Obesity increases the risk of cardiovascular disease and type 2 diabetes. It is recognized as one of the leading worldwide health problems (8,28), as the proportion of obese adults has increased dramatically in developed countries over the past two decades (12,26). Physical activity is considered important in the prevention and treatment of obesity. Inverse relationships between adult obesity and the level of physical activity have usually been reported in cross-sectional studies (4,6,15,20). However, observations regarding the relationships between adolescent physical activity and adult obesity have been inconsistent (2,9,11,14,16,18,21,25,27). Two recent studies examined the effects of changes in physical activity from adolescence to adulthood on adult obesity (9,21). Both studies found that increased prevalence of obesity in adulthood was associated with low levels of leisure-time physical activities in adolescence. In particular, patterns of decreasing activity and persistent inactivity seemed to predict adult obesity. The purpose of the present study was to test the hypothesis that obesity indices measured in young adults are related to their level of physical activity in childhood and adolescence and especially to changes in physical activity from youth to adulthood.
Data were obtained from the Cardiovascular Risk in Young Finns Study, which is a population-based study consisting of a series of surveys of six cohorts born in 1962, 1965, 1968, 1971, 1974, and 1977. The ages were 3, 6, 9, 12, 15, and 18 yr in 1980. All the participants were randomly selected (N = 4320) in the five Finnish university cities with medical schools (Helsinki, Kuopio, Oulu, Tampere, and Turku) and their surrounding communities. The same participants were followed up in 1983, 1986, 1992, and 2001. Restricted sampling was also done in 1989. Physical activity was assessed in all surveys. A more complete description of the study has been reported previously (17,29). The study was reviewed and approved by the ethics committee of each of the five participating universities.
We included the four oldest age cohorts (9, 12, 15, and 18 yr in 1980) in the present analysis. Thus, subjects aged 3 and 6 yr at baseline were excluded. This was due to the different method used to assess physical activity in the two youngest age groups. A total of 2889 boys and girls aged 9-18 yr were invited to participate in the study in 1980, and 2309 (80%) subjects participated and completed the physical activity questionnaire. From this sample, 1665 (72%) participated and completed the questionnaire in 2001. Due to lack of information on obesity indices, the final sample size in this analysis was 1319 subjects (57%; 626 males and 693 females).
Physical activity index.
In 1980, physical activity and participation in sports were assessed using a self-report questionnaire administered individually in conjunction with a physical examination. We assessed the frequency and intensity of leisure-time physical activities, participation in sports-club training, participation in sports competitions, usual way of spending leisure time, and the way of commuting to school (23). A physical activity index (PAI) was computed using these variables. In 2001, the questions assessed the intensity of physical activity, frequency of vigorous physical activity, hours spent on vigorous physical activity, average duration of a physical activity session, and participation in organized physical activity. The original questions relating to the variables of 1980 and 2001 and their scoring and recoding for both PAI are presented elsewhere (24,29). The questionnaires in youth and adulthood were not identical. The main difference concerned the assessment of supervised training. Activities such as sports-club training and sport competitions were assessed in detail only in childhood and adolescence, as these are common activities in young people but not in adults. Despite these differences, the tracking correlations from youth to young adulthood were of the same magnitude as the interage correlations in youth. Cronbach's alpha for the PAI in the 1980 data ranged from 0.49 to 0.76 for males and from 0.44 to 0.69 for females. In the 2001 data, Cronbach's alpha for the PAI varied from 0.74 to 0.85 for males and from 0.59 to 0.85 for females. Reliability showed a consistent improvement with advancing age. The test-retest reliabilities with the PAI values of the previous years (1980-1992), estimated by the stability coefficients of the simplex model, were > 0.70 (22).
