WO2012140322A1

WO2012140322A1 - Method and system for determining the fitness index of a person

Info

Publication number: WO2012140322A1
Application number: PCT/FI2012/050359
Authority: WO
Inventors: Sami Saalasti; Aki Pulkkinen
Original assignee: Firstbeat Technologies Oy
Priority date: 2011-04-12
Filing date: 2012-04-12
Publication date: 2012-10-18
Also published as: EP2696766B1; FI20115351A0; US10123730B2; US10856794B2; US20190029586A1; EP2696766A1; EP2696766A4; US20140088444A1

Abstract

The invention relates to a method and system for determining the cardiorespiratory fitness level of a person, with the aid of freely performed exercise, in which training: • the physiological intensity of the person is measured periodically during the exercise session (5) • the external work output of the exercise session is measured simultaneously relative to the measured intensity (6), • from each period's measured intensity value and external work output, the representativeness of the values measured is determined (8.1, 8.2), in order to determine the fitness level using the following criteria: o the physiological intensity should be stabilized relative to the external work output, o the external work output should be within a selected range, o the physiological intensity should be greater than a selected lower limit, in which case, a fitness level estimate is defined (8.3) for each accepted period as well as a fitness from several estimates.

Description

METHOD AND SYSTEM FOR DETERMINING THE FITNESS INDEX OF A PERSON

The present invention relates to a method and system for determining the cardiorespiratory fitness level of a person, with the aid of freely performed exercise, in which exercise

• the physiological intensity of the person is measured periodically during the exercise session

• the measured intensity is measured simultaneously with the external work output of the exercise session.

In the system, there is

• an interface device containing input devices for entering optional user-specific initial parameters prior to the exercise and a feedback means for giving feedback,

• a memory register for recording the values of the said parameters and computation variables,

• first means for measuring and recording a variable proportional to the physiological intensity,

• second means for registering and recording the external work output simultaneously relative to the measurement of the physiological intensity.

Most wristop devices equipped with a heart rate monitor guide exercise in a selected manner. The user selects an operating mode, which can be, for example, the achievement of a selected training effect, an increase in endurance, or a reduction in weight. User- specific parameters, which are, for example, sex, age, weight, and height, are entered in the device when it is begun to be used. It is easy for the user to set these data in the device. However, it is difficult for the user to provide, for example, information on the maximum heart rate level (HRmax) and the fitness level (V0₂max), as well as the activity class. As such, it is already known that a device (for example, a heart rate monitor) can propose, for example, increasing the HRmax value, if a heart rate value higher than the default value or set value appears in use. Nowadays, web-based services are also popular, in which the data of the heart rate monitor are transmitted through the internet to a server, with the aid of which a computer program, performing physiological computation, is used. Publication US 2007/0082789 A 1 discloses a method for determining a person's performance from physical exercise. The method cannot, however, be utilized during exercise, but instead requires the data to be analysed separately after exercise.

A method is known from publication WO 2009/133248 for determining a free fitness level from exercise, in which heart rate and external output, such as, for example, speed and height together, are measured periodically. A bicycles ergometer and many other exercise devices give the external output directly and, if the speed of running exercise, for example, is measured together with altitude information, it is possible to determine what the external work output is. With the aid of free exercise, as long as there is enough data, the level of the fitness of the user's respiratory and car- diovascular system (cardiorespiratory) fitness level can be determined. However, the method requires relatively long (30 s - 4 min) stable periods. The method is difficult to apply to the automatic determining of fitness level. Publication US 2012/0029370 (WO 2010/112010 Al) discloses a method using short standard loadings to determine fitness level.

The present invention is intended to avoid the drawbacks of the known methods and systems relating to their difficult start-up con- sidering the initial data required or precisely controlled exercise. The characteristic features of the method according to the invention are stated in the accompanying Claim 1 and the features of the system implementing the method are stated in Claim 3. The method according to the invention can effectively utilize short pe- riods of, for example, five seconds, which are defined as represen- tative, in terms of the determining of the fitness level, using selected criteria relating to the simultaneous physical-intensity and external work values. To be able to evaluate fitness on the basis of heart rate, it is important to be able to exclude certain loading stages, which do not produce a reliable result. For example, particularly during recovery, the physiological intensity (e.g., heart rate) is considerably higher when compared to external work output (e.g., running speed) . If fitness were to be estimated from these recovery periods too, the real fitness level would be considerably underestimated.

