more stable emulsion.
Brominated vege: ~'e oil (BVO) and small quanlilies of gum arabic were added to the vitamin D3/Polysorbate premix prior to homogehi~dlion. This was done to increase the specific gravity of the oil phase and avoid phase separdlion or oiling-off of the emulsion. BVO is used in the soda industry as a sl~ er for flavoring oils used in fruit flavored beverages. BVO is a Food Additive (21, CFR 180.30) allowed in an amount not greater than 15 ppm of the ri~ ,ished beverage.
A series of ex~,eri",ents were conducted to evaluate the effect of homogeni~dlion on vitamin D3 recovery and stability. In these ex~ueri",ents liquid beverage concenlldles were pr~pal~d as desc~iL,ed above with the exception of the vitamin D3 addition. All the water soluble componenls were first dissolved in water and a vitamin D3 emulsion, prepared separdl_ly was added at 1% of finished product concenl,dlion and mixed thoroughly. The vitamin D3 emulsion was prepared by cor"l.i"i. ,9 water vitamin D3 and one or more of the following ingredients: Brominated Vegetable Oil (BVO) Polysorbate 80 Gum Arabic (GA), and corn oil followed by homogeni~dlion using a two stage homogenizer. Two dirrt:r~nL sources of vitamin D3 were used: (a) an oil soluble vitamin premix where the vitamin D3 is dissolved in a small amount of corn oil; and (b) a vitamin D3 premix where the vitamin D3 is dissolved in Polysorbate 80 and propylene glycol (PG) (same as bdl~ l-es 4 through 6). One part of the complete concentrate was then dissolved with five parts of water before carbonation. The variables in batches 7-24 are presented in Table 11.

Wo 96/31130 PCTlu~J ~l~ ~C01 Batch Variable (ppm = ppm of product) 7 BVO, Vitamin D3 in Corn Oil, GA (0.14 ppm) 8 BVO, Corn oil, Vitamin D3 in PolyaorL~dte 80, GA (0.14 ppm) 9 BVO, Corn Oil, Vitamin D3 in Polysorbate 80 BVO, Vit D3 in Corn Oil, Polysorbate 80 (0.07 ppm), GA (0.14 ppm) 11 BVO, Vit D3 in Com Oil, Pol~/aorlJdL~ 80 (0.035 ppm), GA (0.14 ppm) 12 BVO, Vit D3 in Corn Oil, Poly:,o,L.ale 80 (0.035 ppm), PG (0.15 ppm), GA
(0.14 ppm) 13 BVO, Vit D3 in Corn Oil, Polyao~L~dl~3 80 (0.07 ppm), GA (0.14 ppm) 14 BVO, Vit D3 in Corn Oil, Polyaort~ale 80 (0.07 ppm), PG (0.30 ppm) BVO, Vit D3 in Corn Oil, PolyaorL~aLe 80 (0.035 ppm) 16 BVO, Vit D3 in Corn Oil, Poly:,o,l,.lle 80 (0.035 ppm), PG (0.15 ppm) 17 BVO, Vit D3 in Com Oil, Pol~aorl~al~a 80 (0.07 ppm) 18 Same as 11 19 Same as 12 Same as 13 21 Same as 15 22 Same as 16 23 Same as 17 24 BVO, Vit D3 in Corn Oil, Polyao~l~ale 80 (0.07 ppm), Fructose (42,000 ppm) The gum arabic used in all batches was Nutriloid Gum Arabic from Tic Gums, Inc. When extra Polysorbate 80 was added to the batches, the percent addition refers to percent of oil in the batch. Batches contain either 3% or 6% extra Polysorbate 80 added. The 20% and 40% refer to co~ nb;"~lions of Polysorbate 80 and Propylene glycol where the Polysorbate 80 content is 3% and 6%. Fructose was added to batch number 24 to see if it would extend the shelf-life of the product which is limited by the degradation of aspartame. In general, the fructose and the various levels of CA 022l7264 l997-l0-02 W O96/31130 PCTrUS~ 1C01 POIy~Or~dLe 80 did not affect the vitamin D3 recovery as the homogeni~dlion step did.
The initial vitamin D3 Recovery (mean = 59.4%) and the mean half-life value (150 days) for batches 7-9, as prt:sef lled in vitamin D3 Recovery Table 6 and Table 12, in-~icat~sl that with few excepLions the homogeni~dLion step significantly improved the initial recovery and stability of vitamin D3 versus previous dLLe,n~L~.

VITAMIN D3 (IU/KG OF PRODUCT) VERSUS DAYS

Days' Corr. Coef. 0.816 0.224 0.506 [Do] 349 324 309 k 0.0098 0.0030 0.0047 tv2 71 231 147 Average Half Life (t"2) of Vitamin D3 for Ldl~ l.es 7-9 is 150 days.

Days after initial vitamin D3 testing. 0-time testing occurred 8 days after the product was manufactured.

The vitamin D3 results for batches 10-17 confirmed that homogenization was necessary. The mean % Recovery for these batches d, ~r"dLically improved to 76.7%
versus all previous batches (Table 6). The overall vitamin D3 stability (mean half-life =
68.6 days) for batches 10-17 as presented in Table 13, was not as good as batches 7-9 (150 days) but was superior in cor"pari~on to batches 1-3(12.6 days).
In orderto confirm the initial vitamin D3 Recovery and stability of batches 10-17 duplicate batches were made (see bdLcl1es 18-19 and 21-24 in Tables 6 and 14). The W O96/31130 PCTÇUS~ SC01 initial vitamin D3 Recovery for balclles 18-19 and 21-24 (mean = 78.2%) corroborated previous recoveries for batches 10-17. Furthermore the vitamin D3 stability of bal. hes 18-19 and 21-24 (mean half-life = 76.7 days) was cor"pal~ble to their respectiveduplicate batches (68.6 days).

W O96/31130 PCTrUS~'OlC01 OD ~ ~ ~ 0 ~ O ~
CD ~ ~ ~ ~ O U~ O 1--c~ 0 r-- OD ~o ~~ ~ ~ ~.
C~ O ~D O

C~ oo U~
O ~ O CO ~
Cl~ o ~n 0 1' o 0 ~ ~ ~ ''I ~ O ~ o ~ ;~:
C ~. ~
G ~ -cs ~
j~ O C~ 0 ~ 0~ ~ 0 0 0 ~ ~ ~
~ ~O ~ ~ ~ ~ O ~ O ~ ~ ~

.C
O C~ , z ~ ~ ~ ~~ -~ ~
c o o >
~0 CD ~ ~ 0 0 .C 0 ~t C~ ~ o ~ o C~
o .C

o ;$ LO U~ ~ ~ o O
I

I ~ ~ 0 ~ ~ o ~ ~I ~ ~ ~ C1~ a~ ~

W O96/31130 PCT~US9''01C01 co ~ ~> o "~ c~ O a~
G ~ '- o ~ o ~ a) C~

C~l 1~ U~ Ct~ C~ O ~ O ~D
~ L~

' ~ ~ G ~ ~ ~ ~ ~o o. ~ ~o, o ~ o ~ ~ Q
s C~ O C
~; ~ N ~ o ~ ~ ~ o ~ o ,~
~ ~ ~~ 2 o s ~t o m o N q~
N OD U~ t) O 1-- 0 ~D ~ ~, Cl ~ a Z 1-- a 1-- cn o o ~~ N ~
~D ~ C~ C'~ O ~ O
~ _.

C~l CO C7) ~ ~ ~ ~o ~ o .
C~ ~ I~ ~ ~ ~5 O ~ ~

m ~ O ~' '~ ~ ~ u a~ ~

W O 96/31130 PCT~USS~'~1C01 Although the shelf life data for batches 10-17 and 18-24 sho~.~d a loss of vitamin D3 as a function time, no siy"iricdnl amount of de_~ dddlion product could be analytically d~lected. Therefore, the main ",echar,i.~", for loss was assumed to be physical ",4,dlion of vitamin D3 to the walls of the container, and/or rapid oxidalion of vitamin D3 and/or isomerization of vitamin D3 to 5,6-trans-vitamin D3. Further studies focused on incit:asi"g the emulsion stability to prevent the l"i~,dlion of the hydrophobic vitamin D3 to the container walls.

d. Use of Gum(s) as an Emulsion Slabili~er The use of gum arabic and gum tragacanth as emulsifying agents for flavor oils in soft drinks is welbo ~ ~' ' ,ed in the soft drink industry. Melillo, "Physical Factors Governing the Sl~ lion of Cloudy Beverages", FOOD PRODUCTS
DEVELOPMENT, June,1977, pp. 108-110. While only gum arabic was used in the ex,ueri",e"ls, exdl", 'es and prototypes disclosed herein, it is u"der:,tood that one skilled in the art could sl ~hstitute appl upridle amounts of gum tragacanth, xdnll Idl I gum or any other apprupridle gum into the products of the pr~senl invention, or thatmixtures of gums may be used in the practice of the preser,l invention.
Gum tragacanth is the dried, gummy exudation obtained from Astragalus gummifer or other Asiatic spe~ ~ s of Astralagus. Tragacanth swells rapidly in either cold or hot water to a viscous colloic'-' sol or semi-gel. The mo'ecl ll:3r weight of the gum is on the order of 840,000 and the molecl~4s are elongated (4500A by 19A) which accounts for its high viscosity. Tragacanth gum is c~lllF ' Ie with other plant hydrocoll~i~'c as well as carbohydrates, most prolei.ls, and fats. Viscosity is most stable at pH 4 to 8 with a very good stability down to pH 2.
Xanthan gum is an exocellular heteropolysaccharide produced by a distinct fermentation process. The bacterium xanthomonas campestris generates the gum on specific organelles at the cell surface by a complex enzymatic process. The molecl ll~r weight for xanthan gum is about two million.
Gum arabic, also known as gum acacia, is the dried, gummy exudate from the stems or branches of Acacia senegal or of related species of Acacia. The most unusual property of gum arabic among the natural gums is its extreme and true solubility in cold or hot water. Gum arabic is a complex calcium, magnesium, and W O96/31130 PCTrUS~6/01G01 potassium salt of arabic acid. It has a main backbone chain of (1 , 3) - linked D-g-'~cl~pyranose units, some of which are sllhstitllt~d at the C-6 posilion with various side chains. The side chains consist of D-g-'nctopyranose, D-glucuronic acid and L-arabofuranose with addilional side chains on the D-galactopyranose of L-rhamnopyranose. The mo' eol ll~r weight is on the order of 250,000.
Gum Arabic is effective in ~ g emulsions and illhiLilillg co-'escence or phase separation by two mechanisms: (a) increasing the viscosity of the continuous (~queous) phase; and, (b) for",i"g strong films around the oil ~llop!_' . A small amount of protein is presenL in the gum arabic as a part of the structure.
A series of experiments were conducted to evaluate various types of gum arabic as the emulsifier system in the vitamin D3 emulsion. Although gum arabic had been evaluated in previous experiments, the usage rate was too low (0.14 ppm) tohave a significant effect. It has been reported that the p,uL~ ceous component is responsible for gum arabic's emulsifying and stabilizing prope, Lies. The vdl ~"'es in b~Lches 25-30 are plt:sehLed in Table 15.