A subsample of participants (N = 102) underwent maximal cycle ergometer exercise testing using a protocol with increments of loading of 20-25 W·min−1 in 2001. Three indicators of exercise capacity (indicating fitness) were measured: estimated maximal oxygen uptake (V̇O2max), mean workload attained during the last 4 min of the test (Wlast4), and hypothetical maximal workload sustainable for 6 min (Wmax6) (3). The validity of the physical activity measurements was studied by correlating childhood and adulthood PAI with the indicators of exercise capacity. The correlation coefficients ranged from r = 0.20 to 0.53, and were strongest between the PAI and Wmax6. In both sexes, childhood PAI correlated with Wmax6: r = 0.39 (P = 0.027) and r = 0.33 (P = 0.044) for women and men, respectively. Similarly, the correlations between adulthood PAI and Wmax6 were r = 0.49 (P = 0.001) and r = 0.53 (P < 0.001) for women and men, respectively. A significant correlation was also found between adulthood PAI and waist circumference, in both sexes, r = −0.14 (P = 0.001) and r = −0.12 (P = 0.001) for women and men, respectively (24).
Participants were divided into three groups according to the distribution of the PAI values. Those who were in the 66th percentile or above were labeled "active,"; participants between the 34th and 65th percentiles were "moderately active,"; and those below the 34th percentile were "inactive."; The same categories were formed in both the 1980 and 2001 data. The cutoff points of the PAI values in both measurements were 5-7 for "inactive,"; 8-11 for "moderately active,"; and 12-15 for "active."; To study the effect of persistent physical activity and changes of activity on adult obesity, the participants were classified into four groups on the basis of the three activity tertiles described above. Participants who belonged to "active"; or "moderately active"; groups both in 1980 and 2001 were labeled "persistently active."; Those who increased their activity from 1980 to 2001, either from "inactive"; to "moderately active"; or "active"; or from "moderately active"; to "active,"; were "increasingly active."; Participants who decreased their activity either from "active"; to "moderately active"; or "inactive"; or from "moderately active"; to "inactive"; were "decreasingly active."; Those who belonged to the "inactive"; group in both 1980 and 2001 were "persistently inactive.";
In 1980, weight was measured with a Seca scale (Vogel & Halke, Hamburg, Germany), and height was measured with a Seca anthropometer. Body mass index (BMI) was calculated as weight (kg)/height (m2). Skinfold thickness was measured at biceps, triceps, and subscapular sites (7). The sum of these three skinfold measures (sum of skinfolds (SSF)) were used to estimate body fatness. In 2001, waist circumference (WC) was measured in duplicate to an accuracy of 1 mm (1,19).
Overweight and obesity in children and adolescents were defined using the international age- and sex-specific cutoff points (5): 19.10 kg·m−2 in boys and 19.07 kg·m−2 in girls at age 9 yr; 21.22 kg·m−2 in boys and 21.68 kg·m−2 in girls at age 12 yr; 23.29 kg·m−2 in boys and 23.94 kg·m−2 in girls at age 15 yr; and 25.0 kg·m−2 in both sexes at age 18 yr. In adults, overweight was defined as a BMI between 25.0 and 29.9 kg·m−2 and obesity as a BMI ≥ 30.0 kg·m−2 (13). Mild abdominal obesity in adults was defined as a WC of 940-1019 mm in men and 800-879 mm in women. Severe abdominal obesity was defined as a WC of ≥ 1020 mm in men and ≥ 880 mm in women. These cutoff points for overweight and obesity according to BMI and WC are widely used in identifying adults with an increased risk of comorbidities (28).
In adults, the self-administered questionnaire included questions on education, occupation, place of residence (urban vs rural), marital status, number of children, and smoking habits. Those smoking daily were considered smokers.
The relationship between youth and adult physical activity and changes in physical activity and adult obesity were analyzed using ANOVA. To evaluate the associations between changes in physical activity and the prevalence of obesity in adults, a multinomial logistic regression analysis was applied, with persistent physical activity as the reference group. Adjustment for potential confounding variables was made to evaluate how the changes in physical activity were independently related to the prevalence of obesity. Data were analyzed for men and women separately. Level of significance level was P < 0.05. All the statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS Inc., version 12, Chicago, IL).
During the 21-yr follow-up, 33.1% of men and 32.0% of women were classified as persistently active, 27.3% of men and 30.3% of women were increasingly active, 28.1% of men and 30.3% of women became decreasingly active, and 11.5% of men and 7.4% of women were classified as persistently inactive. The prevalence of overweight (BMI 25.0-29.9 kg·m−2) was 41.4% in adult men and 24.7% in adult women. The prevalence of obesity (BMI ≥ 30 kg·m−2) was 14.7% in adult men and 12.0% in adult women (Table 1).
