Background:
Birth weight and physical activity are independently associated with cardiometabolic health outcomes. Low or high birth weight are indicators of adverse prenatal development, which may alter physiological response to physical activity later in life. However, few studies have explored the potential interaction between birth weight and physical activity as determinants of cardiometabolic health.
Objective:
We evaluated the hypothesis that high or low birth weight modifies the association of early life physical activity with cardiovascular disease or diabetes later in life.
Methods:
We analyzed data from the National Longitudinal Study of Adolescent and Adult Health (Add Health), a nationally representative cohort of US adolescents followed into adulthood (
n
=20,745) with four data collection waves between 1994 and 2008. Outcomes were assessed in early adulthood: (1) predicted 30-year cardiovascular disease (CVD) risk, computed by a validated algorithm based on objective measures, and (2) prevalent pre-diabetes and diabetes. Using gender-stratified multivariable regression on multiply imputed data, we modeled (1) log-transformed 30-year CVD risk (linear regression) and (2) prevalent pre-diabetes and diabetes (PDM/DM; ordinal regression) each as a function of birth weight (low, normal, high; LBW, NBW, HBW) and self-reported moderate-to-vigorous physical activity frequency (MVPA) in adolescence and young adulthood, adjusting for age, smoking, and sociodemographic factors.
Results:
A greater proportion of women born at LBW had diabetes than NBW and HBW women (10.8% versus 5.9% and 5.4%, respectively). In adjusted analyses, MVPA in adolescence (MVPA1) and early adulthood (MVPA3) were not significantly associated with predicted CVD risk and prevalent pre-diabetes diabetes in men or women overall. However, greater MVPA1 was associated with lower predicted 30-year CVD risk in HBW females (estimated effect coefficient -0.02 [95% CI: -0.03, -0.005,
p
=0.02],
p
=0.05 for HBWхMVPA1 interaction), and the HBWхMVPA1 interaction on PDM/DM approached significance in females (
p
=0.12). In females and males of LBW or NBW, MVPA1 was not significantly associated with predicted 30-year CVD risk or PDM/DM and LBWхMVPA1 interactions were not significant.
Conclusions:
Greater adolescent physical activity was most strongly associated with lower 30-year CVD risk in young women born at HBW. A similar association with prevalent DM/PDM approached significance, with greater adolescent physical activity most strongly associated in HBW women. Females born at HBW may be especially sensitive to the effects of physical activity on reducing risk of cardiometabolic disease later in life, with important implications for disease prevention and health policy.