Batch Variable Gellan Gum, Kelco Products, 100 ppm (in beverage) 26 Gum Arabic, Tic Bev 202, Tic Gums Inc., 2000 ppm (in beverage) 27 Gum Arabic EMULGUM, Colloids Naturels Inc., 500 ppm (in beverage) 28 Gum Arabic Nutriloid, Tic Gums Inc., 2000 ppm (in beverage) 29 Control, Same as Batches 13 and 20 Gum Arabic SPRAY BE, Colloid Naturels, Inc., 500 ppm (in beverage) Batches were prepared to evaluate the stability of various vitamin D3 emulsions in finished beverages. The individual emulsions, prepared separately, were added to beverage concentrates in amounts to yield 1 % by weight in the finished beverages.
The emulsions themselves contained 1-20% by weight of the appropriate gums whichwere first hydrated in aqueous solutions for about two hours at 60~C. (See Table 15 for gums and quantities) The hydrated gum solutions were cooled to 37.8~C or less W O96131130 PCT/US9C/O~C01 before the needed amounts of vitamin D3 were added. The type of vitamin D3 used was liquid vitamin D3 in corn oil oblai"ed from Roche Vildmi, ,s and Fine Chemicals, a division of I lorr",an-LaRoche Inc., Nutley, New Jersey, U.S.A. The pH of the emulsions which contained gum arabic were lowered to pH 4.0 and sodium benzoate was added to preserve the emulsions for e,clended use. The emulsions were then homogeni~ed twice using a two-stage homo~eni~er at 1,500/600 PSI and 3,000/1,000PSI, respectively. For example, batch 27 contained 50 grams of EMULGUM gum arabic hydrated in 950 grams of water, and upon cooling 77.2 milligrams of liquid vitamin D3 in corn oil was ':lended into the gum solution in an amount giving a theoretical fo~ liricdlion of about 825 IU/Kg of ri",shed beverage. The emulsion was preserved by adding 0.3 9 of sodium benzoate and the pH was lowered to 4.0 by adding 1.08 grams of citric acid.
The pe, r..r",ance of the different gums used, as indicated by initial vitamin D3 recovery and stability over shelf-life varied (Tables 6 and 16, respectively).
EMULGUM (batch 27) at 500 ppm concer,l~ dlion gave the best results followed by SPRAY BE, both from Colloids Naturels, Inc.
In generaHt can be said that siy~iril,d~l improver,~enl~ in vitamin D3 stabilitywere observed initially and during shelf-life. The most significant improvement was the stability of vitamin D3 over the shelf life of the product. The average half-life of vitamin D3 for these bdlches was 180 days. It appears that at sufficient concentration, gum arabic can coat the oil dl~Fl~t~ con Idil lil Ig the vitamin D3 to form an emulsion that can be further st7hi~i~Pd by homogeni~dlion using a two-stage homogenizer.
This series of experiments demon:,l, dl~d that gum arabic could be substituted for Polysorbate 80 to minimize initial processi"g loss and improve shelf life stability of vitamin D3.

W O96/31130 PCTrUS9~'OlC01 ,~n o ~ ~ ~ ~ co C~~ ~ ~~ O 0 U~ U~ ~ ct~ C~o U~ o ~ --o~

~ ~Dm ~
u~ o ~ o ~ ~ a~

CD ~ C~ C'' 1'- 0 C'' ~ 0 ~ G ~
llJ C ~ ~, ~ ~ ~ ~~ ~ ~ ~ 8 -o ."
~~ ~,.
C~ c~ ~ c~l C~ ~1 OD ~~ C~ g ~ '-- ~
z ~ ~ 0 N ~) ~ U~ C~ ~ C~l ~ ~O ~ O
~ > 'C-> o a a) ~ ~ ~
g~ C
~ ~ ~ ~ ~ ~ C~l O C~ O

I ~ I
~ ~ ~ C~J o ~ ~ C~l o C~ C

W O96/31130 PCTrUS3~'01G01 e. Use of Commercially Manufactured Vitamin D3 Emulsion In order to evaluate if a sl lit~ ~le vitamin D3 emulsion could be manufactured on a larger scale which would support commercialization of a product accord" ,g to the invention a decision was made to have the vitamin D3 emulsion manufactured by anoutside contractor. Taslt:r"aker Inc. of Cincinnati Ohio U.S.A. which is a providerof flavoring products provided as a special order a vitamin D3 emulsion containing water gum arabic partially h~dluger,dled soybean oil citric acid sodium benzoate and vitamin D3. By actual analysis this cor"r"er- ;ally manufactured vitamin D3 emulsion contains per 10 Kg: (a) about 9.52 Kg of water; (b) about 0.35 Kg of gum arabic; (c) about 0.10 Kg of partially hydlugendled soybean oil; (d) about 0.02 Kg of citric acid; (e) about 0.01 Kg of sodium benzoate; and (fl and at least about 787 000 IU of vitamin D3.
Tastemaker considers the manufacturing procedure it used to be prù~ - ie~a,y to it and did not make that i- ,rur, . .alion available. While the CGI - -- ~ ,er~ ;ally manufactured emu!sjon Gon!~ ed partia!!y hyd!--gPn~!Hd soybean oi! and the se!.f-manu.factur~d emulsion conldil ,ed corn oil (see desc~ i~lion of bal- hes 25-30) it is understood that the invention may be practiced using any sl~ e veget~le oil. Batches of which batch 31 is typical were manufactured as desc.iLled in previous e~,eri...er.l:,. The commercially.manufactured vitamin D3 emulsion was added to the liquid beverage concer~l~dle in an amount to equal 1% by weight of the finished beverage. The beverage concer.l,~le was then added to water at a ratio of 1:5 and carbonated.
The initial vitamin D3 loss for batch 31 was minimal (94.1% recovery) which had surpassed all batches to date. Furthermore the vitamin D3 stability of this batch was superior to all previous balches. As presented in Table 17 the half life of vitamin D3 was 1 390 days.

W O96/31130 PCTrUS9''0~G01 VITAMIN D3 (IU/KG OF PRODUCT) VERSUS DAYS

Days' 0 Not Tested Corr. Coef. 0.218 [Do] 786 k 0.0005 t"2 1,390 days Days after product manufacture, with day 0 being the day on which the product was manufactured.

Acidulants. Acids are commonly used in food and beverages to impart specific tart or sour tastes and to function as preservatives. A combi.-dlion of citric and lactic acids are used in the liquid nul, ilional product of the pr~se"L invention. Citric acid is the most widely used acid in fruit beverages in part because it blends well with these flavors. It is commercially manufactured by fermenl~lion or by synthesis; either may be used in the practice of the present invention. When using fermented lactic acid, a purified form that is free of sugar rcsidues is recommended due to its cleaner taste and clearer appearance. Food grade lactic acid is available in aqueous and crystalline forms.

Sweetener. The sweetener used in the prototype beverages described below is aspa, l~" ,e, but other artificial or natural sweeteners can be used in the practice of the pr~senl invention. Artificial sweeteners that may be employed include saccharin, acesulfame-K and the like. Natural sweeteners that may be employed include W O96/31130 PCTrUS9''01G01 sucrose fructose high fructose corn syrup gl-lcQse sugaralcohols dextrose lode,~l- il ls maltose lactose and the like but other carbohydrates can be used if less sweetness is desired. Mixtures of natural sweeteners or artificial swectcners or natural and a, liri~ ial sw~:etener:j can be used also.
The amount of the l~weetaner effective in a product according to any aspect of the present invention depends upon the particular sweetener used and the sweetness i"lensily desired. In deter",;.,i"g the amount of sweetener any sugar or other sJ~ ~ tener pr~senl in the flavor Gor"ponenl or product matrix should also be taken into consiclerdlion .
Studies have shown that the efficiency of calcium absGrl,lion can be enhanced two-five fold by oral admi";~l,dlion of glucose polymer both in patents with i"lt:sli"al calcium malabsor~lion and in normal subjects. Kelley et al. "Effect of Meal Co",posilion on Calcium Absor,ulion: Enhancing Effect of Carbohydrate Polyer' GASTROENTEROLOGY. 87:596-600 (1984).
In another study using the triple-lumen intestinal perfusion techn:~l le glucosepolymer increased net calcium absor,ulion fourf ~old. Bei et al. "~''ucose Polymer Increases and Equal Calcium Magnesium and Zinc Abso,l,lion in Humans"
AMERICAN JOURNAL CLINICAL NUTRITION. 44:244-227 (1986).
It is understood that a person of skill in the art may make a product in accordal1ce with the invention corlldil lil l9 9l~ ~cose polymers or glucose.

Flavor. As used herein the term "flavor" includes both natural and artificial flavors. The particular amount of the flavor col"ponent effective for i",pal li"g flavor characteristics to the beverage of the present invention can depend upon the flavor(s) selected the flavor i",pression desired and the form of the flavor component. The amount of flavor employed in a product according to any aspect of the present invention is within the skill of one in the art and depends on the flavor i"lensily desired.

Preservatives. Most microbial spoilage of low pH beverages is caused by aciduric and acidophilic organisms like certain vari~lies of yeasts and molds. For this reason preservatives with anti-microbial activity such as benzoic and sorbic acids are added to soft drinks. Usage levels of these acids or their salts range from 0.025 to 0.050 percent depencli"g on the nutritive subsldnces present and the pH of the W O96/31130 PCTrUS~-'01G01 finished beverage. The a"li",ic,ubial activity of these preservatives has been shown to be largely pH dependent. They are least effective under neutrai condilions but their activity increases considerdbly with decreasing pH. For exa" ple, by reducing the pH
value from 4.5 to 3.0, the preservative effect of benzoic acid is i"c,~ased by nearly three times. Only beverages at low pH receive the full benefit from the addilion of preservatives. Woodruf et al., BEVERAGES: CARBONATED AND
NONCARBONATED, The AVI Publishing Company, Inc., 1974, pgs. 143-146. As with most foods, the successful preservation of low pH beverages is dependenl on controlling contamination of ingredients, prucessi"g equipment, and conlc.;"er:i by poler,lial spoilage oryanisrlls~ Splittstoesser in FOOD AND BEVERAGE MYCOLOGY, edited by Beuchatt, pulJlished by Van N~_',dnd Re;.,h~'d, 1987, pgs. 120-122.

Carbonation. The amount of carbon dioxide in a beverage accordi"g to the present invention depends upon the particular flavor system used and the amount of carbonation desired. Usually, carbonaled beverages of the pr~senl invention contai from 1.0 to 4.5 volumes of carbon dioxide. P,~re"~d cd,l,onal~d beverages contain from 2 to 3.5 volumes of carbon dioxide. The beverages of the pr~ser,l invention can be prepar~d by ~landard beverage formulation tecl ",, les. To make a carbonated beverage carbon dioxide can be introduced either into the water mixed with the beverage syrup or into the drinkable diluted beverage to achieve carbonation. Itshould be understood, however, that ca,bonaled beverage manufacturing techniques, when appropriately modified, are also app'i--~lE to noncarbonated beverages.

W O96/31130 PCT~US9~ C01 EMBODIMENTS OF THE INVENTION

Tables 18-21 plesenl bills of materials for manufacturing prototypes of low pH
beverages ru~ liried with calcium and vitamin D3 in acco,dance with some aspects of the invention.

Bill of Malerials for Wild Cherry Fld~rGr.:d Bever~e (For 1000 KG of Beverage) INGREDIENT AMOUNT. KG
Treated Water' (for beverage cGncer,l,ale) 137.82 rol~-ss Mm Ben~oale 0.300 Sodium Citrate (dihydrate) 0.550 Citric Acid (anhydrous) 3.720 Lactic Acid (88%) 3.951 Aspa, l~",e 0.500 Calcium Glycerophosphal~: 8.331 vVild Cherry Color 0.000630 FD&C Red #40 0.0003465 FD&C Yellow #6 0.0002835 Natural & Artificial Wild Cherry Flavor 1.200 Ascorbic Acid 0.300 Vitamin D3 Emulsion2 10.000 Treated Water' (for final blend) 833.33 "treated water" has had the cl-lori"e and alkalinity ~ sted to levels cGr"r"or,ly used in the soft drink industry.
2 This emulsion is described above with regards to batch 31.

CA 022l7264 l997-l0-02 W O96/31130 PCTrUS3~'~1C01 Bill of ~ rials for Orange Fl..~rore.~ ~c~r~a (For 1000 KG of Cc~er ~e) INGREDIENT AMOUNT KG
Treated Water' (for beverage concentrate) 137.62 Boldssium Ben udle 0.300 Sodium Citrate (dihydrate) 0.550 Citric Acid (anhydrous) 3.720 Lactic Acid (88%) 3.951 Aspa"dr"e 0.500 Calcium Glyceruphosphal~ 8.331 Orange Color 0.0001875 FD&C Yellow # 6 0.00140625 FD&C Red # 40 0.00046875 Natural and Artificial Orange Flavor 1.400 Ascorbic Acid 0.300 Vitamin D3 Emulsion2 10.000 Treated Water' (for final blend) 833.33 "treated water" had had the cl.!ari"e and ~ rljusted to levels commonly used in the soft drink industry.
2 This emulsion is desc, ibed above with l~:gar~ls to batch 31.

