Longitudinal Associations Between BMI, Physical Activity, and Healthy Diet

Abstract

An excessive body mass index (BMI) is a growing public health concern resulting from unhealthy lifestyle behaviors. Particularly, a vast body of literature supports significant roles of physical activity (PA) and healthy dietary behaviors (HDB) in reducing BMI. However, a majority of evidence has been accumulated from a cross-sectional data viewpoint which limits our ability to draw causal inferences, and thus, requiring more research efforts to elucidate underlying casual associations based on a longitudinal study design. PURPOSE: This study examined the 7-year longitudinal associations between changes in BMI, PA, and HDB among adults using a parallel latent growth curve model (LGM). METHODS: Prospective, open cohort data collected from 22,885 adults enrolled in workplace wellness programs across over 100 organizations during 2004 through 2013 were used. A healthy cohort between 2004 and 2007 (n=2,579) who provided valid data on BMI, PA, and HDB measures for at least 5 out of 7 follow-up years from the entry was analyzed. Principal component scores (higher indicates better) of PA and HDB were calculated using three questionnaire items, respectively, for each analytic year. Changes in BMI, PA, and HDB across 7 follow-up years were fitted in the conditional parallel LGM after controlling for time invariant covariates using Mplus v7.3. RESULTS: On average, significant increases in BMI and PA were observed across years (SlopeBMI = 0.129, P<.001; SlopePA = 0.008, P=.010), while no significant change in HDB was observed (SlopeHDB = 0.005, P=.124). After controlling for covariates, increases in PA and HDB across years significantly slower the rate of increase in BMI (SlopePA -> SlopeBMI = -1.273, P<.001; SlopeH.Diet -> SlopeBMI = -1.734, P<.001). CONCLUSION: The findings from this study confirm significant roles of PA and HDB to reduce the risk of excessive BMI in later years. Considering that BMI is generally increased as age increases, more systematic efforts to promote PA and HDB among young adults should be made to slow the growth rate of BMI. Further study is also warranted to identify a high risk population who maintains unhealthy lifestyle behaviors across years for developing future intervention strategies.