Importance: Greater caffeine consumption in pregnancy is associated with reduced birth size, but potential associations with childhood growth are unclear.
Objective: To evaluate the associations of pregnancy caffeine and paraxanthine measures with child growth in a contemporary cohort with low caffeine consumption and a historical cohort with high caffeine consumption.
Design, setting, and participants: The Environmental Influences on Child Health Outcomes cohort of the National Institute of Child Health and Human Development Fetal Growth Studies (ECHO-FGS; 10 sites, 2009-2013) was a pregnancy cohort with 1 child measurement between ages 4 and 8 years (follow-up in 2017-2019). The Collaborative Perinatal Project (CPP) was a pregnancy cohort (12 sites, 1959-1965) with child follow-up through 8 years (1960-1974). The current secondary analysis was conducted in 2021 and 2022.
Exposures: Concentrations of caffeine and its primary metabolite, paraxanthine, were quantified from plasma (ECHO-FGS) and serum (CPP) collected in the first trimester. Cut points for analyses were defined by quartiles in ECHO-FGS and quintiles in CPP.
Main outcomes and measures: Child z scores for body mass index, weight, and height were evaluated, as well as fat mass index and percentage and obesity risk measured at 1 time between age 4 and 8 years in ECHO-FGS. In a secondary analysis of the CPP cohort, child z scores and obesity risk longitudinally through age 8 years were evaluated.
Results: In ECHO-FGS (median caffeine intake <50 mg/d), 788 children (mean [SD] age, 6.8 [1.0] years; 411 boys [52.2%]) of women in the fourth vs first quartile of plasma caffeine concentrations had lower height z scores (β = -0.21; 95% CI, -0.41 to -0.02), but differences in weight z scores were only observed in the third quartile (β = -0.27; 95% CI, -0.47 to -0.07). In CPP, beginning at age 4 years, 1622 children (805 boys [49.7%]) of women in the highest caffeine quintile group had lower height z scores than their peers from the lowest group, with the gap widening with each successive year of age (β = -0.16 [95% CI, -0.31 to -0.01] at 4 years; β = -0.37 [95% CI, -0.57 to -0.16] at 8 years). There were slight reductions in weight at ages 5 to 8 years for children in the third vs first caffeine quintile (β = -0.16 to -0.22). Results were consistent for paraxanthine concentrations in both cohorts.
Conclusions and relevance: Intrauterine exposure to increasing levels of caffeine and paraxanthine, even in low amounts, was associated with shorter stature in early childhood. The clinical implication of reductions in height and weight is unclear; however, the reductions were apparent even with levels of caffeine consumption below clinically recommended guidelines of less than 200 mg per day.