W O96/31130 PCTrUS9G/~1C01 Bill of Mal~rials For Peach Fld~or~l Cever~2 (For 1000 KG of Geverdye) INGREDIENT AMOUNT. KG
Treated Water' (for beverage concenl, dlè) 137.42 rOtaCSi~ l~ Ben~Gdle 0.300 Sodium Citrate (dihydrate) 0.550 Citric Acid (anhydrous) 3.720 Lactic Acid (88%) 3.951 Aspa, ld" ,e 0.500 Calcium Glycerophosphdle 8.331 Mohawk Casing Color 0.001250 FD&C Yellow # 6 0.0008125 FD&C Red # 40 0.0004375 Natural and Artificial Peach Flavor 1.600 Ascorbic Acid 0.300 Vitamin D3 Emulsion2 10.000 Treated Water' (for final blend) 833.33 "treated water" has had the chlori"e and alkalinity adjusted to levels commonly used in the soft drink industry.
2 This emulsion is des~,iL,ed above with legalds to batch 31.

WO96/31130 PCTrUS9~'ClC01 Bill of Mdl~rials For Lemon Lime Flavored Ceve...~e (For 1000 KG of ~ve.~.~e) INGREDIENT AMOUNT. KG
Treated Water' (for beverage conce, 11l dla) 138.02 Bot~-cci~ Irn Benzoate 0.300 Sodium Citrate (dihydrate) 0.550 Citric Acid (anhydrous) 3.720 Lactic Acid (88%) 3.951 Aspartame 0.500 Calcium Glycerophosphate 8.331 Lemon Lime Color 0.000630 FD&C Yellow # 5 0.0005796 FD&C Green # 3 0.0000504 Natural and Artificial Lemon Lime Flavor . 1.000 Ascorbic Acid 0.300 Vitamin D3 Emulsion2 10.000 Treated Water' (for final blend) 833.33 "treated water" has had the chlorine and alkalinity ~ljnsted to levels commonly used in the soft drink industry.
2 This emulsion is des~,iL,ed above with l~ald:, to batch 31.

W O96/31130 PCTrUS9~/01G01 PREPARATION OF LIQUID BEVERAGE CONCENTRATE

The concel,l,dled mixture of i~y~dienl~ that make up the beverage is rer~r,t:d to as the beverage concer,l,dle. The liquid beverage concenl~le cor"~lises at least water a source of calcium vitamin D3 gum arabic and vegetable oil. rl~:ferdbly the beverage concenl,dle also cor",~rises vitamin C. If desired the beverage concentrate may also comprise: an acidulant preservative(s) andlor flavoring agent(s) and/or acid stable co!ori"g agent(s). Prototypes of the beverage of the present invention have a weight ratio of total acids to calcium of about 5.1. Prototype beverages of the present invention contained vitamin D3 at levels of about 1.45 x 10~ to about 1.75 x 10~% wlw and calcium at levels of about 1.46 x 10-' to about 1.47 x 10-' wlw.
In this example the liquid beverage concel,l,dle is pr~part:d in a single vessel at ar" -ienl temperature by dissolving the ingredients in water using a ~ sndi"y tank equipped with vigorous ayildliol, capability. A specihc order of addilion shown in Table 22 is followed to aid in disper~i"g the ingredients in an t rr,~ ient manner. Each ingredient should be completely dissolved before the next i"yl~d 3.,lis added.

1. Water 2. rolassium Benzoate 3. Sodium Citrate 4. Citric Acid 5. Lactic Acid 6. Aspall~",e 7. Calcium Glycerophosph~l~

8. Acid Stable Coloring Agent(s) 9. Natural and Artificial Flavor(s) Agent(s) 10. Ascorbic Acid 11. Vitamin D3 Emulsion (vitamin D3 + gum arabic) In commercial beverage manufacturing it is common for beverage concenl,~les to be prepared a day or more (often weeks or months) in advance of blending and filling conl~i"ers with the final product. For this reason the vitamin CA 022l7264 l997-l0-02 WO96/31130 . PCTrUS~'01C01 components may be added to the liquid beverage concenlldlt: just prior to blending with water to complete the beverage in order to prevent unnecessary long term exposure to air.

PREPARATION OF LIQUID BEVERAGE CONCENTRATE
Variations to the beverage concenl,dle manufacturing procedure des.;, iL,ed in EXAMPLE 1 can be made if available mixing vessel sizes are limited and no singlemixing vessel is able to contain the required volume of beverage concenl, dL~e.
Beverages accordi.~g to the prese, ll invention have been manufactured by preparing a plurality of beverage concel ,l, ale component slurries which were thereafter combined by pumping each beverage conce"l, dle componenl slurry to a larger sized tank. The water was divided equally between five clirrerenL beverage conce"l, dl~ component slurries all of which were cor,~ld,llly ~git~t~l A first beverage concel ,l~dle component slurry was made by first adding potassium ben,oale and then sodium citrate to the water. A second beverage concenl, dle cor"ponenl slurry was made by adding to the water in the following order: (a) citric acid; (b) lactic acid: (c) aspal ldr"e; (d) calcium glycerophosphale. A third beverage concenl~ dle cor"ponenl slurry was made by adding the acid stable ool~ori"g agent(s) and then the flavoring agent(s) to the water. A
fourth beverage concenll dle component slurry was made by adding the ascorbic acid to the water. A fifth beverage concenl, ale cor"ponenl slurry was made by adding the vitamin D3 emulsion to the water. The beverage concenl,dle cor"ponent slurries are transferred to a single larger sized vessel in the order in which they have beendescribed. The resultant blend (the beverage concenl, dle) in the larger sized vessel was vigorously agitated for not longer than about two minutes to homogeneously blend the beverage concenl,dlt: component slurries together. A liquid beverage concenl,dle in accordance with the invention should have a pH of 2.84.6 preferably 3.1-3.8. The pH of the prototype beverage concenl,dtes typically ranges from 3.1-3.8. If necessary additional lactic acid is used to adjust the pH of the beverage concentrate to this range.

W O96/31130 PCTrUS~ 1C01 PREPARATION OF CARBONATED BEVERAGE
Deareation and cooling inc, t:ases the beverage's carbonation efficiency and stability because the solubility of carbon dioxide in water is directly propo, lional to carbon dioxide pressure and inversely propo, lional to ter"pe, dLure. The extent of carbonation is expressed in terms of carbon dioxide gas volumes. The number of volumes can be deler",i"ed by cor"pari"g sample readings with carbon dioxide temperature/pressure r~ldLionship charts. Since pressure gauges measure the sum of pressures from all gases the presence of air in the carbor,aLed mix can cause errors in CO2 volume determination unless co~rections are made. A Zahm & Nagel air tester makes it possible to easily measure the pressure and air content of a sample. Tomake such a test the sample conLai"er is pierced :~'ICJ:;, ,9 head space gases to be ~ ~le3scd into a buret filled with 10-20% sodium or potassium hydroxide. The carbon dioxide is absorbed by the basic solution leaving only air inside the burette. The total pressure reading is then corrected for the amount of air ~ ser,L in the burette resulting in the corrected CO2 pressure. The gas volumes of the sample are then deLerl"i.,ed using the corrected pressure.
A beverage in acco,dancê with the invention may be carbonated by either blending the beverage concenL, dLe with cdrlJondLed water or blending the beverage concenL, dLe with water followed by carbonation of the blend. The prototype beverages were manufactured using a 5 to 1 ratio of beverage concenL, dLe manufactured according to Example 2 to non-carbonated water. CarbondLion levels in the ri"ished beverage may range from about 1.04.5 volumes of CO2, dependi, ,9 on flavor or desired sensory attributes. The product is then packaged and sealed in aluminum cans or tinted glass bottles. During the production of the prototype beverages separate in-stream lines of beverage concenL, dLe and water were cor, Ibi"ed in the proper ratio by a continuous metering device known in the art as a volumetric proportioner and then deaerated. The resulting mixture was L,dn:,r~r,~:d to a carbo-cooler where it was cooled and carbonated to approximately 2.5 volumes. The pH of the finished beverage should be in the range of about 3.14 and the pH of the prototypes was about 3.7. The finished product was then filled into standard 12 oz.
aluminum soda cans.
The nutritional profile and initial vitamin D3 Recoveries of the prototype low pH

~3-W O96/31130 PCTrUS96/04601 beverages in accordance with the invention are presented in Tables 23 and 24.

NUTRITIONAL PROFILE OF PROTOTYPE BEVERAGE
SERVING SIZE 1 CAN (355 mL) AMOUNT PER % Daily Value*
SERVING
Calories 0 Total Fat ~ 0 9 0%
Sodium 45 mg 2%
rOt~Ssil~m 25 mg 1%
Total Carbohydrate 0 9 0%
Protein 09 0%
Vitamin C 50% of RDI
Calcium 50% of RDI
Vitamin D 30% of RDI
* Not a siy"irica, IL source of other nutrients.
* Percent Daily Values are based on a 2,000 calorie diet.

VITAMIN D3 (IU/KG OF PRODUCT) (THEORETICAL FORTIFICATION A--810 IU/KG OF PRODUCT) FLAVOR 0-TIME % RECOVERY
Cherry 597 73.7 Lemon Lime 613 75.7 Peach 701 86.6 Orange 580 71.6 Average = 76.9% vitamin D3 Recovery W O96/31130 PCTrUS9'~ C01 CARBONATED BEVERAGE
An alternative embodiment of a liquid beverage concer,l,dle may be prepared according to Example 1 or Example 2 excluding any i"g,~dienl~ other than the water, calcium source, vitamin D3 and gum arabic (eg. the flavorant, and/or the cclDra,)l, and/or the sweetener may be omitted). This liquid beverage concenl, dla may then be combined with another liquid beverage concenl, dle, such as a commercial soda pop concentrate, and the resultant blended beverage concentrate may ll,erearler be combined with Cdl L ondled water, or combined with non-ca, L,onal~d water with the resultant beverage being carbonated in the ",anner desc, iL,ed above in Example 3.

NON-CARBONATED BEVERAGE
A liquid beverage concenl,ale may be prepared by blending a liquid beverage concentrate accor.li.,g to the prese,ll invention, such as described above in Exalllr'os 1 and 2, with non-carl,on-dled water. The resultant blend could then be placed into aluminum soda cans, or light reducing bottles, the head space flushed with nil~uyen gas or carbon dioxide to eli."i"dle oxygen which is harmful to vitamin and colorstability, and sealing the cans in the usual ",anner.

NON-CARBONATED BEVERAGE
An alternative embodiment of a liquid beverage concentrate may be pl~pdl~d accon " ,9 to Example 1 or Example 2 excluding any i"yl eu;enl~ other than the water, calcium source, vitamin D3 and gum arabic (eg. the flavorant, and/or ccloranl, and or sweetener could be omitted), and thereafter t' s ndi.,g the concentrate with fruit juice, vegetable juice, or any other suitable liquid matrix.

~5-W O96/31130 PCTrUS~ 1G01 POWDERED BEVERAGE CQNCENTRATE
The bill of materials for a powdered beverage concentrate in accordance with the invention is p,~ser,led in Table 25.

BILL OF MATERIALS FOR POWDERED BEVERAGE CONCENTRATE
INGREDIENT AMOUNT
Vitamin D3 Emulsion' 350 g Calcium Glyceruphosphdle 291.6 g Lactic Acid Powder (60% lactic acid) 181.3 9 Citric Acid 130.2 g Natural Cherry Flavor 42.0 g Sodium Citrate Dihydrate 19.3 g Aspd, ldr"e 17.5 9 Ascorbic Acid 10.5 9 ' This emulsion is des.i,ibed above with regards to batch 31.

