BMC Public Health (Jan 2023)

Association between physical activity and longitudinal change in body mass index in middle-aged and older adults

  • Laura Cleven,
  • Jeremy A. Syrjanen,
  • Yonas E. Geda,
  • Luke R. Christenson,
  • Ronald C. Petersen,
  • Maria Vassilaki,
  • Alexander Woll,
  • Janina Krell-Roesch

DOI
https://doi.org/10.1186/s12889-023-15119-7
Journal volume & issue
Vol. 23, no. 1
pp. 1 – 9

Abstract

Read online

Abstract Background In middle-aged and particularly older adults, body mass index (BMI) is associated with various health outcomes. We examined associations between physical activity (PA) and longitudinal BMI change in persons aged ≥ 50 years. Methods The sample included 5159 community-dwelling individuals aged ≥ 50 years (50.5% males, mean (SD) age 73.0 (10.2) years at baseline) who were enrolled in the Mayo Clinic Study of Aging (MCSA). Participants had information on PA within one year of baseline assessment, BMI at baseline, and potential follow-up assessments (mean (SD) follow-up 4.6 (3.7) years). Linear mixed-effect models were used to calculate the association between PA (moderate-vigorous physical activity, MVPA; and all PA composite score) and the longitudinal change in BMI, adjusted for baseline age, sex, education and medical comorbidities. In addition to interactions between years since baseline and PA, we also included 2- and 3-way interactions with baseline age to further assess whether age modifies the trajectory of BMI over time. Results We observed a decrease in BMI among participants engaging at a mean amount of PA (i.e., MVPA: 2.7; all PA: 6.8) and with a mean age (i.e., 73 years) at baseline (MVPA: estimate = -0.047, 95% CI -0.059, -0.034; all PA: estimate = -0.047, 95% CI -0.060, -0.035), and this decline is accelerated with increasing age. Participants with a mean age (i.e., 73 years) that engage at an increased amount of MVPA or all PA at baseline (i.e., one SD above the mean) do not decrease as fast with regard to BMI (MVPA: estimate = -0.006; all PA: estimate = -0.016), and higher levels of MVPA or all PA at baseline (i.e., two SD above the mean) were even associated with an increase in BMI (MVPA: estimate = 0.035; all PA: estimate = 0.015). Finally, MVPA but not all PA is beneficial at slowing BMI decline with increasing age. Conclusion PA, particularly at moderate-vigorous intensity, is associated with slower decline in longitudinal BMI trajectories. This implies that engaging in PA may be beneficial for healthy body weight regulation in middle and late adulthood.

Keywords