Obesity in children and adolescents shows an increasing prevalence worldwide (16). Physical inactivity and hence a decrease in energy expenditure is suggested to be one of the main risk factors for developing obesity (16). Ischemic heart disease and cerebrovascular disease are the two leading causes of death throughout the world today (18) and are also strongly related to physical inactivity (3) and a decrease in cardiorespiratory fitness. But today’s common work of promoting physical activity (PA) falls short in the lack of an international agreed measure of PA (5). The measure needed must be able to be used in large-scale population surveys, and self-report is the only feasible method to fulfil that requirement. However, the majority of research studies using self-report have used different or modified instruments (5), most of them not validated against a criterion method (22). Data on physical activity are most often translated into energy expenditure, but type and duration of physical activity are often not presented.
The Minnesota Leisure Time Physical Activity Questionnaire (MLTPAQ) (31) was presented in 1978 as a tool to evaluate energy expended in leisure time activities. MLTPAQ has frequently been used in studies of physical activity in adults, and numerous validation and evaluation studies have been done against other physical activity questionnaires, accelerometers, fitness tests, and criterion methods on adults, with inconsistent results (2,7,9,14,15,20,28,30), probably due to varying degree of unreported time from the participants. In an attempt to limit the unreported time, we suggests in this study to add questions concerning inactivity, such as sleeping time and leisure-time sitting activities (the extended Minnesota Leisure Time Physical Activity Questionnaire (eMLTPAQ)).
Reeder et al. (19), in their study on a cohort of 15-yr-old adolescents from New Zealand, were the first who used MLTPAQ on adolescents. However, the method has not yet been validated in adolescents. Adolescents are midway between childhood and adulthood, but although at the same age, individuals may be at different stages in physical and mental maturity. This puts extra demands on a subjective method to examine physical activity compared with a method used on adults. On the other hand, adolescents, as opposed to adults have a quite homogeneous “occupational” activity pattern, being at school.
The doubly labeled water (DLW) method has made it possible to determine the total energy expenditure (TEE) in free-living subjects. The technique is well established and validated both in animals and humans (24,25), and is today considered to be the golden standard for assessment of TEE, which is assessed with high precision without limiting the subjects’ way of life. The result represents energy expenditure for a relatively long period of time, in contrast to short-term measurements with major limitations in way of life using direct calorimetry.
The aims of this study of a sample of Swedish 15-yr-old adolescents were to validate the TEE estimated from the MLTPAQ with the TEE measured by DLW, and to present and examine the validity of an extended version of the MLTPAQ with additional questions about inactivity during leisure time (eMLTPAQ).
Outline of the study.
Two cross-sectional studies of altogether 2400 adolescents, 1170 boys and 1230 girls, were performed in 1994 and 2000. The aim was to study iron status in Swedish adolescents when the general iron fortification of white wheat flour was removed. To achieve a representative sample of adolescents of Göteborg, great effort was made in selection of schools using socioeconomic area index (high, medium, and low status area). The selection procedure is carefully described elsewhere (27). The studies were school-based, i.e., all data were collected at school, and all students in grade nine in 13 selected schools were invited to participate in the studies, which consisted of a diet history, medical examination, anthropometrical and bone density measurements, and physical activity assessment with eMLTPAQ. In the current study, students in grade 9 in one of the schools (medium status area) included in the two cross-sectional studies were invited to participate.
Eighty-three students (43 boys and 40 girls) were invited to participate in the current study, and 54 (27 boys and 27 girls) accepted the invitation. Measurements of basal metabolic rate (BMR), dosage of DLW, and the eMLTPAQ interview were performed at the Sahlgrenska University Hospital, but 10 of the students refused to go to the hospital. We enrolled 44 adolescents (21 boys and 23 girls) between 15 and 17 yr of age. All subjects and their parents gave written informed consent and the study was approved by the Ethics Research Committee of the Sahlgrenska Academy at Göteborg University.