A powdered beverage concer,l,~le was p,~:par~d by placing the calcium glycerophosphdl~ sodium citrate citric acid lactic acid and ascorbic acid into the chamber of an Aeromatic Top Agglomerator. The powder was then blended for two minutes under medium fll ~ tion. The ter"perdlure was brought to 70~C, the atomization was set at 1 Bar the alo",i~i"g nozle was placed at the highest level of three p~s- !e positions and the fan capacity was set initially at 12 (nominal setting).
Aspa, lar"e was dissolved in approximately 800 ml of warm tap water and a small amount of citric acid was added to achieve a pH of appro~i" ,al~ly 4. The vitamin D3 emulsion and the flavor system were blended by hand with the aspartame solution to yield approximately 1200 ml of liquid. The 1200 ml of liquid was placed on a stir plate and agitated under medium agitation while being sprayed onto the fluidizedpowder for approximately three hours.
As the liquid was sprayed the powder became heavy and it became necessary W O96/31130 PCTrUS96/016Ul to increase the fan capacity to maximum and place the alur,,i~i,,g nozle in the center position. Per actual analysis a Kg of powdered beverage concenl~le conl~i"ed about 83.5 9 of calcium 12.9 9 of vitamin C and 31 900 IU of vitamin D3.
The final powder pa, liul~s were relatively large and brittle and were pulverized before reconstituting with water. The powder was easily reconstitllted (see Example 8) and flavor was typical of a powdered beverage conce, Ill ~le product without thecarbonation. Longer shelf life in this kind of beverage concenl,dl~ is ar,lic;~ d because of the absence of water.

NON-CARBONATED BEVERAGE CONTAINING POWDERED BEVERAGE
CONCENTRATE
Approxi",alely 19.1 grams of the powdered beverage conce"l,dle manufactured in Example 7 were ~ solv0d in a sufficient amount of tap water to yield 1 Kg of beverage. A Kg of the resultant beverage is pr~jected to contain about 1.4 9 of calcium about 0.25 9 of vitamin C and about 607 IU of vitamin D3. As in the case of the liquid form of the powler~d beverage concenl, d~e the acid system can vary depending on the flavor s~l~.;te~l POWDERED BEVERAGE ADDITIVE
A powdered beverage additive may be manufactured by the process desc, iL ed in Example 7 containing at least vitamin D3 a calcium source and vitamin C but if desired omitting sweetener acids flavoring etc. The resultant powdered beverage additive could be added in applopriclle quanlilies to a liquid matrix such as a fruit juice blend offruit juices vegetable juices coffee tea oranys~ le beverage. The powdered beverage additive could be employed in bulk (eg. at an orange juice processing facility) or on a serving by serving basis when provided in single serving size packets.
It should be noted that if a liquid or powdered beverage concenl,~le or beverage additive accordMg to the invention is intended for use in a liquid matrix that may contain any dairy product (for example coffee or tea that may contain cream) a salt of ascor i~ acid should be used in place of ascorbic acid to prevent curdling of the W O96/31130 PCTrUS9G~'~1C01 dairy product.

CALCIUM SUPPLEMENT
A calcium glycerophosphale/vitamin D3/vitamin C tablet suFplen,er,l was pr~par~d by placing about 291.6 9 of calcium glycerophosphate and about 10.5 9 of ascorbic acid (vitamin C) into the chamber of an Aeromatic laboratory batch agglomerator. The powder was then blended for three minutes under medium ~git~tion. The temperature was brought to 55~C, the atomization was set to 1 bar, the ~lullli,illg nozle was placed at the highest of three po~~ ' !e posilions, and the fan capacity was set initially at 9 (nol"i"al setting).
The perial~lLic pump was set at 7 cc/minute and app,uxi,,,~lely 350 9 of vitaminD3 emulsion was sprayed onto the fll ~ ed powder. The commercially manufactured vitamin D3 emulsion desc, iL,ed above with respect to batch 31 was used in this calcium supplement. However, any sl~'-''e dry b'enc'-~'e source of vitamin D, preferablyvitamin D3 or D2, may be used for making a solid calcium suFp'en,er,L accordi.,g to the invention. As the liquid emulsion was sprayed, the powder became heavy and as powderfluM;--~Lio" was depr~ssed the fan speed was incrementally i"cl~ased to 12over 55 minutes to ",ai.,Lai" medium flll; ~ ion. Te",peraL.Ire was also increased to 60~C after 16 minutes. After all the vitamin D3 emulsion was sprayed on the powder, the heat was kept on and the powder was dried for three minutes. Per actual analysis, a Kg of powder for tableting contained about 139.9 9 of calcium, 26.4 9 of vitamin C, and 39,600 IU of vitamin D3.
The final powder particle was a soft agglomerate. No excipients were added to the powder to f~cilit~te the tableting prucess. Using a tablet die of approximately 1/2 inch diameter, 600 9 of the final powder was col"pn:ssed using a Carver model C
laboratory press and an applied load of 200 pounds force. The tablet was easily removed from the die. This process was repe~t~d using 1000 9 and 1500 9 of finalpowder to produce a total of three calcium supplement tablets, 600 9,1000 9, and1500 9, respectively.
A calcium suFplen,ent in solid form in accordance with the invention, comprising calcium glyceruphosphate, vitamin D, and vitamin C, is believed to beadvantageous over prior art calcium supplements because it provides a source of calcium that has a low aluminum content as well as providing vitamin D.

W O96/31130 PCTrUS9''01C01 APPENDIX A

Refer to Figs. 1-7, eyd, I ,9 the pe, ru" "ance of this vitamin D3 assay.
I. OVERVIEW OF THE METHOD
The low pH beverage, the vitamin D3 emulsion and the powder beverage are sapohi~ied with "lelllan-' - pot~c i~rn hydroxide to destroy the fat and releasethe vitamin D3 for exl, d~,lion. The Sdpol ,iried 5dl ", !e 5 are exl, dcled wlth an elllel/penld"e mixture and the extracts are e\dpoldled to dryness using nitrogenand, ecor,-~lil,Jted with iso-octane. Sample extracts are eluted on a cleanup HPLC column (cyanop,u~yl bonded silica), and column _~it~,h ,9 is used to transfer a "slice" of the eluant cor ' ,' ,g vitamin D3 onto an ad.liliunal HPLCanalytical column (al " ,op,u~,yl bonded silica) for final qud"litdlion. The vitamin D3 peak in the sample is qud"lildled using a linear rey,ession external ~Idnddldcurve.
Il. APPARATUS
A. General Apparatus 1. Centrifuge tube glass, 50 ml with teflon-lined screw cap (Corex 8422A).
2. Centrifuge tube glass, 50 ml - conical (Kimax 45176).
3. Centrifuge (IEC Model Centra-HN or equivalent).
4. Water bath -capable of 40(i2)~C and 75(+2)~C.
5. Source of nitrogen (purity ~99.7%) - for evdpo,dlions.
6. Vortex mixer - S/P Mdyl leslil or equivalent.
7. Volumetric flasks - 100 ml, 500 ml.
8. Volumetric pipets - I ml, 2 ml, 3 ml, 5 ml, 7 ml, 15 ml, 30 ml.
9. Repedli"g pipet - "Tilt-a-Pet"
2-25 ml heads (VWR - 53481406) for ell,e,/l,er,ldi,e 2-1000 ml C, len" ,eyer flask reservoirs - size 24/40.
10. Repipet D;_pense, :, - Baxter-P4985 or equivalent -1 ml for KCL solution (Baxter P4985-5) - 5 ml for acelo";t, i!e (Baxter P4985-10) - 6 ml for " ,ell ,anol (Baxter P4985-10).
11. Oxford Macro-Set Pipetter (Baxter - P5079-2, or equiv; Qty =2) 1 - for sample transfer 1 - 4 ml for 45% KOH.

12. Therm-O-Vac - size 14/20 (Cole-Parmer #N-06140-15).

13. Teflon sleeves - sizes 24/40 (Cole-Parmer#N-06139-15).

14. Evapo-Rac E\ dpordlur for 30 mm tubes (Cole-Parmer #N-01 61 0-35).

15. Centrifuge tube rack (Cole-Parmer #N-06737-40).

16. Cooling tray large enough to accG"""ocldle centrifuge tube rack (#N 06737-40).

17. HPLC tubing - 0.040" sla;.,less steel - 2 feet.
~ 18. Caldnces - (a) MettlerAT200 (orequivalent) n ~d ' IEtoat least 0.01 mg (for standards, vitamin D3 emulsion and powdered product.
(b) Mettler PM460 (or equivalent), ~ ~ ~' ~ ' 'e to at least 0.001 9 (for low pH beverage Sdll, 1~S).
19. Glass Stirring Rods.
20. Magnestir Stir Plate - Lab Line #1250 or equivalent.