The loading stage can be evaluated, for instance, on the basis solely of the behaviour of the heart rate (heart rate increases, decreases, or remains constant), or together with external work- output (heart rate increases, decreases, or remains constant simultaneously with the external work output, or in a specific relation to it) . The loading stage can also be evaluated, for example, with the aid of a variable (such as EPOC) depicting a change in the ho- meostasis (at least an increasing variable will eliminate a recovery from the loading stages) .

With the aid of the method according to the invention, a wristop device can be easily made to create the user's fitness level and to update it during use.

If at least the so-called difficult initial parameters, i.e. the parameters depicting the user's internal data, particularly their fitness level, are each taken initially into the calculation as a preselected default value, the system can update their value gradually to the physiologically correct value.

In a preferred embodiment, the user is asked by software means contained in the system or device for external parameters such as sex, age, weight, and height. The fitness level, e.g., the V0₂max, and the maximum heart rate HRmax are taken initially as default values. The HRmax value always begins immediately to slide to the right towards the physiological maximum value when the user exceeds the highest previous maximum heart rate level measured under stress. The updating of the fitness index, on the other hand, demands data, i.e. that the user performs one or even several exercise sessions, which produce periods accepted when determining the fitness index.

In a preferred embodiment, automatic calculation of the fitness in- dex takes place in a simpler manner than that presented in the WO publication. A model is created from empirical material to depict the V0₂max, for example, ml/kg/min, i.e. oxygen consumption per minute per kilogram of weight, using intensity as the variable, in a preferred form from the relative heart rate (%HRmax) and external work output. Next, each heart rate signal datum recorded from the user's free exercise is processed and suitable periods are sought from it, for example, using the following criteria:

• stable loading state

• external work output in the selected range

· heart rate above a set criterion (x % HRmax)

The accepted level should be in a stable loading stage, i.e. the heart rate should be stabilized relative to the external work output. This criterion is preferred met when the EPOC value is rising.

Each accepted period gives a V0₂max estimate from the said model. The final single V0₂max value of a single exercise session is obtained by calculating either the average, the median, or the weighted average from the V0₂max estimates of the accepted periods.

In one embodiment, the user's activity class is also an automatically updated value, so that its preselected default value gradually approaches the correct value. It is preferably determined from a 3-dimensional table, or a function depicting it, in which the axes of the table are activity class, age, and sex. In summary, it can be stated that the invention permits the fitness level to be determined from the user's arbitrary exercise performance, provided there are sufficiently intensive portions in it. When applying the invention in a wristop device or similar, the two main benefits of the invention are, firstly, the easier start-up of the device or product and, secondly, the automatic increase in precision and calibration to the fitness level through use. If previously it has been desired to exploit information on the fitness level in feedback or guidance, the fitness level has had to be estimated indirectly, making the feedback given or the functionality imprecise. The fitness level can also be measured precisely, for example, using laboratory tests, and after this entered manually into the device, but this is not very practical and demands consid- erable effort and arrangements from the user, so that it is seldom practical .

The fitness level varies and/or changes continuously due to the effect of several different factors, such as, for example, stress and tiredness, as well as the recovery state, illnesses, or aging. Thus, if the user's fitness level is calibrated automatically, it will ensure the accuracy, correctness, and suitability to the user in different situations, of the feedback given to the user. If no allowance is made for changes taking place in the fitness level, or if wrong estimates of the fitness level are used, the feedback can be erroneous and, in the worst case even detrimental to the user.

CENTRAL CONCEPTS User-specific initial parameter: Information depicting the person, which can be, for example, age, height, weight, sex, maximum heart rate (HRmax), fitness level, i.e. maximum oxygen consumption (V0₂max), activity class, or some other personal information. External work output: External work output is a person's external measured work output achieved when moving, such as, for example, a person's running speed, or running speed on a specific piece of ground, rising or falling angle of a treadmill, or measured using an output measuring device in free performance, or some other movement, for example, using an acceleration sensor, or output measured in some other way. External work output can also be calculated from a person's specific known performance, for example from the running time over a known distance, based on the length of stride when run- ning, step contact, or step density. External work output can be converted, for example, converted into oxygen consumption (VO_:) or energy consumption, in which case it will depict the amount of oxygen or energy consumed when a person moves at the output in question. Thus, external work output can be measured or estimated, for example, as continuous data during exercise, or as a single average depicting the entire exercise session.