Physical activity in youth was inversely associated with both measures of obesity in youth in females but only with BMI in males. Physical activity in youth was not associated with adult obesity. Adult physical activity was related to both measures of adult obesity among females but only with WC among males (Table 2).
The effects of the changes in physical activity from childhood and adolescence to adulthood on adult BMI and WC are shown in Table 3. In both sexes, the participants who were persistently or increasingly active had lower WC values than those who were decreasingly active.
The results of the multiple logistic regression analysis showing the influence of physical activity changes on overweight and obesity measured by BMI are presented in Table 4. Women who had been decreasingly active from youth to adulthood had higher probability of being overweight (OR = 1.79, CI = 1.02-3.16) and obese (OR = 2.09, CI = 1.03-4.26) than women who had been persistently active. These relationships remained significant after adjusting for confounding variables. In addition, women who had been persistently inactive had higher probability of being overweight than women who had been persistently active.
The associations between the physical activity tracking patterns and the risk of abdominal obesity are shown in Table 5. Both men and women who had been decreasingly active were more likely to have mild and severe abdominal obesity in adulthood than those who had been persistently active. In women, these associations remained significant after adjusting for potential confounders.
We found that physical activity history over 21 yr was significantly related to abdominal obesity in men and women in unadjusted analyses. Subjects who reported that their physical activity level had decreased from youth to adulthood were at increased risk of being obese in adulthood compared with subjects who reported being persistently active or increasingly active. In women, these relationships remained significant even after adjustment for potential confounders. These observations suggest that changes in physical activity patterns during the lifetime may play an important role in the development of adult obesity in women.
In general, the results support the findings of many previous cross-sectional studies (4,6,15) and longitudinal studies (9,18,21,27) on the relationship between physical activity and obesity. Two recent studies examined how changes in physical activity from youth to adulthood relate to the risk of obesity. Hasselström et al. (9) showed in an 8-yr follow-up study that a change in physical activity from ages 17 to 25 was inversely related to WC and fatness in men (9). Tammelin et al. (21) performed a 17-yr follow-up study and found that becoming inactive during the transition from ages 14 to 31 was associated with overall obesity in both sexes and with severe abdominal obesity in women (21). In the present study, being decreasingly active seemed to be a stronger risk factor for obesity than being persistently inactive. This may be related to a general adverse change in health habits associated with the reduction of physical activity level. It has been hypothesized that increasing one's physical activity might reduce weight gain through increased energy expenditure or favorable changes in adipose tissue (20), while decreasing one's physical activity might increase sedentary behaviors and reduce energy expenditure (15), especially when combined with the increased body fat distribution caused by increased energy intake and unhealthy dietary habits (12).
An unexpected result was that being persistently inactive was not associated with a higher level of adult obesity compared with the persistently active reference subjects. This finding conflicts with the results of two other studies that found that persistent inactivity predicted an increased level of obesity (9,21). In those studies, follow-up time was shorter (8 and 17 yr) and the participants were younger (25 and 31 yr as adults) than those in the present study. In the present study, however, the number of subjects in the persistently inactive group was much smaller than in other groups.
Obesity tracks significantly from youth to adulthood (13,16). We observed that physical activity in youth was inversely related to obesity in youth. Therefore, considering obesity in youth as a confounding variable was important. The relationship between activity history and adult obesity remained significant after controlling for youth obesity and other confounders in women. This suggests that changes in physical activity patterns during the lifetime may have an independent effect on the development of adult obesity in women. In men, the relationships between physical activity patterns and obesity were diluted after controlling for potential confounders. This may indicate that part of the effect of physical activity change on obesity in men is mediated by concomitant changes in sociodemographic and lifestyle habits, which in turn are known to be associated with obesity (12,18).
Physical activity patterns were related to BMI in women but not in men. This sex difference may be explained by differences in body constitution. Women have a higher percentage of body fat than men (12), and men have more skeletal muscle in the upper body, both in absolute terms and relative to body mass, than women (10). Thus, BMI may be a better marker of adipose tissue in women than in men.
