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

The study investigates whether device-measured moderate-to-vigorous physical activity (MVPA) modifies the relationship between birth weight and cardiometabolic risk factors in youth. Grounded in the DOHaD framework, prior research links low birth weight to later cardiovascular risk and high birth weight to later obesity. Physical inactivity is a known cardiovascular risk factor, while higher-intensity activity is associated with more favorable cardiometabolic profiles in children and adolescents. However, previous studies on whether physical activity modifies birth weight–risk associations are few and show inconsistent results. The purpose here is to test effect modification by MVPA across multiple cardiometabolic outcomes in a large, diverse cohort of children and adolescents.

The introduction summarizes evidence that: (1) low birth weight, as a proxy for fetal growth restriction, is associated with increased adult cardiovascular disease risk; (2) high birth weight predicts higher obesity risk; (3) physical inactivity elevates cardiovascular risk, with substantial mortality risk reductions seen even at lower activity volumes; and (4) in youth, higher-intensity physical activity is consistently linked to lower cardiometabolic risk. Two prior studies on effect modification by physical activity in youth yielded contradictory results and included few outcomes, highlighting the need for larger, more comprehensive analyses.

Design: Pooled analysis of individual participant data from 12 international studies (nine from ICAD 2.0 plus MoBa, PANCS, and ASK), comprising 9,100 children and adolescents. Studies included both prospective birth cohorts and cross-sectional cohorts. Participants: Children and adolescents clustered around ages 9–10 and 15–16 years. For analysis, participants were split by median age into children (≤11.6 years) and adolescents (>11.6 years). Inclusion required data on birth weight, valid accelerometer-derived PA (≥3 valid days), and at least one cardiometabolic outcome. Exposures: Birth weight measured at birth (in several cohort studies) or retrospectively reported by parents (in cross-sectional studies), expressed in kilograms. Device-measured physical activity using ActiGraph accelerometers (uniaxial models GT1M/7164; triaxial GT3X+ re-integrated to uniaxial) collected over 4–7 days. Data harmonized to 60-second epochs; non-wear defined by extended strings of zero counts with allowance for brief interruptions; a valid day required ≥480 minutes wear time; participants needed ≥3 valid days. MVPA threshold defined as ≥2296 counts per minute (average minutes/day). Vigorous PA (VPA, ≥4012 cpm) used in sensitivity analyses. Outcomes: Cardiometabolic risk factors measured at the same time as PA, including systolic and diastolic blood pressure (mean of repeated, resting, automated measurements), fasting lipids (LDL-cholesterol, HDL-cholesterol, triglycerides), fasting glucose and insulin with insulin resistance computed by HOMA2-IR, and waist circumference measured by standardized protocols. A clustered cardiometabolic risk score was created by averaging age-group standardized values of mean arterial pressure, triglycerides, LDL/HDL ratio, and HOMA-IR. Covariates: Sex and highest parental education (dichotomized into up to/including compulsory vs any post-compulsory education). Additional adjustments included waist circumference and height (for blood pressure models) and age at follow-up. Statistical analysis: Multilevel linear regression models with study as a random effect assessed associations of birth weight and MVPA with each outcome. Interaction between birth weight and MVPA was tested by including a product term (birth weight × MVPA). Sensitivity analyses examined interaction using VPA. Analyses were stratified by age group (children vs adolescents).

- Most associations between birth weight and cardiometabolic outcomes were not modified by MVPA. - Statistically significant interactions: - Children: birth weight × MVPA for waist circumference (p = 0.005), indicating MVPA may attenuate the positive association between higher birth weight and greater waist circumference. - Adolescents: birth weight × MVPA for HDL-cholesterol (p = 0.040). - Sensitivity analyses using VPA suggested additional interactions: - Children: birth weight × VPA for diastolic blood pressure (p = 0.009). - Adolescents: birth weight × VPA for LDL-cholesterol (p = 0.009) and triglycerides (p = 0.028). - Independent associations (from presented regression outputs) showed MVPA was associated with more favorable profiles (e.g., lower waist circumference and triglycerides, higher HDL, lower HOMA-IR), and higher birth weight related to lower blood pressure and some lipid/glucose measures, but the modification by MVPA was generally minimal. - Overall, MVPA was beneficial across the birth weight spectrum, with limited evidence that MVPA modifies birth weight–risk associations beyond specific outcomes noted above.

The study addressed whether MVPA changes the relationship between birth weight and cardiometabolic risk in youth. Findings indicate that MVPA generally does not modify these associations, suggesting that the cardiometabolic implications of birth weight are relatively consistent regardless of daily MVPA levels. Notable exceptions were observed for waist circumference in children and HDL-cholesterol in adolescents, where higher MVPA appeared to mitigate higher-risk profiles associated with higher birth weight (waist circumference) and was linked with HDL variation. Sensitivity analyses with VPA suggested that higher-intensity activity may play a role in modifying associations for diastolic blood pressure (children) and lipids (adolescents). Importantly, MVPA itself was consistently associated with better cardiometabolic profiles, reinforcing recommendations to promote physical activity in all youth independent of birth weight history. The results highlight the dual importance of optimal prenatal growth and regular physical activity for cardiometabolic health.

MVPA does not consistently modify the associations between birth weight and cardiometabolic risk factors in children and adolescents. However, MVPA may attenuate the positive association between higher birth weight and waist circumference in children, and higher-intensity activity (VPA) may influence associations with specific outcomes (blood pressure and lipids). MVPA is beneficial across the birth weight spectrum. These findings underscore the importance of both healthy prenatal growth and promoting physical activity throughout childhood and adolescence. Future research should explore causal pathways using longitudinal and interventional designs, examine dose–response and intensity-specific effects, and assess whether timing of activity (e.g., across developmental windows) differentially influences birth weight–risk associations.

- Birth weight was measured at birth in some studies and retrospectively reported by parents in others, introducing potential recall and measurement heterogeneity. - The pooled dataset combined prospective and cross-sectional studies; many associations are cross-sectional, limiting causal inference. - Heterogeneity in measurement protocols across studies (e.g., device models, blood sampling methods) despite harmonization may introduce residual methodological variability. - Accelerometer data were harmonized to uniaxial 60-second epochs, which may misclassify short bursts of activity common in children and adolescents. - Inclusion required ≥3 valid accelerometer days and available outcomes, which may introduce selection bias and affect generalizability.