Fitness level: A person's aerobic, i.e. endurance condition, the (cardiorespiratory) condition level, or performance capacity of the respiratory and cardiovascular system. Fitness level can be depicted using parameters (fitness index), which can be, for example, maximum oxygen intake (V0₂max), METs ( =V0₂max/3.5 ) , maximum work output defined in watts, for example from a bicycle ergometer test, or using an output measuring device in free exercise, maximum run- ning speed in a maximum running test, or in free running taking into account altitude data, a Cooper test result, the fastest time for running a specific distance (.e.g., 3, 5, or 10 km), or some other external measured endurance performance depicting maximum work output. The fitness level can be measured or estimated, for example, using various direct or indirect fitness tests, for example, using a free fitness test like that described, various evaluation methods, for example Jackson et al. 1990, using various questions (e.g., "are you able to run 3 km non-stop?", "are you able to walk with a reasonably strenuous output non-stop for 10 minutes?", or in a similar manner), on the basis of the answers to which the fitness level can be estimated, or else using the person's own subjective evaluation.

Physiological intensity of exercise: This tells the output at which the person moves. The physiological intensity of exercise depicts, for example, the heartbeat frequency of the person's heart, i.e. their heart rate. The intensity can be scaled relative to the person's fitness level, i.e. their maximum performance, so that it tells at what output the person moves relative to their own maxi- mum. The heart rate can be scaled at the person's maximum heart rate and a percent intensity of the maximum heart rate (HRmax) can be obtained. Other variables depicting physiological intensity can be, for example, the person's frequency of respiration, their oxygen consumption (V0₂), or their ventilation. The intensity of exer- cise can also be the external measured work output, or activity, measured, for example, using an acceleration sensor in a preferred form relative to the person's maximum external work output. Intensity can also be estimated on the basis of the person's subjective feeling. Thus, the physiological intensity can be measured or esti- mated, for example, as continuous data during exercise.

The loading stage of exercise: In exercise performance, there are several different loading stages, during which the body's physiological state and functions, i.e. the physiological intensity, are regulated to correspond in the best possible manner to correspond to the external work output. The loading stages are, for example, the on response, i.e. the body's response to the start of exercise; the steady state, i.e. the stabilized physiological intensity when the external work output is constant; and the off response, i.e. recovery, when the external work output decreases or stops altogether. In freely performed exercise, the various loading stages alternate naturally with each other, partly merging according to, among other things, the shapes of the ground (uphill, downhill) and the performance of the exercise (breaks for instance for traffic lights, etc.) . In terms of evaluating fitness on the basis of heart rate, it is important to exclude certain loading stages that do not produce a reliable result. For example, especially during recovery the physiological intensity (e.g., heart rate) is considerably higher relative to the external work output (e.g., running speed) . If fitness were to be evaluated also during their recovery periods, the real fitness level would be considerably underestimated. A loading stage can be evaluated, for instance, on the basis of heart rate behaviour (the heart rate rises, falls, or remains constant) solely, or in combination with the external work output (the heart rate rises, falls, or remains constant simultaneously with the external work output, or in a specific relation to it) . A loading stage can also be evaluated, for example, with the aid of a variable (such as EPOC) depicting a change in homeostasis (a variable that is at least increasing will exclude the recovery loading stage) .

Figure 1 shows a flow diagram of a system monitoring a user's physical state,

Figure 2 shows the system of Figure 1 applied to guiding training,

Figure 3 shows graphically the determining of the fitness level,

Figure 4 shows a flow diagram in the calculation of energy consumption,

Figure 5 shows a flow diagram in the calculation of training effect,

Figure 6 shows a set of curves for determining training effect,

Figure 7 shows a way to calculate a fitness index, and

Figure 8 shows a block diagram of the system in a wristop device .