W 096/31130 PCTrUS9~/01CO1 21. Teflon Coated Stir Bars - 2" length.
22. Beakers - 600 ml 800 ml 1000 ml.
23. C ~ r - Hewlett Packard-11 C or equivalent.
24. R~rl igerdlùr (freezer cor,lpdl l~ hl optional) for storage of :~Ldnddrd:~
at 4(+4)~C.
25. Lighting Peq~ ",el,b Ultra-violet shields - F40T12 - Dayton Plastics Inc. for white fluo,~scenL bulbs.
26. Scoop - 1/8 ~ pOOIl.
B. HPLC Instru, - ,enldliun 1. Columns: Guard (4.6 x 30 mm) Cyano - Rainin - Cat. #CS-GU
Cartridge holder- Rainin - Cat. #140-200.
Cleanup - Chromegabond Cyano (4.6 x 250 mm 311 60 Ang~l,u,,,s) -ES Industries.
Analytical - Hypersil APS ll (4.6 x 250 mm 3 1~, 120 Anstroms) -Keystone.
2. Pumps: Two consld"l flow pumps capable of ope, dli"g at 5 ml/min and up to 6000 psi (Beckman 110B with pulse dd""~ener or equivalent).
3. Dt~ .lul:.: Cleanupsystem-fixedorvariable wa I l~hyll~
capable of ",onilc,ri"g 254 nm or 264 nm ( Waters 440 or equivalent).
Analytical system - Variable wa~/~lehylll detector capable of ",on . i"g at 264 nm ~ 0.0025 AUFS. Under normal ope,dli"g col, ~s the short term noise should be less than 3% of the 5T vitamin D3 :.Ldnddl d peak height (Waters 486 or equivalent).
4. Injector: Alcott/Mi~;,ul"e,ili~ s 728 orequivalent.
5. Column Oven: Capable of 35~C - 100~C and + 1.0~C settings andaccuracy. Storage for 2 x 250 mm columns and one 30 mm guard column.
6. SY.jtCI ,9 Valve: HPLC column s~. it~;l, ,9 valve with at least 6 ports. Has a working range up to 6000 psi (Mi~ilu,,,eriLi~;~ 732 or equivalent).
7. Recorder: One 10 mV ~bcordi"g device for the cleanup HPLC output and either a ,~co,der or an illL~:yldLul for the analytical HPLC system. A data system capable of l"oniLori"g acquiring and ,~p,ucessi"g two chal-nei~ of data is strongly ,~cor"",ended.
8. Solvent Reservoir: 10 Liter - Common to both cleanup and analytical HPLC systems (VWR #KT953901-W O96/31130 PCTrUS9G/~SC01 1003 or equivalent) Ill. REAGENTS
A. Standard Rer~l~nce Material - Vitamin D3).
1. USP ,~rt~ nce sld"dd,d #1310 (Ch~ rul = vitamin D3).
Consult current USP literature for the current lot number. Potency =
40 000 IU per mg. Store at 2~C to 8~C. Care must be used in opening the sealed ampules to avoid introducing glass r,dy",~"L~ into the !~ldnddl d. Vitamin D3 must be used from an open ampule i"""ed;at~ ly and di~.cd,ded.
B. Chemic31s 1. Amyl Alcohol Analytical Reagent r~co"""end M " Icklud UN 1987.
2. 1\1e;1ldl1ol HPLC Grade ~ cor"",end Burdick & Jackson #230.
3. Iso-octaneHPLC Grade ~ecor"",and Burdick & Jackson #362.
4. PentaneHPLC Grade ,t:co",lllend Burdick & Jackson #312.
5. Diethyl Ether Anhydrous ,~co"""end M " luhludL UN 1155 6. Potassium Hydroxide 45% solution ~ecor"",end Baker#3143-03.
7. Sodium Ascorl,dL~ RecGr"" ,end Aldrich #26 855-0.
8. Acelor,- ;'~ HPLC Grade ,eco"""end Burdick & Jackson #015.
9. Ch' ~ ~Orul,,, HPLC Grade , ~co" " "end Burdick & Jackson #048 10. Potassium Chloride Reco"""end Mal'- lulhudl#6838-500~NY.
11. n-Butyl Chloride HPLC Grade rt:cGIllll~and Burdick & Jackson #034.
C. Solutions 1. HPLC Mobile Phase Vol~""~L, - -'Iy pipet 40(_0.1) ml of n-butyl chloride 20(_0.1) ml of amyl alcohol and 10(_0.1 ) ml of ch' o~. ", into 4000 ml of iso-octane.
Mix well. Make four liters at a time - roughly equivalent to 1.0% n-butyl chloride + 0.5% amyl alcohol + 0.25% ch'~rurur,,, in iso-octane.
Use for both cleanup and analytical HPLC systems. Cor" 'e: ~y fill the 10 liter reservoir prior to each day's analysis.
2. Exll d~ liun Solutions #1 - 20:80 ell,er/~nldi-e: Mix 200 ml of diethyl ether with 800 ml of pentane. This is sufficient for up to 20 Sdll, es (2 X 25 ml per sample is required). Prepare fresh daily.
#2 - 33:67 t:ll ,erl,l enldne: Mix 250 ml of diethyl ether with 500 ml of pentane. This is sufficient for 28 Sdll, !e s (25 ml per sample is required). Prepare fresh daily.
3. KCL Solution Prepare a 10% KCL solution using distilled water. Mix 50 9 of pOpc~illrn chloride with distilled water and dilute to a 500 ml liter volume. Store at room ~el,,perdL.Ire and e. dlion date is one month from date pr~pd,~d if kept tightly capped.
D. Pl~pdldLion of Vitamin D3 Stdl~ddl~ls NOTE: Work under UV-shielded, whife, fluorescent bulbs wifh the ulfra-violef shields described in Secfion 11.25 if possible. If unprotected white W O96131130 PCTrUS~6/01C01 lights are used, extra precautions must be taken to keep an solutions of Vitamin D3 protected from light by covering containers with aluminum foil or by using amber low-actinic yld s~ .~ c.re. Sl~ndah~:, are also heat sensitive and should only be briefly removed from the refrigerator for immediate use.
NOTE: Due to the ",ell,o.l'~- suscel ' ' ' 'y to low level conldn ,a~ , an volumetric flasks must be rinsed with iso-octane prior to preparation of Vitamin D3 standards.
1. Stock Standard (AppluAilll.~t~ly 1,920 lU/ml) Weigh 24(+1) mg of vitamin D3 into a 500 ml volumetric flask.
Dissolve and bring to volume using iso-octane. F-, - dlion date is two weeks and it must always be stored at 2~C to 8~C when not being used to prepare the i"lt:r",edidl~ :~ldl1ddl-1.
2. I"L~""edi~le Standard - ISTD (AppruAi,,, ~t~ ly 27 lU/ml) Pipet 7.0 ml of stock :~ldnddld into a 500 ml volumetric flask. Dilute to 500 ml with iso-octane. ': ~: dlion date is 10 hours for plepdldliùn of the working :~ldnddld~, but can be used for 2 months (at room l~l I "~erdlure) for 1~ - h ,9 the, el~l ,lion times on the cleanup HPLC system.
3. WorkingSI~nddld~-3T 5T 15T 30T
lion is 1 week. Store at 2~C - 8~C).
3T = Pipet 3.0 ml of ISTD into a 100 ml volumetric flask and dilute to volume with iso-octane (d~J~JIuAilll~ly 0.8 lU/ml).
5T - Pipet 5.0 ml of ISTD into a 100 ml volumetric flask and dilute to volume with iso-octane (app,uAilll..~ly 1.3 lU/ml).
15T = Pipet 15.0 ml of ISTD into a 100 ml volumetric flask and dilute to volume with iso-octane (app~uAill..~t~ly 4.0 IU/ml).
30T = Pipet 30.0 ml of ISTD into a 100 ml volumetric flask and dilute to volume with iso-octane (a~u~JluAilll._~ly 8.0 lU/ml).
IV. PROCEDURE
Sample Pl erjd, dlion 1.a. Low pH Beverage (300 IU/KG - 900 IU/KG) Accurately weigh (to the nearest 0.001 g) 12.5 9 of the low pH
beverage into a 50 ml centrifuge tube (Corex #8422A) and proceed to #2 in the Saponir,. dliun Section.
b. Vitamin D3 Emulsion (-100 00 IU/KG) Accurately weigh (to the nearest 0.0001 9) 0.19 of the bulk emulsion into a 50 ml centrifuge tube (Corex #8422A). Add 10 ml of d i~ - I/de.or,i,e water. Proceed to #2 in the Saponirudlion Section.
c. Powder Product (-35 000 IU/KG) Accurately weight (to the nearest 0.00019) 0.39 of the powder product into a 50 ml centrifuge tube (Corex #8422A). Add 10 ml of distilled/deiol ,i~e water. Proceed to #2 in the Saponircalion Section.
2. Add about 0.4(~0.1) 9 of sodium ascorl,dL~ (1/8 level teaspoon) and vortex 10 seconds.
NOTE: Sfart with a low vortex speed and increase vortexing speed with each sLIcces.civc step (i.e., affer the addition of methanol, and again affer the addition of 45% KOH).
3. Add 6(+0.3) ml of methanol and i" ""edicl~ly vortex for 15 seconds.

4. Add 4(iO.3) ml of 45% potassium hydroxide solution tightly cap the tube, and i"""edi~t~ly vortex for 20 seconds.
5. Place the tubes in a p,t:hedL~:d water bath at 75(i2)~C for 30 minutes. The tubes should be vortexed for 5 seconds at the 10 and 20 minute intervals.
6. After 30 minutes remove the tubes from the water bath and place in ice water for a minimum of 30 minutes to bring them rapidly to room l~r"~.e,dLure.
EAII d~il~n 7. Add 5(iO.3) ml of ac~:Lunil,.' to each tube, cap and vortex at a " ,oderdlt: speed for 5 seconds.
8. Add 25(i1) ml of 20% ether/80% pentane mixture and shake in a wide ser, ' ' cular arc across the front of the body 20 times.
Invert the tubes with each stroke.
9. Briefly centrifuge at ",ode~dL~: speed (app,uAi",~ ly 300 x G for 1 minute) to COI" ' : layer sepd,dLion.
10. Draw off the clear ether/,~,enldi)e using the si~hor, ,9 apparatus and vacuum. (See Figure 1). Transfer the top layer to a 50 ml conical centrifuge tube (Kimax #45176) leaving behind 4 to 8 ",'"' "ele,:, of the ell,er/,~,enldne mixture. Avoid transfer of any middle layer or aqueous (bottom) layer.
NOTE: If any of the middle or aqueoLIs bottom /ayer is accidentally l,dn~-r~"~d, the sample must be discarded and the assay r~,pe~tf~d 11. To avoid sample to sample cGr,ld",' ,alion rinse the si~,hon' ,9 apparatus with 5 - 7 ml of pentane and add this rinse to the dm,~
12. Evaporate the L~dnsrt~ d ~II,er/per,ldne layer in the warm wate bath (40 i4~C) with nitrogen to about 2 ml to allow for e~ nal l,dnsre, ,. (See Figure 2 for the Evapo-Rac e~,dpordlion apparatus.) 13. Repeat the ~Xtl d~liol- in steps #8 - #12 once cor, ' :. ,' ,9 the extracts in the same 50 ml conical centrifuge tube.
NOTE: Be careful not to overflow ~he 50 ml Corex centrifuge tubes with the 2~ ml eAl,dclion solutions (#1 or#2) during the 2nd and 3rd ~l,d~,lions. Should this occur, the sample mustbe discarded and the assay repeated.
14. For the third eAIId~;lion, follow steps #8 - #12 using 25 ml of 33%
ether/67% pentane solution (not the 20% ether/80% pentane solution) 15. Evaporate the co",' ..,ed ~AIId.;lions to dryness. Remove the centrifuge tubes from the water bath as soon as e\,dpordlion is col " !~: The extract should appear clear or as a white or slightly yellow film. Make sure that the extract is cor" ' ' 'y dried before ~ ~con:,lilution. The tubes may have to be gently tapped to CGI I l, ' ~ ' the e~/apordlion.
16. I~ edidhly r~con:,lilute with 2.0 (iO.006) ml of iso-octane with a class A volumetric pipet. Be careful to thoroughly rinse down the walls of the tube. The tube should be tightly capped to prevent e\,dpoldLion and vortexed 5 seconds to mix.
17. Finally, add 1 ml of the KCL solution to each sample and touch to the vortexer briefly to mix. Tightly cap and centrifuge at moderate speed (app,uAi",c~ly 300 X G) for 1 minute to con, '~t phase sepa,dlion. If using a centrifuge equipped with a s~ ;"g; ,g bucket rotor place the centrifuge tubes on the outside perimeter of the rotor. This is to CA 022l7264 l997-l0-02 WO96/31130 PCTÇUS9''01C01 prevent the conical tubes from brt:al~i"g. Transfer only the top layer to a vial and tightly cap. Be careful ~ to transfer any of the saturated KCL solution.
NOTE: The sample exfract must be analyzed within 24 hours from fime of p,t:pa,dlion. If fhe HPLC system encounters problems, the aufc san, '~r vial should be immediately stored below 8~C affer pr~:pardlion for up to 48 hours. No sample can be reinjected affer an aborted HPLC analysis if it was leff in the auLusai"N~r at room temperature ovemight.

18. inject onto the eq~ b.dlt:d HPLC system (section V).

V. HPLC CONDITIONS
A. Cleanup HPLC System - See Figure 3 for configuration.
1. Column: Ch,u,,,egabond Cyano (4.6x250 mm 31~) with CS-GU
guard column (4.6X30 mm).
2. Eluant: 1.0%1-chlorobutane + 0.5% amyl alcohol + 0.25%
cl ,'c uror' ' H n iso-octane.
3. Run Time: Slice dele", laLion =app~u~illl~tuly20 minutes.
4. Flow Rate: 1.5 ml/min.
5. Injection Volume: 25û~1.
6. Column Heater: 40(_1)~C.
7. Detector: 254 nm or 264 nm.
8. Recorder~ y, dLùr or data system (pl t:rel l ~d).
9. Column Actuated by timed control from injection point to Switch: c - n. Slice time window should be no greater than 1.0 minute (with 0.1 minute accuracy) for c~ n of vitamin D3.

