Association between physical activity and visceral adiposity index (VAI) in U.S. population with overweight or obesity: a cross-sectional study.

BACKGROUND

Previous studies have revealed the effects of different physical activity (PA) types on visceral adipose tissue (VAT) accumulation in individuals with overweight/obesity. However, the independent association (especially the dose-response relationship) between PA and VAT in individuals with and without overweight/obesity remains less explored. Visceral adiposity index (VAI), calculated from waist circumference, body mass index (BMI), triglyceride and high-density lipoprotein cholesterol, is a novel indicator of VAT. This study aims to elucidate the association between PA and VAI in participants with and without overweight/obesity.

METHODS

Participants who are overweight or obese and with complete data on VAI, PA, and other essential covariates from the National Health and Nutrition Examination Survey (NHANES) database (2015-2018) were included in this study. PA was evaluated by the PA questionnaire and converted into metabolic equivalent task (MET) hours per week (MET-h/wk) based on the suggested MET scores. Multivariate linear regression models were used to identify the association between PA and VAI. Subgroup analyses, combined with interaction tests and restricted cubic spline (RCS) regression analyses, were utilized to explore the stability and nonlinearity of PA-VAI association, respectively.

RESULTS

A total of 4, 312 participants with complete data on PA and VAI was included in this study, with 3, 441 of them being overweight or obese. After adjusting for all potential covariates, increased PA was found to be significantly associated with remarkable lower level of VAI in all participants (β = -0.0004, P = 0.003), participants with (β = -0.0013, P = 0.012) and without (β = -0.0004, P = 0.003) overweight/obesity. Subgroup analyses and interaction tests revealed that the PA-VAI association was not modified by other covariates in individuals with overweight/obesity. Furthermore, RCS analyses revealed that PA was significantly, linearly and negatively associated with VAI in all participants, participants with and without overweight/obesity (all P < 0.05, all P for nonlinearity > 0.05). Noteworthily, as opposed to individuals without overweight/obesity, PA was significantly associated with lower VAI in participants with overweight/obesity after exceeding the threshold of 150 MET-h/wk.

CONCLUSION

Increased PA was significantly associated with lower level of VAI, but a higher level of PA (> 150 MET-h/wk) was needed to obtain significantly lower level of VAI in individuals with overweight/obesity.