Birth weight, cardiometabolic risk factors and effect modification of physical activity in children and adolescents: pooled data from 12 international studies

Abstract

Objectives: Low and high birth weight is associated with higher levels of cardiometabolic risk factors and adiposity in children and adolescents, and increases the risk of cardiovascular diseases, obesity, and early mortality later in life. Moderate-to-vigorous physical activity (MVPA) is associated with lower cardiometabolic risk factors and may mitigate the detrimental consequences of high or low birth weight. Thus, we examined whether MVPA modified the associations between birth weight and cardiometabolic risk factors in children and adolescents.

Methods: We used pooled individual data from 12 cohort- or cross-sectional studies including 9,100 children and adolescents. Birth weight was measured at birth or maternally reported retrospectively. Device-measured physical activity (PA) and cardiometabolic risk factors were measured in childhood or adolescence. We tested for associations between birth weight, MVPA, and cardiometabolic risk factors using multilevel linear regression, including study as a random factor. We tested for interaction between birth weight and MVPA by introducing the interaction term in the models (birth weight x MVPA).

Results: Most of the associations between birth weight (kg) and cardiometabolic risk factors were not modified by MVPA (min/day), except between birth weight and waist circumference (cm) in children (p = 0.005) and HDL-cholesterol (mmol/l) in adolescents (p = 0.040). Sensitivity analyses suggested that some of the associations were modified by VPA, i.e., the associations between birth weight and diastolic blood pressure (mmHg) in children (p = 0.009) and LDL- cholesterol (mmol/l) (p = 0.009) and triglycerides (mmol/l) in adolescents (p = 0.028).

Conclusion: MVPA appears not to consistently modify the associations between low birth weight and cardiometabolic risk. In contrast, MVPA may mitigate the association between higher birth weight and higher waist circumference in children. MVPA is consistently associated with a lower cardiometabolic risk across the birth weight spectrum. Optimal prenatal growth and subsequent PA are both important in relation to cardiometabolic health in children and adolescents.

Fig. 1. Illustration of the interaction between birth weight and MVPA on waist circumference (children) and HDL-cholesterol (adolescents).

Predicted values of cardiometabolic outcomes (a: waist circumference, b: HDL-cholesterol) across birth weight and the 25th, 50th and 75th percentile of MVPA in children (a) and adolescents (b) from regression models with significant interaction between birth weight and MVPA (p < 0.1). Adjusted for highest parental education, sex, age and waist circumference (when not the outcome). HDL high density lipoprotein, MVPA moderate to vigorous physical activity MVPA children: 25th percentile = 39.6 min/day, 50th percentile = 57.7 min/day, 75th percentile=80.0 min/day. MVPA adolescents: 25th percentile = 25.3 min/day, 50th percentile = 39.8 min/day, 75th percentile = 58.7 min/day.

Fig. 2. Illustration of the interaction between birth weight and VPA on diastolic blood pressure (children), LDL-cholesterol (adolescents) and triglycerides (adolescents).

Predicted values of cardiometabolic outcomes (a: diastolic blood pressure, b: LDL-cholesterol, c: triglycerides) across birth weight and the 25th, 50th, and 75th percentile of VPA in children (a) and adolescents (b, c) from regression models with significant interaction between birth weight and VPA (p < 0.1). Adjusted for highest parental education, sex, age (height for diastolic blood pressure) and waist circumference (when not the outcome). HDL high density lipoprotein, LDL low density lipoprotein, MVPA moderate to vigorous physical activity, VPA vigorous physical activity VPA children: 25th percentile = 7.7 min/day, 50th percentile = 14.4 min/day, 75th percentile = 24.5 min/day. VPA adolescents: 25th percentile = 4.5 min/day, 50th percentile = 10.5 min/day, 75th percentile = 20.3 min/day.

Fig. 3. Illustration of the association between birth weight and the clustered cardiometabolic risk score across MVPA (no interaction between birth weight and MVPA).

Predicted clustered cardiometabolic risk score across birth weight and 25th, 50th, and 75th percentile of MVPA from regression model with interaction term (birth weight x MVPA) in children (a) and adolescents (b), p value for interaction between birth weight and MVPA; children p = 0.774, adolescents p = 0.948. Adjusted for highest parental education, waist circumference, sex and age. Clustered cardiometabolic risk score calculated from summing standardized values for MAP (mean arterial blood pressure), triglycerides, LDL/HDL-ratio and HOMA-IR, divided by four (number of variables). MVPA moderate to vigorous physical activity Children MVPA: 25th percentile = 39.6 min/day, 50th percentile = 57.7 min/day, 75th percentile = 80.0 min/day. Adolescents: 25th percentile = 25.3 min/day, 50th percentile = 39.8 min/day, 75th percentile = 58.7 min/day.