B. Analytical HPLC System - Figure 3 for configuration.
1. Column: Hypersil APS ll (4.6X250 mm, 3~).
2. Eluant: 1.0%1-chlorobutane + 0.5% amyl alcohol + 0.25%
ch'~ uru"" in iso-octane.
3. Run Time: Approxi",at~ly 35 minutes.
4. Flow Rate: 1.5 mVmin.
5. Column 40(_1)~C.
Heater:
6. D~L~uLion: 264 nm ~0.0025 AUFS, (Waters 486).
7. Recorde!: Reco"""end the use of an i"Lt:y,dLur or data system for reprocessing.
8. Frll ~ b.dLt: the columns and obtain a stable te s ~ ,e. Inject the i"L~---,edidLt: ~Ld"cla,d ISTD (no column switch) at least 3 times until a consi:,L~r,L r~tenLion time (r~:LenLion time +0.02 minutes) is e~ - -ed on the cleanup HPLC (Figures 4 & 5). Always verify the cleanup HPLC ,~ :L~r,Lion time within 1/2 hour before analysis of :,Ldnddl ds or Sdl I, ' C The run time is appruxi" ,~ly 2û
minutes however the time required to eq~ b.dL~ the columns with fresh eluant is dppl u~ tuly 2 hours.
9. After d~Lel " , ,g the ~~l~"Lion time of the il lL~I " ,edidl~ Ldn-.ldl d ISTD
on the cleanup column set the slice window (i.e. transfer of vitamin D3 from the cleanup column to the analytical column). This is done by CA 022l7264 l997-l0-02 W O96/31130 PCTrUS96/O~G01 setting the s~r ;~ h ,9 valve to switch the vitamin D3 from the cleanup column to the analytical column at 0.10 minutes before the vitamin D3 first elutes from the cleanup column until 0.10 minutes after the vitamin D3 peak returns to '~ c s -' Ie on the cleanup column. See Figures 4 and 5. Slice window times should not exceed 1.0 minute -using a minimum among of time (genel 'Iy 0.8 - 1.0 min.) necessaly to collect all the vitamin D3 while preventing the transfer of any other illl~ re~ g cGnll)onent~.
See Figures 6 & 7 for the cleanup and analytical HPLC
ClllUllldlUyldlllS of a 15T working :~Ldl~ddll~.
Vl . HPLC ANALYSIS
A. Upon verifying eqll ' b.dLion of the HPLC system and e -~ 1 '' lg the - " n window, inject three (or four) working :~Ldnddlds (3T, 5T, 15T, 30T) and then the sample extracts. The three (or four) working sLdl Idards should be injected once again at the end of the run.
Only single il l, - ns of each sample are required.
Vll. CALCULATIONS (Use only peak heights for reporting purposes) Note: Peak height is required for quantitafion as small amounts of ~ass ' ,e noise can cause large area differences.
A. C ' llation of Working Standard Concel lL~ dLions 1. C-'~"'-l~ the concel lLl dLion of the vitamin D3 in working :,Ldn.ldl d~
3T, 5T, 15T, 30T from the F~ ;. lg equation:
lU/ml = (W) (P) (7) (PV) = (W) (PV) (0.0112) (500) (500) (100) where: W = weight of vitamin D3 ~Ldnddl .1 in mg.
P = 40,000 lU/mg for vitamin D3 PV= final pipet volume for working ~Ldnddld~. ~
= 3for3T
5 for 5T.
= 15for15T.
30 for 30T.
Example: for a 5T sldnddl d p, t:pdlt:d from a stock solution that con ~ Ied 24.00 mg vitamin D3, the concerlLIdLion is c?lclll-~-d as follows:
lU/ml = (24.0ûi r4Q.Q00) (7) (5) = 1.3440 lU/ml (500) (500) (1QQ) - B. Ca' ll~tion of the Standard Curve Using Linear Reyl~ssion and the Qudl,LiLdLion of Vitamin D~in Samples 1. The peak heights of each It:spe-.Li~/e level of the working :,Ldnddl d are averaged. A lineam~yl~:ssion line is c ~cll~-~od by using the average peak heights (y-axis) and the concenLIdlion (x-axis) forthe It:s,cecLi~e working :~ldllddr~J.
Example: A lineam~:ylt:ssion line (vitamin D3 peak heights versus concel,L,dLiol-) for264 nm channel is pl~serlLed below. Two ill,- -ns (b~g l n lg and end of run) were made per each level of working :,Ldndal d.

WO96/31130 PCT/U~ 1C01 Working ~ ~onc. ~ lo. Avg.,Peak IWl~nl i l~ ~ject Heigh t 5T 1.3440 2 2.6396 15T 4.0320 2 8.0789 30T 8.0640 2 16.5839 Slope = 2.07775 y-i, lk~ JL = -0.20754 Corr. Coef. = 0.99994 2. The 5dlll les should be qua"LildLed by L,ldul~ 9 the :-ld"darda around the 5dl l l 1 ?~
C. Low cH:BeveragQ.and Vitamin D3 Emulsion. and Powder Product C~ tinn 1. Per Weight Basis - IU/Kg:
Vitamin D3 (lU/kg) = ~C) (V) (1000)#
'(S) (X) where: C = Vitamin concellLIdLion (lU/ml) from sLdllddld curve.
V = Volume (ml) of iso-octane to l~con:,LiLute extracts.
1000 = converts grams to k- -_ dlll~
S = Sample size in grams.
X = 0.86 for 75~C sa~.or,iricdLion factor for thermal isomel i~dLion of vitamin D3 to previtamin D3.
# = Sl Ihstitl ltp 100 for 1000 to convert to lU/100g.
Example: A 12.533 9 (S) low pH beverage sample was ~e- on:,LiLuted in 2 ml (V) of iso-octane which gene,dL~d a peak height of 6.2330. The co"~:,por,di"g vitamin D3 concer,L,dLiol1 (C) cL ~ Ied from the previously c~cl~ -d ~Ldncldl.J curve was 3.0998 lU/ml. The final concerlLIdLioll would be ~ t-d in the / i.lg manner:
Vitamin D3 (lU/kg) = (3.0998) (2) (1000) = 575 lU/kg (12-533) (0.86) The Simultaneous Determination of Calcium (Ca) in a Low pH Beverage by ICP-AES Using a High Solids Nebulizer A. THEORY
1. Inductively coupled plasma atomic e~ sion spe- LIu,ll~L,y (ICP-AES) is an atomic speuLluscopic technique that has several advd"Ldges CGI l lpdl ~d to atomic aL,sol ~,Lion: P'~ ,L cleL~uLion limits a broad CA 022l7264 l997-l0-02 W O96/31130 PCTrUS3~'01C01 linear dlion range of over four orders of magnitude for most elements, minimal il,telr~:r~nces and the ability to del~rll, ,e several ele. "er,L~ in the sample simultaneously under one set of operdLi"g co~ s. These advdllLdges lldllsldLt: into less sample pl~pdldlion, ~ I b. dtion and analysis time for the analyst.
2. The ICP-AES instrument consists of three cG"")onellL~. sample introduction device, Iul- hbux and spe- I,ur"eL~r. Most cor"")only Sdl l l !es are introduced in the form of solutions which are neh~ ~d (broken into tiny droplets) and passed into the torch with a stream of argon. In the lur.:l)bux 1-2 kW of radio-frequency power is coupled from a copper coil (inducfor) into a small region inside a quarlz tube (torch), through which argon flows. The power density in this region is high enough to heat the argon until it ionizes and, since the region is at dll "osphe, ic pressure, there are sufficient cc ~ens with other argon atoms to instantly ignite a plasma with a temperature of about 10,000 K.
3. The ".i- ur,l~L~l-sized droplets from the nebulizer enter the bottom of the torch and pass through the cooler (6000 K), darker central region of the plasma called the axial channel. Here water is evd~JordLed, and the l~rr , ,9 dry pdlLicles of analyte are vdpo,i~ed and dLu",i~ed (",-'e ~ ~les broken down into atoms) by the heat of the plasma in just a few 1" o - e nds. FYcit-~'ion and ioni~dLiol~ of the outer ele~ LIuus of the atoms occurs; the intensity of the e r";~s;on that results from the dePYici~ on of these atoms and ions is p, upu, lional to the conce"l, dlion of analyte in the original solution. Thus, ~ b. dlion consists of measuring the intensity of analyte er"-s n for slandd,ds of known concenL,dlion.

4. Light emitted by the ICP is ~ C'lf ~ d by a lens in the s~.e..l,u",eLer and focused onto a dirr,d-.lion grating which di~pe,~es the light into its cor"l,onenL wav~l~nyll ,:,. The emitted I ddidliun u a~elenyl h resolved from all the analyte ele."e"l:. is c-- sr: d sim~ neously by several del~ulul ~ placed in front of the grating and converted into an ele~ Irical signal. A data system relates these signals to the concenl,dtions of the ~le."enl:, in the sldnddld:~ and ~ 5 the analyte conce"l,dlion in the Sdll, ' S
5. The particular instrument used in this method features a movable ehl,dnce slit conl,.- ed by a high ,~:s ~ tion stepper motor called SAMI (Scanning Accessory for Multielement Instru",er,ldlion).
Moving the entrance slit slightly changes the angle of inc;dence upon the grating, and slightly changes the wavulel,yllls incident upon the exit slits. This feature allows the user to perform bachyluund co"~ulion in the sample matrix by sLIL,lldulill9 the e",;~sion bachyl ound just off the peak center.
6. This method employs a speedy dilution pr~pd, dlion of Sdl l e S with a ~ surfactant and dilute acid. A special kind of nebulizer, called a maximum diasolvcd solids n-h~ : (MDSN) or high-solids nebulizer is required to provide long term operdlion without clogy; ,9. Rec~use the viscosity of :,ldnddl.ls and Sdll, les iS quite dirr~ r~"t an internal WO96/31130 PCTrUS5f'~C01 aLdnddld must be used to co",~ensdle for the poorer neb~ tion erri..;en-_y of the high solids Sdl ", !e ~ Cobalt is added to each sLdnda,d so that they are exactly 20.0 mg/L Co. C; ' b.dLion consists of measuring the analyte/Co ratio in the standards as a function of analyte concenLIdLion. An exact quantity of cobalt is added to each sample so that if they were diluted to 50.0 mL, their cobalt conce"L,dLion would also be 20.0 mg/L. Note, however, that the analyte/intemal sLdndd, ~J ratio in the Sdl,, 'e s will not change with the total volume, and so volumetric ware is not necessaly for the sample pl epal dLion. When the SOr - a. e asks for the "sample volume" to A a dilution factor, the analyst should enter 50 mL, the volume that would make the conce,lLIdLioll of cobalt in the Sdl~, ' s equal to that in the aLdnddlda.