According to Figure 1, in the system according to the invention, additional attention is paid to the start-up 1, compared to the start-up 1' of conventional use. After the loading 2 of the program and the initiation of the registers, during the first operating session the optional default values are generated, point 3. Later, loading 2 ' the program leads directly to processing 4 of the computation registers, in which at the same time measurement 5 of inten- sity and measurement 6 of external work output are initiated, for example (position and altitude data obtained from a GPS service) . The computation registers contain continuous real-time information on the user's intensity and external work output. The user's intensity is usually the measured heart rate value. The external work output can be obtained in many ways, depending on the kind of the exercise and the devices used. A GPS device used in connection with wristop devices provides altitude data, which is one alternative. The altitude data is optional and is not necessary if the exercise session takes place on level ground. The exercise device, such as, for example, a bicycle ergometer, can provide the external work output directly.

In a preferred embodiment, after initiation of the program has been commenced, on the first time it asks the user for external data, such as sex, age, weight, and height. Alternatively, these too are default values, which the user corrects if they wish.

Conventionally, the device has some selected operating mode 7, i.e. manner of use, which aims at some desired end result, for example a selected level of energy consumption, and provides feedback 9 on it .

Because particularly the fitness level is required in the calculation of many of the target variables, it is calculated automati- cally with the aid of calculation devices 8 belonging to the system. Other data, such as maximum heart rate HRmax and the activity class, are preferably updated in the same way.

In Figure 2, the usual operating mode 7 of Figure 1 is exercise guidance. The target of the exercise can be, for example, achieving a selected training effect, or exercise for reducing weight. In order to achieve the target, the user is provided with continuous feedback, i.e. whether the intensity is suitable or too high/low for achieving the target within the set time. The guidance feedback is calculated on the basis of the register data at point 10. The provision of the guidance feedback 11 is checked by a conditional statement 11, if the exercise has been completed. If the exercise is still under way, a return is made to the calculation of the guidance feedback, otherwise exit to the end 12 of the routine. In Figure 2, it should be noted that the automatic update routine (8, 8') can also obtain a parameter value from the exercise, which is automatically updated in the computation register. This can be, for example, the value of the fitness index, which is calculated only after the exercise and not continuously.

Figure 3 illustrates the calculation of the fitness level value by means of a preferred method. The graph in the upper part of the figure shows the momentary speed (km/h) in running exercise and its mean value as the exercise proceeds. In the lower part of the fig- ure, the relative heart rate value %HRmax (0 - 100 % of the maximum heart rate value), the altitude in metres, and the computed fitness level estimates e (V0₂max ml/kg/min), calculated in five-second intervals, are shown on the same time scale. At about four minutes, the acceptable data are sufficient and there is a thick line in the graph showing the fitness level computed from the points e obtained. Each value of the fitness level is obtained from the modelled function by entering into it the relative heart rate and external work of the accepted segment. The modelled function is obtained with the aid of numerous test exercise sessions performed by test persons. The exercise sessions of the selected group were monitored for a period of several months and at regular intervals their fitness level was measured using a clinical method. From the data obtained and previously known information on the relation between the external work output and oxygen consumption, a modelled function depicting fitness level was formed, with heart rate and external work as the variables. The data collected also included sex, age, and weight, by means of which the modelled function can be focussed more precisely on each user. The calculation of the fitness level estimate from simultaneous physiological-intensity and external-work values is computed, in one embodiment, as follows. It is generally known, that, by comparing the HRmax heart rate to the theoretical work output and in turn by extrapolating the maximum heart rate, a rough estimate of the fitness level can be obtained.

In the literature, a basic equation is known for running exercise: V0₂ma (ml/kg/min) = 11.1 *speed_ms+5.3333 Here, the following equation is derived from it, which takes slope into account:

V0₂max= (- cl*hr/maximal_hr+ (l+cl))*(ll.l* (c2*angled+l ) *speed_ms+5.3333) V0₂max ml/kg/min, the momentary V0₂max value i.e. estimate

hr BPM instantaneous, momentary heart rate level

maximum hr BPM, the person's background parameter (maximum heart rate level)

angles in radians, momentary running angle

speed_ms m/s, momentary running speed

cl, c2, parameters optimized from empirical material

A corresponding equation is created for each form of external work output, in such a way that the calculated V0₂max corresponds to the real value.

Example of application 1 (Figure 4). Calculation of energy consump- tion . Automatic determining of the fitness level can be used, for example, in the calculation of energy consumption. The calculation of energy consumption generally requires the person's age, height, weight, sex, and the activity class that depicts the level of the person's physical activity. On the basis of these data, it is possible to estimate, using, for example, the equation developed by Jackson et al . (1990), the person's fitness level, i.e. the maximum oxygen consumption (V0₂max or METmax) , which can be used in calculating the energy consumption, see Figure 4.