BMR was measured by indirect calorimetry using a ventilated-hood system. The equipment used was a DeltatracTM II Metabolic Monitor (Datex, Helsinki, Finland). The equipment was calibrated with gas mixtures of known O2 and CO2 contents according to the manufacturer’s instructions before each measurement. All subjects were measured after an overnight fast, and they arrived from their home by car or public transport. The subjects were instructed to limit their physical activity the evening before the measurement. After a 30-min rest in the supine position, BMR was measured during 30 min when the subjects were awake in the supine position. The measurements were performed in an environmental temperature between 22 and 23°C. The BMR for each subject is based on the mean of the last 25 min of the measurement. BMR was also predicted using the equations shown in Table 1 (11,17,26,34).
Body weight was measured, with subjects wearing light clothing without shoes, to the nearest 0.1 kg. Height was measured and determined to the nearest centimeter using a horizontal headboard with an attached wall-mounted metric rule. Body mass index (BMI) was calculated from body weight and height. A pair of Harpenden skinfold caliper was used to measure biceps, triceps, subscapular, and suprailiac skinfold thickness in triplicate. All anthropometric measurements were performed by the same specially trained and experienced nurse.
TEE from the DLW method was measured over a period of 14 d. Sample analysis and calculation procedures have been described elsewhere (29). At day 1, a baseline urine sample was collected for determination of background isotope enrichment. Then the subjects ingested a weighed mixture of deuterated and oxygen-18 enriched water, corresponding to 0.05 g of deuterium oxide (2H2O) and 0.10 g of oxygen-18-water (H2 18O) per kilogram body weight. Seven urine samples were collected by a member of the research group and the exact voiding time was registered in school, or at home during weekends, on day 2, 3, 4, 8, 13, 14, and 15. Samples were analyzed in triplicates on a Finnigan MAT Delta Plus Isotope-Ratio Mass Spectrometer (ThermoFinnigan, Uppsala, Sweden). TEE from DLW (TEEDLW) was calculated by the multipoint method by linear regression from the difference between elimination constants of deuterium and oxygen-18, with the assumptions for fractionating and a respiratory quotient of 0.85 (13). The relationship between pool size deuterium (ND) and pool size oxygen-18 (NO) was used as a quality measurement of the DLW analysis as proposed by the International Atomic Energy Agency (13).
All interviews were performed by the same well-trained interviewer. The interviewer asked the participants about what type of leisure-time physical activities (LTPA) they had been doing during the last year. Then the participants estimated the duration of the activities performed in minutes per week for each season. To be able to calculate energy expenditure (EE) for LTPA, the time reported for each activity was multiplied with a MET value (1) and predicted BMR per minute.
In addition to the original MLTPAQ, the participants also answered questions about sleeping time, time spent watching television/video, computer time, and time spent doing school work at home (the extended MLTPAQ). This was multiplied with predicted BMR and MET values for these activities (1). EE is calculated from equation 1:
We define the time not included in equation 1 as unreported time (UT), i.e., it is not recognized by the participants neither as a LTPA nor as sleep, TV/video, computer, and homework. The main part of the unreported time was time spent at school, which presumably did not consist of high-intensity activities. TEE was calculated from equation 2, where the unreported time was given a MET value of 1.5:
Physical activity level (PALeMLTPAQ) was calculated as TEEeMLTPAQ divided by predicted BMR. However, because there is a possibility that MET for unreported time differs from 1.5, we calculated the MET value for the unreported time from the TEEDLW as described in equation 3:
The Shapiro-Wilk test of normality showed that most of the variables concerning physical activity were not normally distributed. All data are therefore presented both as median (range) and mean (standard deviation). Simple linear regression and a paired t-test were used to compare predicted BMR and BMR measured by indirect calorimetry. Spearman’s correlation coefficient was calculated to study correlations between MLTPAQ/eMLTPAQ and DLW. Difference between TEEDLW and TEEeMLTPAQ were plotted against the mean TEEDLW and TEEeMLTPAQ (the Bland-Altman comparison technique (4)), and Spearman’s correlation coefficient was used to study correlations between the difference and the mean. The mean difference between methods and the 95% limits of agreement (mean difference ± 2 SD) were calculated. Statistical calculations were performed by using the SPSS for Windows (version 10.0; SPSS Inc., Chicago) software program.