B. MATERIALS
1. Instrument a. Inductively Coupled Argon Plasma C., I;;,aiOn Spe-_L,u"leLer, ARL
Model 3560 or Accuris b. Ryton V-groove n~bu . ARL#173259-0000 or Precision Glass #510-50 only c. Spray chd"lLe,. ARL#173142-0003 or Precision Glass #110-34 or equivalent d. ICP torch: ARL#1390û9-0003 or Precision Glass #100-05 or equivalent 2. General Labu, dLùly Ftll 'i, l~enLlFc fi ' ' ~ s a. Analytical balance b. Fume hood c. Di "c ~ ~'e, flat-buLLur,,ed, 50 mL plastic centrifuge tubes with caps (Baxter C3902-14 or equivalent) d. Plastic coated rack suitable for holding many centrifuge tubes e. 125 mL, 250 mL, and 1 L plastic bottles for storing sLdndd,da.
polymethyl~.~, llene (PMP) or equivalent f. D;spc--'-'- plastic transfer pipets-3.5 mL capacity 9. E ppendo, r pipet or equivalent, 1000 I~L capacity with tips h. 50 mL 1~ r:. ~t~rorequivalent i. Plastic di_penser bottle (PMP or equivalent) fitted with a Teflon-constructed di_penser top with arlju-' ' ' volume between 1-10mL; dispenser may be fitted to concer,L,dLed HCI bottle directly j. Magnetic stir plate and Teflon coated l"ay"etic stir bars k. 1 L and 250 ml volumetric flasks: glass or plastic (PMP or equivalent) I. Class A volumetric flasks: 2,4,5,10,15,20,25,40,50 mL
m. Options: 1 mL digital pipet with tips, Rainin EDP-Plus or 1 mL, glass volumetric pipet or equivalent 3. Chemicals/SLdndd,-ls Unless otherwise noted, the r 'lO.~ g chemicals should be stored at room temperature. Their expiration date is one year affer the date they are frst opened. Upon ~Yp.. diiun the chemicals must be either discarded or re-W O96/31130 PCTrUS9-'01C01 evaluated.
a. High purity stock sLdrlddld solutions (NIST or NlST-L,~ e) 10 000 mg/L Ca 10 000 mg/L Co 1000 mg/L Co. These stock aLdnddl d solutions expire on the date given by the manufacturer.
b. Hyd,u..l1 ~ric acid J.T. Baker BlA-grade or equivalent c. Triton X-100 Kodak scintillation-grade or equivalent d. Argon gas minimum 99.996% purity e. High purity water ~ 'i, - ~-treated or equivalent c. iNSTRUMENT~t OPERATIN~ ~ON~ITl~)NS
1. The -~ lenyll,s that have been used are listed in the table below.
The instrument should be installed with identical chdnnt l. if p~ s ~ !~
be~o~ ~se the sensitivity of the line and the pc s ~y of i"l~:~ re(ences can change if a different line is e", '~,/cd for analysis.
ELEMENT WAVELENGTH (nm) TYPE ORDER
Ca 317.93 ion 2 2. Typical ranges of operdLi"g conclilions for the ARL 3560 are listed below.
a. Incident power: 1200-1400 watts b. ~fl~ct~d power: ~5 watts c. Snout argon gas flow: on d. Coolantargon pressure: 30-40 psi e. Plasmaargon pressure: 20-30 psi f. Nebulizer argon pressure: 30-46 psi (-0.6-0.7 Uminute if a mass flow conl,. -. r or other type of flu~nlelt:r is used to regulate flow) 9. Pe,: pump flow rate: dial setting which cor,~aponds to -2.5 mUmin. (dependa on make and model of pump) h. Fe~ i ~ pump tubing: 1.12 mm l.D. red/red P.V.C.
1\1_.ul~ne orequivalent 3. Software pdl dl I I~L~:ra. These are stored in the TASK files which perform e ~.dLion and sample measurement and must not be altered.
a) InLeyldlions (on-peak): three 5 second i"l~:y,dLions per sample b) InLt:gldLions (off-peak): two 5 second illLeyldLions taken at-80 SAMI units off peak center D. STANDARD AND SOLUTION PREPARATION
Store at room temperature unless ull,en~;se noted.
1. 2% HCI rinse: Mix conce"L,dL~d HCI with high purity water in the appruAi",dle ratio of 20 mL acid to 1000 mL total volume of solution.
Use a plastic COI, ,er of a size app,up,idL~ to the volume of solution p,~pdlt:d. For exd", !e to prepare 20 L of 2% HCI fill a 21 L carboy with high purity water to the 20 L mark and add 400 mL HCI to the water. F-, ..dLion: 6 months.
2. Triton X-100 solution (appru,~i" Id~ly 5%): Add about 700 mL high purity water to a 1 L plastic bottle con ,i"g a Teflon-coated stirring bar. Place the bottle on a ",ay"etic stirrer and begin stirring at a W O96/31130 PCT~US9''01~01 Illode~dl~ speed. Slowly add 50 mL Triton X-100 from a graduated cylinder. vVhen the Triton is .li;.:.olved, fill the boKle dppl U~il ,..Jtuly 1000 mL with high purity water. Transfer to 1 L plastic boKle fiKed with a Teflon-constructed .li~panser with - 5 ~-'-' ' -. volume from 1-10 mL.
F- dLi~n: 6 months.
3. Tld.litional :~Ldnddld solution pl~pdldliun Prepare 1 literofthe d,u~luplidl~ sldnddld solution. Add the i" "- ' ' amount of 10,000 or 1000 mg/L stock :~Ldllddld solution to a 1 L volumetric flask using a Class A pipet. Then add app,u,~i",~.~ly 800 mL of high purity water, 2.00 mL of 10,000 mg/L cobalt internal :,Ldl Iddl d (Class A pipet), and 20 mL (r~ F ~ or di~",enser) of hyd, u~ loric acid to each flask.
Add 50 mL of Triton X-100 solution from the di "~enser to each flask and then fill the flasks to volume with high purity water, slowly to avoid forming suds. Agitate well and transfer to clean, dry 1 liter storage boKles. Dispense as needed into 125 mL storage boKles to use at the instrument. F-, dlion: 6 months.
4. Standard blank solution: Prepare 1 liter of a sldndard blank at the same time, and from the same reagent batches, as the above ~ldnddl d:~. Add 800 mL high purity water to a 1 L volumetric flask;
2.00 mL of 10,000 mg/L cobalt internal sldnddld (Class A pipet), and 20 mL (~F, ~$ 3r) of hy.l,uch'-ri., acid to the flask. Add 50 mL of Triton X-100 solution from the di~."enser to the flask and then fill the flask to volume with high purity water, slowly to avoid forming suds.
Agitate well and transfer to a clean, dry 1 liter storage boKle.
Dispense as needed into 125 mL storage boKles to use at the instrument. ':., dliOI) 6 months.
5. Intemal ~Ldllddld ~t:r~,~nce blank solution (ISRB): Prepare 1 liter of an ISRB at the same time, and from the same reagent batches, as the above "ldnddl d~. Add 800 mL high purity water to a 1 L
volumetric flask and 20 mL (l ~ F i~ ctt~r) of hy.l, u.,L' - ric acid to the flask. Add 50 mL of Triton X-100 solution from the di~penser to the flask and then fill the flask to volume with high purity water, slowly to avoid forming suds. Agitate well and transfer to a clean, dry 1 liter storage boKle. Dispense as needed into 125 mL storage boKles to use at the instrument. F- ~ dLiOn 6 months. Note that the ISRB
does not contain cobalt, but the :~Ldl Iddld blank does. The ISRB is analyzed before any :~Ldl Iddl ~ or S'dl I, !e i, the purpose is to subtract the intensity of analytes found in the ,~agenl:, (Triton X-100 solution, HCI, and water) from the analyte i"l~nsilies found in the sldllddld:~
and Sdll, E. PROCEDURE
1. Standard I la, l'" lg a. All boKles used for storage of standard solutions must first be soaked in 10% (v/v) HCI for a minimum of three hours, followed by multiple rinses with high purity water. Air dry or rinse several times with the :~Ldl)ddld. When reusing the boKles for a new batch of the same :~Ldl Iddl d, no acid soak is necessary - simply rinse several times with high purity water and then several times with small portions of the fresh :~ldnddld.

b. As the working ~Idnddrd:~ in the 125 mL bottles are used up, simply refill the bottles from the 1 L ~Idnddlds prt:pared in D. 3.
c. Because there are many Sdl ", !e ~, the most efficient way to add the cobalt internal sldllddld to the Sdlll, 'es is with a 1 mL digital pipet.
2. Sample Pl ~Jdl dlion a. Refill the reagent containers before p~:pdli"g Sdl~ ~ '-s so that the same batch of r~agenl~ can be used for all Sdlll,''_S and the blank.
b. Remove the caps and arrange the empty 50 mL tubes, with labels, in the rack b -g- "~;"9 with the sample blank and two tubes for each sample.
c. Transfer sample to a plastic storage container. Place these cor,' ~ ,el:, directly on a Illayll~lic stirring plate and add a Teflon coated stirring bar. Set the stirrer at an i, llel 1 l ,edidlt: speed.
After a minimum of one minute of ayildlion begin to withdraw the sample for v. ~ I, ,9 with a ~ - - ~ '~ 'e plastic transfer pipet.
d. Carefully weigh and record to the nearest 0.0001 9, 5 g of sample into the plastic tubes. The sample blank tube is left empty at this point. Add the ~ ;"g r~agenl~ to each tube, including the blank, in this exact order:
(1) Add 2.5 mL of Triton X-100 solution using the d;_penser bottle.
(2) Add app,~.Ai", ~ly 45 mL of high purity water (3) Add 1.00 mL of the 1000 mg/L cobalt internal :,ldn-ldl .I with either a - - ' dl~d digital pipet (p,t:r~r,e:d) or a Class A
pipet.
(4) Add 1 mL of conceril,dLed hyd,uch' ic acid with an Cppendo, r pipet or from a Teflon di~penser bottle.
(5) Add high purity water until the total volume in each tube is app, uAill,at~ly 50 mL. Put the caps on and shake the tubes thoroughly.
3. Instrumental Analysis a. The f. " . ;. ,9 instructions refer to some general cha, d~;Leri:.Lics of the PLASMAVISION SOri dlt:, which is currently used on all instruments running this method, but no attempt has made to describe specific key sequences needed to perform these procedures, since that i,,ru,,,,dLion is provided in training.
Equivalent ope, dLions must be pe, rur" ,ed with other versions of the SOr~
b. Turn on the plasma and allow a thirty minute warm-up time before r-''b dlion. Turn on the computer and printer and start the SOri ~ . Begin pumping 2% HCI rinse solution through the nebulizer.
c. Perform an instrument configuration before the first caGrdliun is made for each 8-hour shift. This will check computer-instrument commu"i.,dlions and check the SAMI motor. Watch the motor to be sure that it turns properly.
~ d. Check the optical ' _ " "enl using the 150 mg/L calcium :~ldnddl d. This procedure will insure that the SAMI motor is operating properly and that the - ' b.dlion will always be pe, ru, l l ,ed near the exact center of each analyte peak. Perform CA 022l7264 l997-l0-02 W O96/31130 PCTAJ~f'01C01 this procedure before the first --' b.dLion is made once during each 8-hour shift. Choose appl ou, idLe setup options and then run the profile. The measured peak centers for the element to be measured must be within +6 SAMI units of the current SAMI
profile position. If this result is not oL: ~ ,ed consult the supervisor: either a new default SAMI profile position needs to defined, or the instrument requires service.
e. Select the apprupridle task and the applupridle calibration sequence file name and begin '-b. ~Lion. Aspirate the :~Ldnddl d solutions into the plasma starting with the ISRB solution plepd,~:d in D.5. The software prompts for each sldnddld by name. Be alert to any error l"essages If an error occurs, write down the ",essage and consult the supervisor. After the last sLdl)ddl d has been run save the data and have the soft--- .b C~ h te a linear ,eyl ession for each element. Print the r-' ~.dLion data, which su"",ldli~es the element illlensities in each sLdllddld the colleldLion coerri~ "L, and the c-'c~ d conce"L,dLions of the ele",eril:- in each :,ldnda,d.
f. Enter the section of the Sur~l a~ e to analyze the Sdl " le s Set the print options to print whatever docL""elr,ldlion is required.
Select a name for the file that will store sample results; if no file name is chosen, the proy,d", will store the data under the task file name by default.
9. Shake the Sdl " ~ 'e s i" " "ed;~ly before introduction into the ICP.
h. HIT THE ENTER OR RETURN KEY AFTER ENTERING THE
WEIGHT, VOLUME AND NAME OF THE SAMPLE. Note that the PLASMAVISION software prompts for the sample weight and volume before the sample is introduced and the sample name after it has been analyzed. The dilution volume for Sdl, !e s will always be entered as 50 mL , eyal I - of the actual volume in the sample tube.
i. Make sure that the sample introduction tube is placed in the 2%
HCI rinse solution for at least 2-3 seconds t ~ n Sdl " 'e s Analyze the :,landd, d~. and the Sdl l le s in the following order:
(1) Analyze the i"le""ediaLe check sLdl-ddld solutions.
(2) Analyze the ",eage,~l blank" (or sample blank co, ' ~ ,s cobalt). The i"LensiLies of the analytes in this blank will be suLLIduLed auLu",_'; 'Iy from the illLensiLies found in the Sdl l l 'e (3) Analyze each sample in d~
j. Results can be lepo,Lt:d in any convenient concer,L,dLiun units, dependi"g upon how the tasks are p, uyl dl "" ,ed. Note: in cases in which the analyst has entered the e-' b.dLion ~Landd,d~
into the task as "mg/L" and has entered the dilution volume as "50 mL" and the sample weight in grams the sample results will be in units of ,ug/g. The actual printout will show whatever units are plUy~dl "",ed into the task for each element. The unit "~ug/g"
is prerer,ed to "ppm" for, epOI Lil ,9 sample results because the latter term is dll b: _ous In labordLo~ ies where sample results for this method are cor"monly, ~ o, led as "ppm" it must be ~" ,der:,Luod that this really means " ~ Uyl al l 15 of analyte per gram of sample."