In an automatic system, a person can only be asked, for example, their age, sex, and weight, so that the rest of the person's background data are set as certain default values, on the basis of which calculation of V0max takes place using the aforementioned Jackson equation. On the basis of age, it is possible to use the traditional equation to calculate the maximum heart rate 210 - (age*0.65) = maximum heart rate HRmax, which is the necessary variable for performing a fitness test. After this, V0₂max is computed from exercise performed by the person, which gives a more precise value for the original estimate of V0₂max . This more precise estimate of V0₂max is then used in the calculation of the energy consumption .

Age and weight can also be preset, in which case the person will not need to enter any personal background data. The automatically calculated and updated V0₂max keeps the system accurate.

Oxygen consumption can also be calculated in a corresponding manner to energy consumption.

Figure 4 shows an example of the computation model for energy consumption, in which the V0₂max value calculated automatically from exercise sessions reduces or entirely eliminates the entering of the personal background data set at the start . Thus the product is considerably easier to start up, and in addition the product always produces accurate information on energy consumption, even though the person's fitness changes.

1. RR data, i.e. a time series of the time always between two consecutive heartbeats.

2. The heartbeat interval data is corrected by automatic error correction .

3 Respiratory frequency information is calculated from the heartbeat interval data.

4. The heart rate per minute is calculated from the heartbeat interval data.

5. The heart rate is converted to % of the person's maximum heart rate, using the person's maximum heart rate. The maximum heart rate is defined on the basis of the person's age, if the person does not know it directly.

6. The on/off kinematic information (loading stage) is calculated on the basis of the respiratory frequency and %HRmax.

7. % V0₂max or METmax intensity is calculated, i.e. the physical loading exertion level as percentages relative to the person's maximum .

8. The RQ (respiratory quotient) value is calculated, which gives the ratio of fats and carbohydrates as energy sources.

The RQ relates to a known way to calculate energy consumption on the basis of oxygen consumption. Generally, oxygen consumption (litre/minute) is multiplied by five, giving the energy consumption in calories.

9. The relative intensity % V0₂max is converted to absolute oxygen consumption by multiplying it by the person's V0₂max (maximum oxygen consumption) . If the person does not know their VO_:max value, it can be calculated using Jackson et al.'s 1990 equation, on the basis of the person's background data. The V0₂max value can also be obtained on the basis of an automatic fitness test. All the person's background data can be set as default values, in which case the automatic fitness test will make the V0₂max value more precise, as exercise sessions accumulate. 10. The calorific equivalent is calculated, which tells on the basis of the RQ value how much energy is produced per litre of oxygen consumed.

11. The momentary energy consumption kcal/min is calculated.

12. The sum of energy consumption kcal is calculated.

Example 2 of application (Fig. 5) . Calculation of training effect.

The automatic determining of fitness level can be used, for example, in the calculation of the training effect. Typically, age and activity class from the person's background data are used in the calculation of the training effect, see Figures 5 and 6. The activity class can be determined on the basis of the result of a free fitness test, V0₂max, together with age and sex, so that the person need not known their activity class, see Table 1.

Figure 5. Calculation of training effect.

2. The heartbeat interval data is corrected by automatic error correction .

4. The heart rate per minute is calculated from the heartbeat interval data.

7. % V0₂max or METmax intensity is calculated, i.e. the physical exertion level as percentages relative to the person's maximum . 10. EPOC (Excess Post-exercise Oxygen Consumption) is calculated at the present moment t, using V0₂max intensity (7), ^' as well as the previous EPOC value (8) and the duration (9) of the time interval being examined.

The exercise's EPOC peak, i.e. the highest EPOC value so far, is calculated.

The EPOCpeak value is converted to Training Effect by using the activity class according to Figure 3. If the person does not know their activity class, it is set to a specific value as a default, or is calculated from the V0₂max value in the manner of Table 1, using either an automatic fitness test, or the V0₂max value estimated by Jackson et al . 1990. All of the person's background data can be set to be default values, in which case the automatic fitness test will make the V0₂max value more precise as training accumulates.

