Article Text
Abstract
Background High birthweight and rapid weight gain (RWG) in early life are independently and positively associated with obesity in childhood and adulthood. It is unclear whether RWG can affect BMI growth from childhood to adolescence. RWG is common among low-birth-weight infants, especially following intrauterine growth retardation. However, whether the association between RWG and BMI trajectory differs by birthweight groups is not well understood. We aimed to investigate the effect of RWG (birth-3 y) on BMI trajectory (5–14 y) and whether associations differed by birthweight groups.
Methods We used data from the Millennium Cohort Study, a representative sample of children born in the UK in 2000–2002. We included term singletons with available information on weight gain (birth-3 y) and 1+ BMI measurement between 5 y and 14 y (n=11,522). Weight at birth and 3 y were converted to internal standard deviations scores (SDS). RWG was defined as change in weight SDS from birth to 3 y >0.67SDS, which is commonly used and represents the width of each percentile band on standard growth charts. Mixed effects fractional polynomial models were applied to estimate the effects of RWG on BMI trajectories, for boys and girls separately, before and after adjustment for maternal pre-pregnancy, maternal smoking during pregnancy, parity, ethnicity, breastfeeding, early introduction to solid foods, maternal education, and family income. The analysis was further stratified by low birthweight (LBW) status.
Results Boys and girls who experienced RWG were heavier than their non-RWG counterparts at 5 y by 1.21 kg/m2 [95% CI 1.10 to 1.33] and 1.25 kg/m2 [1.13–1.38] respectively; continued to gain weight more rapidly between 5 y and 14 y, especially in childhood (5–7 y) by 0.2–0.3 kg/m2 per year. These differences persisted after adjustment for potential confounders. The effect of RWG on BMI was particularly higher among non-LWB children, whose estimated BMI trajectories exceeded the International Obesity Task Force overweight reference lines. For example, at 14 y the estimated mean BMI was 24.41 kg/m2 [23.91–24.45] for non-LBW RWG girls (vs. 22.07 kg/m2 for LWB RWG girls), exceeding the gender- and age-specific cut-off point of 23.34 kg/m2 for overweight. Sensitivity analysis using 3-category birthweight variable (low/normal/high) showed that the trajectories of high-birthweight children were similar to those of normal-birthweight children.
Conclusion RWG from birth to 3 y was positively associated with BMI in early childhood and adiposity growth throughout childhood to adolescence. This positive effect was particularly detrimental for non-LBW than LBW children, which may have important public health implications for preventing early life RWG among term appropriate- and large-for-gestational-age children.