Thirty-five (17 girls and 18 boys) participants completed the whole study. Five girls and two boys were not willing to take the DLW while one girl and one boy did not complete the urine samples. The median age of the students was 15.8 yr and the median BMI was 20.4 kg·m−2. Anthropometric data are shown in Table 2. The two largest reported times from the eMLTPAQ method were sleeping time and time for watching TV/video. Median reported LTPA were 35 min·d−1 (Table 3). Predicted BMR by the equation of Molnar et al. (17) was the one that correlated best (r2 = 0.73, P < 0.001) with the measured BMR and had the lowest residual standard deviation, indicating the best fit (Table 4). This equation was therefore chosen in the calculation of TEE from MLTPAQ. Compared with measured BMR, the predicted values underestimated BMR in 24 of the 35 students (Fig. 1), with a mean difference of 0.2 MJ·d−1 (95% limits of agreement: −0.9 to 1.3 MJ·d−1). The relationship between ND and NO fell well into the accepted range between 1.015 and 1.060 (Table 5). Calculated PALeMLTPAQ was 1.37, compared with 1.77 from the DLW method using measured BMR and 1.85 using predicted BMR (Table 5). The Spearman’s correlation coefficient between TEEDLW and EELTPA was 0.49 (P < 0.01). Including the questions about inactivity increased the correlation to 0.65 (P < 0.01). Adding information about the unreported time led to a correlation between TEEDLW and TEEeMLTPAQ of 0.73 (P < 0.01). However, as presented in Figure 2, the eMLTPAQ underestimated TEE with a mean difference of 2.8 MJ·d−1 (95% limits of agreement: −0.1 to 5.6 MJ·d−1). This indicates that 1.5 may be a too low MET value for the unreported time. Calculated MET for unreported time (from equation 3) ranged from 1.26 to 5.25 with a median value of 2.44. Spearman’s correlation coefficient between the difference and the mean of the two methods was 0.10 (P = 0.58), indicating that there was no systematical error in TEE calculated from the eMLTPAQ. Among the girls, however, there seemed to be a positive correlation, i.e., a high TEEDLW increased the underestimation in the TEEeMLTPAQ (r = 0.55, P = 0.023). This was not the case for the boys (r = 0.11, P = 0.66).
This is the first time the frequently used MLTPAQ has been validated against a criterion method in adolescents. Inclusion of questions about physical inactivity and information about the unreported time to the original MLTPAQ increased the correlation with the TEE from the DLW method. We found a good correlation between the two methods, indicating that eMLTPAQ can be used to rank adolescents’ TEE. However, TEE is greatly underestimated as shown from the Bland-Altman method. This is not a surprising or unusual finding because the method only concerns leisure time activities, not total day activities. Starling et al. (30) used MLTPAQ to assess TEE in 67 older individuals and found MLTPAQ to underestimate EE for physical activity by 56% in women and 62% in men, compared with DLW. MLTPAQ underestimated EE for physical activity by 78% in the current study (71% in boys and 88% in girls). Adding questions about inactivity and information about unreported time reduced the underestimation to 51% (45% in boys and 58% in girls). A lower variation in reported physical activity in girls compared with the boys may explain that the girls underestimated more than the boys.
In the current study we found a correlation of 0.73 between TEE estimated by eMLTPAQ and DLW. Other validation studies, using fitness, accelerometers, energy intake, other questionnaires, or registration of physical activity, have reported correlations ranging from 0.11 to 0.57. In a validation study of 24 adult men, Conway et al. (7) combined the MLTPAQ with the Tecumseh Self-Administered Occupational Activity Questionnaire and energy cost of sleeping. They found, compared with DLW, an r2 of 0.38 (r = 0.62) between the methods and a difference in mean TEE of 1%. However, contrary to our findings, they found that overestimation of TEE was more frequent than underestimation. This may be due to the fact that they used one questionnaire for occupational time and one questionnaire for leisure time, thereby getting information about a larger part of the day compared to a situation using the MLTPAQ only.