There are some limitations in the present study that should be considered when interpreting the results. Diet is an important determinant of obesity. Unfortunately, detailed information on diet was not available from the majority of the study subjects. Without information on diet, it is not possible to assess whether diet could have confounded the results. In the Young Finns Study, the level of physical activity has been assessed at several time points during the follow-up. To have a large enough study sample with statistical power, only data collected in 1980 and 2001 were included in this analysis. We have previously shown, however, that the interage stability coefficients estimated at 3-yr intervals by the Simplex model (sample tracking correlations corrected by reliability coefficients) varied from 0.65 to 0.99 among males and from 0.53 to 0.97 among females (22). This indicates that physical activity behavior is rather stable in youth and suggests that one measure of physical activity during childhood and adolescence may represent the overall activity in youth reasonably well.
The questionnaires for children and adults were not identical. When measuring physical activity using a questionnaire, it is difficult to harmonize the questions targeted at different age groups. For example, the perceived intensity of equally intense physical activity may differ between children and adults. Although children participate in activities that seem very strenuous, their perceived intensity is low and only increases as they get older (23). Adults can more reliably assess the intensity of their physical activity. Therefore, in adults, questions concerning hours and frequency of intensive physical activity may be more relevant than in children. Children, on the other hand, are able to report reliably the frequency of their participation in organized sport, which plays an important role in physical activity among children. Although the questionnaires for young people and adults were different, the tracking correlations from youth to young adulthood were similar to the interage correlations in youth. This suggests that physical activity can be measured using different kinds of questions, provided that the questions are relevant.
In summary, we found that physical activity history over 21 yr was significantly related to abdominal obesity in men and women. These findings suggest that changes in physical activity patterns during the lifetime may play an important role in the development of adult obesity. Persistent participation in regular physical activity from youth to adulthood could aid in the prevention of abdominal obesity in adulthood.
This study was financially supported by the Academy of Finland (grant 53392), Social Insurance Institution of Finland, Ministry of Education, Turku University Foundation, Special Federal Grants for Turku University Hospital, Juho Vainio Foundation, Finnish Foundation of Cardiovascular Research, Emil Aaltonen Foundation, Finnish Medical Foundation, and Finnish Cultural Foundation.
1. Åkerblom, H. K., J. Viikari, O. T. Raitakari, and M. Uhari. Cardiovascular risk in young Finns study: general outline and recent developments. Ann. Med.
31(suppl 1):45-54, 1999.
2. Andersen, L. B., and J. Haraldsdottir. Tracking of cardiovascular disease risk factors including maximal oxygen uptake and physical activity from late teenage to adulthood. An 8-year follow-up study. J. Intern. Med.
3. Arstila, M., O. Impivaara, and J. Mäki. New ergometric reference values for clinical exercise tests. Scand. J. Lab. Invest.
4. Blair, S. N., and C.Bouchard. Physical activity and obesity: American College of Sport Medicine consensus conference. Med. Sci. Sports Exerc.
5. Cole, T. J., M. C.Bellizzi, K. M. Flegal, and W. H. Dietz. Establishing a standard definition for child overweight and obesity world wide: international survey. BMJ
6. Cooper, A. R., A. Page, K. R. Fox, and J. Misson. Physical activity patterns in normal, overweight and obese individuals using minute-by-minute accelerometry. Eur. J. Clin. Nutr.
7. Dahlström, S., J. Viikari, H. K. Åkerblom, et al. Atherosclerosis precursors in Finnish children and adolescents. II. Height, weight, body mass index, and skinfolds and their correlation to metabolic variables. Acta Paediatr. Scand.
8. DiPietro, L. Physical activity in the prevention of obesity: current evidence and research issues. Med. Sci. Sports Exerc.
9. Hasselström, H., S. E. Hansen, K. Froberg, and L. B. Andersen. Physical fitness and physical activity during adolescence as predictors of cardiovascular disease risk in young adulthood. Danish Youth and Sports Study. An eight-year follow-up study. Int. J. Sports Med.
23(suppl 1):S27-S31, 2002.