According to Figure 6, the training effect can be determined on the basis of the training's highest EPOC value as well as the person activity class. The person's activity class can be determined automatically on the basis of the automatically updated fitness level (V0₂max) . The following Table 1 shows the V0₂max values according to age and sex for determining the activity class. These two sub- tables together form a 3-dimensional table, in which the axes are age, sex, and activity class. In each cell contains a V0₂max value determined by these variables. The activity class is found by searching for the correct column with the aid of age and sex, from which the known V0₂max value is sought, on the row of which the activity class is sought.

Table 1. From Table 1, the person's activity class V0₂max can be determined on the basis of sex and age. The activity class can, in turn, be used, for example, to determine the training effect from the EPOC value . Example 3 of application. Selection of training programme. The automatically determined fitness level can be used in the selection or adjustment of the training programme. Typically, the person must estimate their activity class, or the activity class is determined on the basis of the person's activity history. On the basis of the result V0₂max of the fitness test determined on the basis of a free fitness test, the training programme can be selected directly to suit the person's fitness level, or the activity class can be calculated on the basis of Table 1 and the programme determined on its basis, see Table 2.

Table 2 shows training programmes (table row) . The desired training programme can be selected on the basis of the activity class calculated (Table 1) from V0₂max ("original activity class" in the table) .

In a preferred solution, the V0₂max data obtained on the basis of an automatic fitness test from several training sessions by the same person is used. In the best case, the reliability of the estimation of the V0₂max of each individual training session, and how much reliable material has been found temporally from the training, can be used in the weighting. If the amount of reliable data is 5, 10, 15, 20, 25, 30, 40, or 50 minutes, the corresponding weighting coefficients as percentages are 10, 20, 30, 40, 50, 60, 80, and 100 %, in that order. For example, if reliable material has been obtained from 4 minutes of the entire training in the automatic test, this new V0₂max value is added to the V0₂max average values of the previous test using a weighting coefficient of 8 %, and correspondingly the amount of reliable material by 30 min with a weighting coefficient of 60 %. Naturally, it is preferable for the V02max of the first training session to be calculated with a weighting coefficient of 100 %, because it probable that this value will be more correct than the preselected V0₂max value, the V0₂max value calculated using the equation of Jackson et al. 1990 on the basis of the person's background data, or the V0₂max value calculated using the equation of Jackson et al. 1990 on the basis of the person's preselected background data.

Figure 7 shows the calculation of a fitness index, for example, during running or walking. The person's maximum heart rate (HR ax) , heart rate (heartbeats per minute), speed (e.g., m(k)/h), and altitude (metres) are entered in the computation register. Each period being examined must be reliable, in order to create a fitness index. Reliability can be examined, for example, by setting reliabil- ity for known influencing factors, in this example external work (W) , angle altitude, EPOC gradient, physiological intensity, limit values (cl, c2... c9), within the framework of which the variables must remain during the period being examined, for the period to be accepted. If the period is determined to be reliable, a fitness in- dex is calculated for it, with the aid of external work (W) and physiological intensity (Int) . So that the real fitness index can be shown reliably, the total time or number of the accepted periods must exceed the time/value (clO) regarded as reliable.

According to Figure 8, in a typical application (e.g., wristop device) the implementation comprises an assembly built around a central processing unit (CPU) 32. A bus 36 transmits data between the central unit 32 and the other units. The input unit 31, ROM memory 31.1, RAM memory 31.2, keypad 18, PC connection 37, and output unit 34 are connected to the bus.

The heartrate sensor 12 and some sensor 30 registering external output are connected to the input unit 31, which handles the sensor's data traffic to the bus 36. Optionally, the PC is connected to a PC connection 37. The output device, for example a display 15, is connected to the output unit 34. In some embodiments, voice feedback is created with the aid of a voice synthesizer and a loudspeaker 35, instead or, or in addition to the feedback on the display. The sensor 30 measuring external work can, in fact, comprise a_. group of sensors, which are used together to define the external work done by the user. All of the default values of the optional parameters are preferably stored in a ROM memory, or more specifically, e.g. in an EEPROM (Electrically Erasable Programmable Read-Only Memory) memory.

For example, the user's "external" data:

sex man, age 35 years, weight 75 kg, height 180 cm.

User's more demanding data:

fitness level (V0₂max) 40 ml/kg/min; Activity class 4. In a web service, the default values of the parameters are preferably recorded in self-service software.