We found a PALDLW at 1.83 for boys and 1.77 for girls, which is at the same level as reported in Swedish adolescents by Bratteby et al. (6), with a PAL of 1.92 for the boys and 1.78 for the girls. This may indicate that the participants in the current study constituted a representative sample of Swedish 15-yr-old adolescents. This is also confirmed by predicting the PAL using the newly published data from Hoos et al. (12). In our study population (with a mean age of 15.7), the predicted PAL is calculated to be 1.79, which is very close to the PAL of 1.80 found in the present study. Compared with the only study published using the MLTPAQ in adolescents (19), we found a shorter LTPA duration. We reported a median of 35 min·d−1, whereas Reeder et al. (19) reported a mean participation time of 84 min·d−1. If we disregard that our data did not follow a normal distribution and look at mean participation time in the current study, it is 50 min·d−1 (Table 2). Reeder et al.’s study was performed in New Zealand, and that population might be more physically active than in Northern Europe. A more probable explanation to the conflicting results is the time gap between the studies. Our study was executed in the year 2002, whereas the New Zealand study was performed in 1987, and during these 15 yr a decrease in physical activity has been suggested (10,23). In contrast to the current study where interviews asking for physical activities were performed, Reeder et al. used a cue card to prompt recall of activities, which may have contributed to the higher activity participation reported.
We used the predicted values of BMR that correlated and fitted best with the measured BMR in the calculation of TEE from eMLTPAQ. We found that most of the prediction equations overestimated the BMR, compared with the results from indirect calorimetry, as others have presented earlier (8,21,33). The equation from Molnar et al., however, gave a lower BMR compared with the indirect calorimetry (P < 0.05). This is one of the contributing factors to the underestimation in TEE shown in Figure 2. An alternative approach would have been to use the measured BMR in the TEE calculations from eMLTPAQ, which probably would have resulted in a smaller discrepancy between TEEeMLTPAQ and TEEDLW. However, in epidemiological studies of physical activity, a predicted BMR is a more realistic value to obtain, compared with indirect calorimetry.
Another factor of decisive importance is the choice of MET values for the different activities reported by the students. We have used the MET factors suggested by Ainsworth et al. (1) that are based on published and unpublished studies of energy costs of physical activities. Torun (32) concluded that from the age of 15 yr, the energy costs of physical activities are similar to those of adults. Still, we have to consider that even if skateboarding, for example, has a given MET value of 5.0, there are a number of different ways to do skateboarding, probably with a wide range of intensities. All time-reported skateboarding is given a MET value of 5.0, which may be one of the largest errors in studies transforming physical activity into energy expenditure.
To try to limit the size of the unreported time, we added questions about the physical inactive time (sleeping time, TV/video/computer time, schoolwork at home time) to the original MLTPAQ. To calculate the TEE, we choose the MET value 1.5 for the unreported time, as we supposed that the unreported time mainly would be composed by school time. The DLW results showed that this was too low and suggested a median MET value of 2.44 for the unreported time. However, there are large individual differences ranging from a MET of 1.26 to 5.25. The typical Swedish school time involves a lot of time in the standing position (to be published), which may be an explanation to the higher observed MET value. Another explanation can of course be that the individuals with the highest calculated MET for unreported time may have had larger difficulties to remember and report the true amount of LTPA as well as the inactive time. This was the fact for the three individuals (two boys and one girl) with the highest calculated MET for unreported time (5.25, 4.31, and 3.83, respectively). They were the single three individuals reporting the longest daily duration of watching television (8 h·d−1, 5.7 h·d−1, and 4.6 h·d−1, respectively). Either they overreported their time spent watching television or they watched less television during the two DLW weeks than normal. The three lowest calculated MET for unreported time was 1.26, 1.76, and 1.78 (three boys). The student with the lowest calculated MET value for unreported time possibly overreported his LTPA (3.2 h·d−1), whereas 1.76 and 1.78 is close to 1.5, which we predicted it to be.
