There are multiple etiologies of obesity, and attempts to curb the rising prevalence of obesity by addressing any single etiology are notoriously unsuccessful. Addressing physical activity, or the lack of it, seems a promising approach because studies clearly indicate that we are a sedentary society. Indeed, physical activity levels decrease as children become older: girls follow this diminishing pattern of activity more than do boys, and African American girls do so more than do white girls (1). Teenaged African American girls report metabolic equivalent (MET representing measurable physical activity) values approximating zero (1), and, as might be expected, increasing body mass indexes (BMIs) are inversely associated with METs in that report. There is a well-reported relation between time spent viewing television and BMI values, but, even if the basis of this relation is unknown, it is reasonable to suppose that the time spent in the sedentary behavior of television viewing is time that could have been spent in physical activity (2).

Despite this understanding of the relation between physical behavior or sedentary behavior on the one hand and obesity on the other, it has been difficult to tie changes in activity patterns to the development of obesity in a cause-and-effect manner when we leave the realm of television viewing. This may be due to the large effects exerted by a subtle change in nutrition or activity as well as to our inability to reliably measure such small changes. In this issue of the Journal, Ekelund et al (3) present a cross-sectional portion of their longitudinal Stockholm Weight Development Study in an attempt to correlate physical activity with weight gain in 17-y-old subjects in Stockholm. Their study uses the Bodpod (an air-displacement plethysmograph) to determine fat mass and fat-free mass, rather than relying solely on BMI, which cannot separate those 2 variables. They found that only 4% of the variation in body fat was related to the level of self-reported physical activity, and they found even this small relation only in males. This group of researchers, as well as others, previously showed that a rather low degree of variance in BMI was due to physical activity (eg, 0.5% variance was due to vigorous physical activity). For the girls in the current study, fat mass was more closely related to the mother’s fat mass than to the girls’ physical activity, and that was also shown in earlier studies (4).

In the current study by Ekelund et al, self-reporting is invoked as the measure of physical activity in the subjects. The authors correctly stated that self-reported activity might reflect errors that were due to the perceptions of the respondents. In this era, when all popular media outlets broadcast the merits of weight control and an active lifestyle (presumably Sweden is similar to the United States in this respect), an inactive subject might inflate his or her perceived activity levels, which would result in overreporting. In the United States, there is evidence that preteenagers will report their activity levels and dietary intake as being more in line with a perceived socially desirable answer than with the actual facts (5). Ekelund et al attempted to validate the self-reported results by testing 11% of the population more objectively with the use of recording accelerometers, and they found good agreement of the results obtained by using the 2 methods. Recalled activity might be the best we can do in a large epidemiologic study, even if we prefer objective measurements and even if the authors have made a reasonable effort to justify their results by using both methods so that the influence of social issues may be lessened in their population.

More than 85% of these authors’ population of Stockholm adolescents had normal BMIs, and only a minority of subjects were in the overweight or obese range, so their data may not represent relations between physical activity and fat mass or BMI in the most severely affected of their obese subjects. The range of METs in 17-y-olds of normal BMI might be too narrow to show the relation in those with the highest BMI, and thus the authors’ results in the normal subjects may not be reflective of results that would be obtained in those with a greater fat mass. Parenthetically, it must be noted that the Swedish population in their study and the general Swedish population have a prevalence of overweight [the Centers for Disease Control and Prevention (CDC) uses the term "at risk for overweight" for BMIs in the 85th–95th percentile for age] of 40% and a prevalence of obesity (the CDC uses the term "overweight" for BMIs >95th percentile for age) of <25% of the prevalences found in 17-y-olds in the United States (6).

Despite these findings, can increased physical activity provide a path out of the increasing prevalence of obesity in children? Clearly, physical activity is not likely to counteract a poor diet. It would take >1–2 h of extremely vigorous activity to counteract a single large-sized (ie, 785 kcal) children’s meal at a fast food restaurant, and there are few children (or adults) who can maintain such a pace; moreover, the balance is only worsened if there are repeated such meals (for the energy content of common foods, including fast foods, see: http://www.nal.usda.gov/fnic/etext/000020.html). In addition, reduced sedentary activity is suggested as an achievable goal, whereas the institution of vigorous physical activity after a sedentary life is not likely to be a maintainable intervention (7). There is no risk to a decrease in sedentary activity or to a reasonable and sustainable increase in physical activity, and there is benefit in terms of improved insulin sensitivity, among other factors. Thus, even if we have difficulty in directly and substantially linking physical activity to fat mass in a study such as that of Ekelund et al, the risk-benefit ratio is so high that there is no justification for not supporting an increasingly active lifestyle. Indeed, the Institute of Medicine recently listed just this approach as one intervention in a comprehensive recommendation (8). It appears that many small changes, rather than one overall solution, will be needed if we are to change our aggregate march to higher BMI values.

References

1. Kimm SY, Glynn NW, Kriska AM, et al. Decline in physical activity in black girls and white girls during adolescence. N Engl J Med 2002;347:709-15.
2. Ludwig DS, Gortmaker SL. Programming obesity in childhood. Lancet 2004;364:226-7.
3. Ekelund U, Neovius M, Linné Y, Brage S, Wareham NJ, Rössner S. Associations between physical activity and fat mass in adolescents: the Stockholm Weight Development Study. Am J Clin Nutr 2005;81:355-60.
4. Ekelund U, Sardinha LB, Anderssen SA, et al. Associations between objectively assessed physical activity and indicators of body fatness in 9- to 10-y-old European children: a population-based study from 4 distinct regions in Europe (the European Youth Heart Study). Am J Clin Nutr 2004;80:584-90.
5. Klesges LM, Baranowski T, Beech B, et al. Social desirability bias in self-reported dietary, physical activity and weight concerns measures in 8- to 10-year-old African-American girls: results from the Girls Health Enrichment Multisite Studies (GEMS). Prev Med 2004;38(suppl):S78-87.
6. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA 2004;291:2847-50.
7. Epstein LH, Saelens BE, Myers MD, Vito D. Effects of decreasing sedentary behaviors on activity choice in obese children. Health Psychol 1997;16:107-13.
8. Koplan JP, Liverman CT, Kraak VA, eds. Preventing childhood obesity: Health in the Balance Committee on Prevention of Obesity in Children and Youth. Washington, DC: Institute of Medicine, 2004.