W 096/31130 PCTrUS9''~SC01 A-3 VITAMIN C (L-ASCORBIC ACID) DETERMINATION
A. SAMPLE SIZE AND PRODUCT APPLICABILITY
Samples should be as uniform and ~I:pr~senldli~re of the product as pc ~ ~- !e . Sampling should be pe, r~""ed i"""edi..~ly after a gentle mixing or stirring to prevent inaccurate Sdll ,9 due to sl,~ rical;oll~ All sample weights must be ,t:cGrded to at least three siy"ir,canl figures.
Sample sizes for low pH beverages are ~'~t~d from the ~ ri,.g equation.
Sample size = 350/E
where:
Sample Size is the II ,eor~ al sample size, in grams; E is the ~ -e~--d ascorL-.~ acid conce"l,dlion in ", _ dl"s per liter or h-__ dl" ,~spe- ti~ely, as is and; 350 is the desired amount, in " Uy,d",~ (mcg), of ascorL:
acid in the sample prepd,dlion. The net conversion factor for ", oy~"~
to " _ dlll~ and h'-_ dlll:~ to grams is unity.
B. THEORY
In this method the amount of L-ascG,L:~ acid present in the sample is dett:", ,ed by co- ~ ~ "~I, ic titration. A coulo" ,~I, ic method of analysis measures the quantity of ele~;t, i~ ity required to carry out a cl ,e"
reaction. If the reaction is 100% efficient the p~cs~ge of one Faraday of ele :L, i.;ily will cause the reaction of one equivalent weight.
In this case iodine is coulo",~l, 'Iy gene,dL~d from iodide. The iodine then oxidizes the L-ascG,L ~ acid to dehy.l,uasco, ~ acid. When enough iodine has been produced to oxidize all the L-ascorL ~ acid in the sample, an excess of iodine will occur. This excess of iodine signals the equivalence point, and is del~c1~d by two con~ld"L polenLidl ele~,udes.
The quantity of ele~, i~iily used is given by the product of current times the time to reach the end point of the co~ Ldc titration. Thus the amount of iodine used is equal to the number of equivalents of L-ascorL:- acid and the amount of L-asco, L acid can be ~
Trichlor~,acelic acid is added to the sample to plec;~iLdl~: the protein and to ., ,c,;~, , the acidic con n necessal y for a quantitative reaction.
C. APPARATUS
Analytical Balance ~ Beakers 100ml, graduated ~ B-i"h",al1l1 E585 Polarizer Cable for Double Platinum Wire Electrode, Brinkmann cat.
no. 20-97-738-8 or equivalent CA 022l7264 l997-l0-02 W O96/31130 PCT~US96/O~G01 ~ Cabies for Platinum Foil Cle~ udes, B~i"h~"dnn cat. no. 20-97-770-1 and 20-00-853-9, or equivalent ~ Chart Paper ~ Dc~
~ Diâ,~:~,~ ' '3 Pipets ~ Disr ~ tips for pipettor ~ Double Platinum Wue E lecL,ude, Blillh.,'dnl1 cat. no. 20-92-3504, or equivalent ~ CleuLI ude Holder ~ [ppendo, r pipet or equivalent, 200 mcl ~ Isolation Tube, outer iidr"~ler 20 mm, 125 mm long, Pore Size C, Ace Glass Company cat. no. 7209-16; OR outer d;dlllt:L~I 12mm, 125 mm long, Pore Size E, Ace cat. no.
7209-10. Size of isolation tube dependa on size of ele~,L,udes used.
~ Keithly Model 225 Constant Current Source OR Keithly Model 220 Constant Current Source ~ Magnetic Stir Plate ~ Pipettor; 5 ml - OxFord, Wheaton, CppendG, r, or Fil " , p _:
Pipettor or dis"ense" 10 ml and 30 ml - Oxford, Wi ,edl~on, Lab Industries or equivalent ~ Platinum Foil E Ic~,L,udes, (2), Bril,hlllann cat. no. 20-92-110-2, orequivalent ~ Sample vials, 5 ml with screw caps ~ Shields: yellow or clear shields with a cutoff of 385 nanc" I I~L~I ~
Strip Chart Recorder; Kipp & Zonen or equivalent ~ Teflon-coated Stir Bars UlLldsoh;~, Bath Vacuum Flask, 2,000 mL
Volumetric flasks, 100, 500 ml, 1000 ml with :,Luuper:, W O96/31130 PCTrUS9C~'~lC01 D. REAGENTS
1. CHEMICALS
L( I ) asco,L acid (USP Rt:r~r~nce Standard, Official Lot); store in a de~ r M_Lapho:,~.horic acid; ACS or equivalent Potassium iodide; ACS or equivalent Sodium sulfate, anhydrous granular; ACS or equivalent Trich!cruacc:Lic acid; ACS or equivalent Il. SOLUTIONS
NOTE: All solutions, sa",, '-s and ~Ldnddld:~
must be pr~pa,ed and stored under UV shielded or yellow sh- - 'c' o d lighting unless otherwise stated (see APPARATUS).
NOTE: De ~sed water should be used for the 0.1M pot~csiurn iodide solution to prevent air w~iddLiùl- of 1- to 12. Degas water by placing deiolli,dd water into a vacuum flask, and placing it under vacuum for 15 minutes with sonicdliùn.
1. Poldasium lodide (0.1M) Weigh 8.3 (~0.5) 9 of pot~csi~rn iodide into a 500 ml volumetric flask. Dissolve and dilute to volume with deg~-cserl. deiol,i~ed water. Store in a tightly sLuppe,~d brown bottle at room L~r"perdLure. Do not store for more than one (1 ) week. Discard if solution acq~ s a yellow tinge.
2. L-Ascorbic Acid Stal~la-J (2000 mg/L) Store sLdllddld bottle in a d~icc~ r to prevent moisture absor~,Lion. Weigh accurately 0.200 (+0.0005) 9 and transfer quantitatively to a 100 ml volumetric flask.
Dissolve and dilute to volume with 3% metaphosphoric acid. The sLdndal d can be made fresh just before use each day, or can be stored in small vials in the freezer.
Standard stored in the freezer is good for two months.
3 . T. ich lor~acelie Acid (1 M) Weigh 163 (+0.5) 9 of trichloruac~Lic acid into a 1 liter volumetric flask. Add 500 ml deg~-csed deior,i,ed water.
Swirl until di~solved, then dilute to volume. This reagent ~ may be stored for one month at room l~r"perdL.Ire.
4. Sodium Sulfate (1M) Weight 142 (+0.5) 9 of sodium sulfate into a 1 liter WO96/31130 PCT~US9'~;C01 volumetric flask. Add app,u~i,"aLely 750 ml of d~ioni~ed water and mix until di~solv~d. Dilute to volume with deio~ ed water. This reagent'may be stored for six months at room len"~erdLure.
5. ~ t ,rhosph~r;. Acid (3%) Accurately weigh 15.0 (iO.10) 9 of ",~Ld,i~l-ospl-oric acid and quantitatively transfer to a 500 ml volumetric flask.
Add approxi",~t~ ly 250 ml deioni~ed water and swirl until ",~Ld~.ho~,horic acid is .li~.solved. Dilute to volume with water. This solution may be stored for one week under , ~r, i~e, dLi~ (2-8 ~C).

E. PROCEDURE
Instrument Settings - Figures 8 ' 10 and 11.
a. Recorder Chartspeed/input 1 mmlsecond or5 cm/min deper, ,9 on chart~eco,der/l volt b. Poldli~er Constant puL~:nLial 150 mV
E585 sensitivity 10 ", Udlll~J~
c. Keithly 1.56 ma Constant Current Source 1. Fill the isolation tube co" , ,9 the cathode 01e~ L, ude with 1 M sodium sulfate. The sodium sulfate continuously moves from the isolation tube through the glass frit into the sample solution and must be frequently r~ I n;sl,ed.
2. An instrument check should be done at the start of each day.
Using an [ppendo, r pipet (or equivalent) pipet 200 r" . li~ :, of the 2000 mg/L L-asco, L ~ ~ acid :~Ldnddl d into a 100 ml beaker. Follow steps 5 through 12 of the procedure.
Follow steps 1 through 4 of the ~'qtions. Compare the ê~,uel il l l~l, 'Iy deLt:", ,ed conce, ILI dLion to the Ll ,eor~Lic al concer,L,dLion. The results should be within 4% of the e~ d value. If not run another instrument check using fresh ,~ager,La and :,Ldnddld. If the e,~l,eri",er,Ldl result still differs by more than 4% from the ex,ueuLed :~Ldnddld value consult the method supervisor.
3. Agitate Sdll es well before and during Sdll ,9. Liquid Sdl, - s must be freshly opened. Analyses must be cGr, ~ t~ d within 20 minutes after the co" ~ ,er is opened.
Samples are weighed directly into the beaker unless otherwise stated.
4. vVhile swirling the sample add 5.0 (+0.1) ml of 1M
Ll i.:hlo, oact:Lic acid (TCA) to the sample. Swirl for 30 seconds to cor" !~- 'y pl~u;~JiLdl~ the protein.
5. Add 30.0 (+0.5) ml of 0.1 M pof-qcsi- Im iodide to the sample.
6. Add deg~-csed dEioni~ed water to apl ruxi" Idt~ly the 60 ml CA 022l7264 l997-l0-02 mark.
7. Add a stir bar. Lower the ele_t,udes into the sample soiution.
Make sure that all the ele~;t, udes are i" " "er:,ed. The sodium sulfate level in the isolation tube must be at least 2 cm above the sample level. Adjust the " Id~l ,etic stirrer speed so that stirring is vigorous but no air is entrained.
8. Switch the chart r~:corder on. Switch the poldl i~er on. Adjust the base line of the chart ,~co"~er to 10% of full scale.
9. Switch the con~ld"L current source on.
10. Titrate until excess iodine is produced i" ~ by a rising current curve. Stop the titration when the rising current curve has reached at least 70% of full scale on the chart rt:corder paper.
~ 11. Switch the r~corder chart speed to off and the con~LdnL
current source to standby.
12. Remove the ele~t,udes from the sample and rinse well with deior,i~ed water.
F. CALCULATIONS
1. ExLIdpoldL~ the linear portion of the rising current curve to the base line to locate the end point. The portion of the curve b~:t~r _en 70% and 30% of full scale will be linear. (Figure 12).
2. Count the number of ce"Li",~t~r:. from start of titration to the end point to the nearest 0.1 cer,Li",~ r.
3. Convert this di~Ad,)ce to seconds of titration time.
4. C ~ ~ the amount of L-asco, L ~ - acid present in the sample by the ~ ~.; ,9 formula:
C=mxixt = ixtxR
nxsxF s Vvhere:
C = concellL,dLion of L-ascorL: acid in mg/l or mg/kg (mcg/ml = mg/l and mcg/g = mg/Kg) m = 176 (gram ."~ weight of L-asco,L~- acid) n = 2 (change in valence) current in ~" IIIJa t = time in seconds F = 96487co~- "L,s/equivalent R = prupo, Lion ~y consLa"L for m n and F=0.912 mcg mA-sec s = sample size in g EXAMPLE: Assume a 3.0 9 sample of a low pH beverage was analyzed and the measured length of titration on the strip chart was 19.0 cm. Chart speed was 1 mm/sec.
The current during the titration was 1.56 mA

W O96/31130 PCTrUS9''~1C01 s where i= 1.56mA; s= 3.0 9; R= 0.912 mcg/mA-sec; and t= 19 cm x 1 sec x 1Omm = 190 sec 1 mm 1 cm C = (1.56mA) x (0.912 mcg/mA-sec) x (190 sec) 3.09 = 90 mcg/g = 90 mg/Kg

Claims (18)

CLAIMS:

1. A beverage concentrate comprising:
a. a source of calcium;
b. vitamin D;
c. vegetable oil; and d. a gum.

2. A beverage concentrate as described in claim 1 wherein the source of calcium is calcium glycerophosphate.

3. A beverage concentrate as described in claim 1 wherein the source of calcium is calcium citrate malate or calcium carbonate.

4. A beverage concentrate as described in any of claims 1-3 wherein said gum is selected from the group consisting of gum arabic, gum tragacanth and xanthan gum.

5. A beverage concentrate as described in any of claims 1-4 wherein said vegetable oil is selected from the group consisting of corn oil and partially hydrogenated soybean oil.

6. A beverage concentrate as described in any one of claims 1-5, and further comprising vitamin C.

7. A beverage concentrate as described in any one of claims 1-6, and further comprising lactic acid.

8. A beverage concentrate as described in any of claims 1-6 wherein said concentrate is in a dry powdered form intended for reconstitution with an aqueous solution.

9. A liquid beverage comprising the powdered beverage concentrate of claim 8 reconstituted with an aqueous solution.

10. A liquid beverage as described in claim 9 further comprising an acidulant, said liquid beverage having a pH in the range of about 2.8 to 4.6

11. A beverage concentrate as described in any of claims 1-6 further comprising water such that said concentrate is in a liquid form intended for dilution.

12. A liquid beverage or concentrate as described in any of claims 1-11 further comprising a sweetener.

13. A liquid beverage or concentrate as described in any of claims 1-11 further comprising a glucose polymer.

14. A liquid beverage or concentrate as described in any of claims 1-11 further comprising potassium benzoate.

15. A liquid beverage or concentrate as described in any of claims 1-11 further comprising a flavoring agent.

16. A liquid beverage as described in either of claims 9 or 10 wherein the beverage is carbonated.

17. A calcium supplement in solid form comprising calcium glycerophosphate, vitamin D and vitamin C.

18. A calcium supplement in solid form comprising calcium glycerophosphate, vitamin D3, vegetable oil, vitamin C, and a gum selected from the group consisting of gum arabic, gum tragacanth and xanthan gum.