When Adolescents Drop the Ball Sustainability of Physical Activity in Youth Mathieu Bélanger, PhD, Katherine Gray-Donald, PhD, Jennifer O’Loughlin, PhD, Gilles Paradis, MD, MSc, James Hanley, PhD Background: A majority of youth do not attain the recommended levels of physical activity. To develop interventions that will be more efficient at enabling healthy levels of physical activity during adolescence, a better understanding is needed about which specific types of physical activity adolescents are more likely to sustain and when they tend to stop participating in each specific type of activity. Methods: From 1999 to 2005, 1276 adolescents, initially aged 12–13 years, completed a 7-day physical activity recall every 3 months during each of 5 years of secondary school. The prevalence of participation in each of 29 specific physical activities in each of Grades 7–11 was computed. Survival analyses were used to estimate when adolescents who reported each activity at baseline tended to discontinue the activity. All analyses were conducted in 2008. Results: The prevalence of participation in most activities declined over the 5 years; it did not increase for any activity. Within 2 years of baseline, the majority of adolescents discontin- ued participation in most activities in which they had reported participation at baseline. Sustained participation in a specific activity related both to its intensity (90%, 73%, and 40% of girls and 77%, 86%, and 60% of boys sustained participation in light-, moderate-, and vigorous-intensity activities, respectively) and its format (41% and 89% of girls and 69% and 90% of boys sustained participation in team and individual physical activities, respectively). Conclusions: Participation in almost all physical activities declined during adolescence. The time of discontinuation varied across activity types. Promoting activities that attract and sustain secondary school students may improve physical activity levels throughout adolescence. (Am J Prev Med 2009;xx(x):xxx) © 2009 American Journal of Preventive Medicine Introduction A lthough the benefits of regular physical activity are well established, levels of physical activity remain below recommended levels. 1,2 Physical inactivity is more prevalent than any other modifiable risk factor, 3,4 and almost all youth, including 90% in Canada, do not attain the recommended 90 minutes of moderate-to-vigorous intensity physical activity per day. 5 This could result in a substantial future burden on health, given that low levels of physical activity during youth may lead to physical inactivity during adulthood. 6,7 Adolescence is generally characterized by sharp de- clines in physical activity 8,9 and by changes in the types of activity engaged in. 10 –12 With the possible exception of household chores and active transportation (i.e., walking or bicycling) to and from school, 10,11 few young individuals sustain involvement in specific physical ac- tivities throughout adolescence. 13 However, participa- tion in specific physical activities in adulthood is con- siderably more likely among individuals who engaged in the activity during adolescence. 14 The indication that participation in specific types of physical activity declines at different rates originates from cross-sectional studies, 15 repeat cross-sectional studies, 16 and longitudinal studies that usually have few follow-up waves. 10 –12,14,17 These studies are limited to reports of the proportion of adolescents who engage in different types of physical activity according to age or to indices of the correlation between early and later participation in specific activities. Few studies report when adolescents discontinue participation in specific types of physical activity. From the Centre de formation médicale du Nouveau-Brunswick, Université de Moncton, Université de Sherbrooke, and Beauséjour Research Centre, Regional Health Authority A (Bélanger), Moncton, New Brunswick; the Department of Epidemiology, Biostatistics and Occupational Health (Bélanger, Gray-Donald, Paradis, Hanley) and the School of Dietetics and Human Nutrition (Gray-Donald), McGill University; the Research Institute of the McGill University Health Centre (Paradis); the Department of Social and Preventive Medicine, Université de Montréal and Centre de recherché du Centre Hospi- talier de l’Université de Montréal (O’Loughlin); and Institut national de santé publique du Québec (O’Loughlin, Paradis), Montreal, Quebec, Canada Address correspondence and reprint requests to: Mathieu Bé- langer, PhD, Centre de formation médicale du Nouveau-Brunswick Pavillon J.-Raymond-Frenette, Université de Moncton, Moncton NB Canada E1A 3E9. E-mail: mathieu.belanger@umoncton.ca. UNDER EMBARGO UNTIL MAY 26, 2009, 12:01 AM LOCAL TIME 1 Am J Prev Med 2009;xx(x) 0749-3797/09/$–see front matter © 2009 American Journal of Preventive Medicine • Published by Elsevier Inc. doi:10.1016/j.amepre.2009.04.002