In conclusion, MLTPAQ correlated well with TEE from DLW in adolescents, and inclusion of information about inactivity time increased the correlation. Due to difficulties in estimating the intensity of unreported time and an underestimation of the BMR, eMLTPAQ seems to underestimate TEE. However, we consider eMLTPAQ to be valid in ranking Swedish adolescents’ energy expenditure and to describe patterns of leisure time physical activities.
The authors are grateful to The Ingabritt and Arne Lundberg Foundation for the financial support.
The authors gratefully acknowledge the assistance of Dr. Lars Ellegård and Mrs. Elisabeth Gramatkovski for analyzing and calculating the doubly labeled water results, and Mrs. Annica Alklind for performing the anthropometric measurements.
1. Ainsworth, B. E., W. L. Haskell, A. S. Leon, et al. Compendium of physical activities: classification of energy costs of human physical activities. Med. Sci. Sports Exerc. 25: 71–80, 1993.
2. Albanes, D., J. M. Conway, P. R. Taylor, P. W. Moe, J. Judd. Validation and comparison of eight physical activity
questionnaires. Epidemiology 1: 65–71, 1990.
3. Blair, S. N., H. W. Kohl, N. F. Gordon, R. S. Paffenbarger, Jr. How much physical activity
is good for health? Annu. Rev. Public Health 13: 99–126, 1992.
4. Bland, J. M., D. G. Altman. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307–310, 1986.
5. Booth, M. Assessment of physical activity
: an international perspective. Res. Q. Exerc. Sport 71: 114–120, 2000.
6. Bratteby, L. E., B. Sandhagen, H. Fan, G. Samuelson. A 7-day activity diary for assessment of daily energy expenditure
validated by the doubly labelled water method in adolescents. Eur. J. Clin. Nutr. 51: 585–591, 1997.
7. Conway, J. M., M. L. Irwin, B. E. Ainsworth. Estimating energy expenditure
from the Minnesota Leisure Time Physical Activity
and Tecumseh Occupational Activity questionnaires: a doubly labeled water validation. J. Clin. Epidemiol. 55: 392–399, 2002.
8. De Lorenzo, A., I. I. Bertini, A. Puijia, G. Testolin, C. Testolin. Comparison between measured and predicted resting metabolic rate in moderately active adolescents. Acta Diabetol. 36: 141–145, 1999.
9. Elosua, R., J. Marrugat, L. Molina, S. Pons, E. Pujol. Validation of the Minnesota Leisure Time Physical Activity
Questionnaire in Spanish men: the MARATHOM Investigators. Am. J. Epidemiol. 139: 1197–1209, 1994.
10. Heath, G. W., M. Pratt, C. W. Warren, L. Kann. Physical activity
patterns in American high school students: results from the 1990 Youth Risk Behaviour Survey. Arch. Pediatr. Adolesc. Med. 148: 1131–1136, 1994.
11. Henry, C. J., S. Dyer, A. Ghusain-Choueiri. New equations to estimate basal metabolic rate
in children aged 10–15 years. Eur. J. Clin. Nutr. 53: 134–142, 1999.
12. Hoos, M. B., W. J. M. Gerver, A. D. Kester, K. R. Westerterp. Physical activity
levels in children and adolescents. Int. J. Obes. 27: 605–609, 2003.
13. International Dietary Energy Consultancy Group. The doubly-labelled water method for measuring energy expenditure
: technical recommendations for use in humans. Vienna: International Atomic Energy Agency (IAEA), Publication NAHRES-4, 1990.
14. Jacobs, D. R., B. E. Ainsworth, T. J. Hartman, A. S. Leon. A simultaneous evaluation of 10 commonly used physical activity
questionnaires. Med. Sci. Sports Exerc. 25: 81–91, 1993.
15. Leon, A. S., D. R. Jacobs, G. Debacker, H. L. Taylor. Relationship of physical characteristics and life habits to treadmill exercise capacity. Am. J. Epidemiol. 113: 653–660, 1981.
