Abstract
PURPOSE: The strength of the association between obesity and mortality rate (MR) varies by body mass index (BMI) and sociodemographic groups. We test the hypothesis that the association between obesity and MR varies, in part, due to the moderating effect of parental BMI and birth weight. METHODS: Data come from the 1958 National Child Development Study, an ongoing longitudinal dataset initiated in 1958 with baseline measures of birth weight from 18,059 infants born in Great Britain over 1 week. We tested whether the association between BMI and MR was moderated by parental BMI and birth weight using generalized additive proportional hazards models. RESULTS: The association between adult BMI and MR was moderated by birth weight and maternal BMI, such that the association between BMI and MR was weaker among individuals with a higher birth weight (P = .0148) and stronger among individuals born to mothers with a higher BMI (P = .032). At any given level of BMI approximately greater than 25, individuals with low birth weight or born to mothers with a higher BMI, had a higher MR. Paternal BMI did not significantly modify the relationship between BMI and MR (P = .5168). CONCLUSIONS: Results suggest that the relationship between obesity and MR is modified by birth weight and maternal BMI.