Purpose: This study aimed to investigate the association between maternal fish consumption during pregnancy and BMI in children and the development of this association between birth and 14 years of age, taking into account relevant mother and child covariates.
Methods: The study population consisted of 3684 Dutch children born in 1996-1997 who participated in the PIAMA birth cohort study. Maternal fish consumption during pregnancy and the child's body weight and height (up to 11 times) were reported by questionnaire. Generalized estimating equations were used to investigate whether BMI of children differed according to maternal fish consumption during pregnancy.
Results: The crude overall association between maternal fish consumption during pregnancy and BMI in children was non-significant (P = 0.17), but differed by the child's age (P interaction = 0.03). Children of mothers who consumed fish ≥1×/week during pregnancy (n = 909) had statistically significant lower mean BMI z scores than children of mothers who never consumed fish (n = 1025) at the ages 4, 7, 8.5, and 11.5 years. Adjustment for maternal covariates (particularly pre-pregnancy BMI) attenuated the differences, which remained statistically significant at the age of 7 years only (mean difference in BMI z score: -0.14 95 % CI -0.25; -0.03). Additional adjustment for child covariates hardly affected the results.
Conclusions: In a population with relatively low fish consumption, higher fish consumption by pregnant women seems rather an indicator for more healthy maternal characteristics in general than a causal factor for the lower BMI in their children.
Keywords: BMI; Children; Fish; Maternal consumption.