16. Livingstone, B. Epidemiology of childhood obesity in Europe. Eur. J. Pediatr. 159 (Suppl. 1): S14–S34, 2000.
17. Molnar, D., S. Jeges, E. Erhardt, Y. Schutz. Measured and predicted resting metabolic rate in obese and nonobese adolescents. J. Pediatr. 127: 571–577, 1995.
18. Murray, C. J., A. D. Lopez. Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet 349: 1498–1504, 1997.
19. Reeder, A. I., W. R. Stanton, J. D. Langley, D. J. Chalmers. Adolescents’ sporting and leisure time physical activities during their 15th year. Can. J. Sport Sci. 16: 308–315, 1991.
20. Richardson, M. T., A. S. Leon, D. R. Jacobs, Jr., B. E. Ainsworth, R. Serfass. Comprehensive evaluation of the Minnesota Leisure Time Physical Activity
Questionnaire. J. Clin. Epidemiol. 47: 271–281, 1994.
21. Rodriguez, G., L. A. Moreno, A. Sarria, J. Fleta, M. Bueno. Resting energy expenditure
in children and adolescents: agreement between calorimetry and prediction equations. Clin. Nutr. 21: 255–260, 2000.
22. Sallis, J. F., B. E. Saelens. Assessment of physical activity
by self-report: status, limitations, and future directions. Res. Q. Exerc. Sport 71: S1–14, 2000.
23. Schlicker, S. A., S. T. Borra, C. Regan. The weight and fitness status of United States children. Nutr. Rev. 52: 11–17, 1994.
24. Schoeller, D. A. Measurement of energy expenditure
in free-living humans by using doubly labeled water. J. Nutr. 118: 1278–1289, 1988.
25. Schoeller, D. A., E. van Santen. Measurement of energy expenditure
in humans by doubly-labeled water method. J. Appl. Physiol. 53: 955–959, 1982.
26. Schofield, W. N. Predicting basal metabolic rate
, new standards and review of previous work. Hum. Nutr. Clin. Nutr. 39 (Suppl. 1): 5–41, 1985.
27. Sjöberg, A., L. Hallberg, D. Höglund, and L. Hulthén. Meal pattern, food choice, nutrient intake and lifestyle factors in The Göteborg Adolescence Study. Eur. J. Clin. Nutr.
28. Slattery, M. L., D. R. Jacobs. The inter-relationships of physical activity
, physical fitness, and body measurements. Med. Sci. Sports Exerc. 19: 564–569, 1987.
29. Slinde, F., L. Ellegård, A. M. Grönberg, S. Larsson, L. Rossander-Hulthén. Total energy expenditure
in underweight patients with severe chronic obstructive pulmonary disease living at home. Clin. Nutr. 22: 159–165, 2003.
30. Starling, R. D., D. E. Matthews, P. A. Ades, E. T. Poehlman. Assessment of physical activity
in older individuals: a doubly labeled water study. J. Appl. Physiol. 86: 2090–2096, 1999.
31. Taylor, H. L., D. R. Jacobs, Jr., B. Schucker, J. Knudsen, A. S. Leon, G. Debacker. A questionnaire for the assessment of leisure time physical activities. J. Chronic Dis. 31: 741–755, 1978.
32. Torun, B. Energy cost of various physical activities in healthy children. In: Activity, Energy Expenditure
and Energy Requirements of Infants and Children, B. Scürch and N. S. Scrimshaw (Eds.). Lausanne: International Dietary Energy Consultancy Group, 1989, pp. 139–184.
33. van der Ploeg, G. E., S. M. Gunn, R. T. Withers, A. C. Modra, J. P. Keeves, B. E. Chatterton. Predicting the resting metabolic rate of young Australian males. Eur. J. Clin. Nutr. 55: 145–152, 2001.
34. World Health Organisation, Fao, and Unu. Energy and protein requirements. Geneva: WHO, Technical Report Series 724, 1985.