10. Janssen, I., S. B. Heymsfield, Z. Wang, and R. Ross. Skeletal muscle mass and distribution in 468 men and women aged 18-88 yr. J. Appl. Physiol.
11. Kemper, H. C., G. B. Post, J. W. Twisk, and W. van Mechelen. Lifestyle and obesity in adolescence and young adulthood: results from the Amsterdam Growth and Health Longitudinal Study (AGAHLS). Int. J. Obes. Relat. Metab. Disord.
12. Lahti-Koski, M. Body mass index and obesity among adults in Finland. Academic dissertation. Publications of the National Public Health Institute (KTL) A 12/2001, Helsinki: Hakapaino Oy, 2001.
13. Laitinen, J., C. Power, and M. R. Järvelin. Family social class, maternal body mass index, childhood body mass index, and age at menarche as predictors of adult obesity. Am. J. Clin. Nutr.
14. Lefevre, J., R. Philippaerts, K. Delvaux, et al. Relation between cardiovascular risk factors at adult age, and physical activity during youth and adulthood: the Leuven Longitudinal Study on Lifestyle, Fitness and Health. Int. J. Sports Med.
15. Martinez-Gonzalez, M. A., J. A. Martinez, F. B. Hu, M. J. Gibney, and J. Kearney. Physical inactivity, sedentary lifestyle and obesity in the European Union. Int. J. Obes. Relat. Metab. Disord.
16. Parsons, T. J., C. Power, S. Logan, and C. D. Summerbell. Childhood predictors of adult obesity: a systematic review. Int. J. Obes. Relat. Metab. Disord.
17. Pulkki-Raback, L., M. Elovainio, M. Kivimäki, O. T. Raitakari, and L. Keltikangas-Järvinen. Temperament in childhood predicts body mass in adulthood: the Cardiovascular Risk in Young Finns Study. Health Psychol.
18. Raitakari, O. T., K. V. K. Porkka, S. Taimela, R. Telama, L. Räsänen, and J. Viikari. Effects of persistent physical activity and inactivity on coronary risk factors in children and young adults-the cardiovascular risk in young Finns study. Am. J. Epidemiol.
19. Raitakari, O. T., M. Juonala, M. Kähönen, et al. Cardiovascular risk factors in childhood and carotid artery intima-media thickness in adulthood. The Cardiovascular Risk in Young Finns Study. JAMA
20. Ross, R., and I. Janssen. Physical activity, total and regional obesity: dose-response considerations. Med. Sci. Sports Exerc.
33(6 suppl):S521-S527, 2001.
21. Tammelin, T., J. Laitinen, and S. Näyhä. Change in the level of physical activity from adolescence into adulthood and obesity at the age of 31 years. Int. J. Obes. Relat. Metab. Disord.
22. Telama, R., E. Leskinen, and X. Yang. Stability of habitual physical activity and sport participation: a longitudinal tracking study. Scand. J. Med. Sci. Sports
23. Telama, R., J. Viikari, I. Välimäki, et al. Atherosclerosis precursors in Finnish children and adolescents. X. Leisure-time physical activity. Acta Paediatr. Scand.
24. Telama, R., X. Yang, J. Viikari, I. Välimäki, O. Wanne, and O. T. Raitakari. Tracking physical activity from childhood to adulthood: a 21-year follow up study. Am. J. Prev. Med.
25. Twisk, J. W., H. C. Kemper, and W. van Mechelen. The relationship between physical fitness and physical activity during adolescence and cardiovascular disease risk factors at adult age. The Amsterdam Growth and Health Longitudinal Study. Int. J. Sports Med.
26. U.S. Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General
. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 199; pp. 133-135.
27. Williamson, D. F., J. Madans, R. F. Anda, J. C. Kleinman, H. S. Kahn, and T. Byers. Recreational physical activity and ten-year weight change in a US national cohort. Int. J. Obes. Relat. Metab. Disord.
28. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation (WHO Technical Report Series 894)
. Geneva: World Health Organization, 2000; pp. 5-68.
29. Yang, X., R. Telama, L. Laakso, and J. Viikari. Children's and adolescents' physical activity in relation with living environment, parents' physical activity, age and gender. Acta Kinesiol. Universitatis Tartuensis