In these embodiments, it would be as such also possible to use some other method than that described above as a fitness test. However, the fitness test according to the invention provides several advantages in terms of automatic updating. It can be completely integrated in many standard-model wristop devices and demands substantially fewer calculation stages than the method according to the WO publicatio .

The invention can be applied, for example, in the following applications: wristop device, mobile/cellphone application or device, fitness device, computer software, or web service. REFERENCES

Jackson, A.; Blair, S . ; Mahar, M.; Weir, L . ; Ross, R.; and Stute- ville, J.; 1990: "Prediction of functional aerobic capacity without exercise testing". Medicine and Science in Sports and Exercise, 22 (6) : 863-870.

McArdle, W.D.; Katch, F . I . ; and Katch, V.L.; 2000: "Exercise Physiology: energy, nutrition, and human performance", 5^th ed. Baltimore, Williams and Wilkins.

TABLE 2

Claims

1. Method for determining the cardiorespiratory fitness level of a person, with the aid of freely performed exercise, comprising the following stages:

• the physiological intensity of the person is measured periodically during the exercise session (5)

• the external work output of the exercise session is measured simultaneously relative to the measured intensity (6)

· at each time, it is determined whether the physiological intensity has stabilized relative to the external work output,

• from each period's measured intensity value and external work output, as well as from the stability information obtained, the representativeness of the values measured is determined (8.1, 8.2), in order to determine the fitness level using the following criteria:

o the physiological intensity should be stabilized relative to the external work output,

o the external work output should be within a selected

range,

o the physiological intensity should be greater than a selected lower limit,

in which case, a fitness level estimate is defined (8.3) for each accepted period and, if the total duration/number of the periods exceeds a selected time/value, a fitness index is defined with the aid of the fitness level estimates.

2. Method according to Claim 1, characterized in that

• a variable, such as the EPOC value, depicting the change in cumulative homeostasis, is calculated continuously

• the value of the variable depicting homeostasis should be increasing, in order for the period to be accepted.

3. System for determining the level of the cardiorespi atory fitness level of a person, with the aid of freely performed exercise, which system includes

• an interface device containing input devices (18) for entering optional user-specific starting parameters before training, and a feedback, device (15, 35) for providing feedback,

• a memory register (31.2) for recording the values of the said parameters and the calculation variables,

• first means (12) for measuring and recording a variable proportional to physiological intensity,

• second means (30) for registering and recording external work output simultaneously relative to the measured physiological intensity,

characterized in that the system further includes

• third means for determining whether the physiological intensity has stabilized relative to the external work output,

• fourth means (8.1, 8.3) for determining the representativeness of the values of the intensity and external work recorded from simultaneous periods, in order to accept or reject each period in terms of fitness level, and

• means (8.3) for calculating a fitness-level estimate from the accepted periods, and to define a fitness index from the calculated fitness level estimates.

4. System according to Claim 3, characterized in that, in terms of determining the fitness level, the representative periods are those periods that have the following properties:

• the physiological intensity must have stabilized relative to the external work output

• the external work output must be within a preselected range,

• the physiological intensity must be greater than a set criterion (x % HRmax ) .

5. System according to Claim 4, characterized in that the system includes means for calculating continuously a variable, such as an EPOC value, depicting the change in cumulative homeostasis, in which case the fourth means are arranged to accept periods, in which the variable depicting the change in homeostasis is increasing .

6. System according to any of Claims 3 - 5, characterized in that the system includes means to determine the user's activity class, with the aid of initial data and a defined fitness index, using preselected criteria.

7. System according to Claim 6, characterized in that the said preselected criterion for determining the activity class comprises a 3-dimensional table with the following variables: sex, age, and fitness index.

8. System according to any of Claims 3 - 7, characterized in that the external work output is arranged to be determined on the basis of running speed and slope.

9. System according to any of Claims 3 - 8, character i zed in that the system is arranged to perform updating of the fitness index after each training session.

10. system according to Claim 9, characte ized in that the system is arranged to calculate the fitness index from two or more training sessions and to weight the fitness index of each training session according to how many accepted periods are found from the training session.

11. System according to any of Claims 3 - 10, character i zed in that the system is arranged to calculate the energy consumption during training .

12. System according to any of Claims 3 - 11, characterized in that the system is arranged to calculate the training effect (TE) created by the training.

13. System according to any of Claims 3 - 11, characterized in that the system is arranged to define a training programme according to the calculated fitness index.