Maternal and paternal relationships with offspring bone mass: response to Harvey et al.

Abstract

Dear Editor, We thank Harvey et al. for their comment [1] regarding our paper comparing relationships of maternal and paternal smoking during pregnancy with offspring bone measures at age 10 years in the Avon Longitudinal Study of Parents and Children (ALSPAC) [2]. We agree that the associations the authors report of maternal smoking and pre-pregnancy BMI with offspring whole body bone mineral content (BMC) at birth in the Southampton Women’s Survey, without similar paternal effects [1], provide some support for intrauterine influences on fetal bone development. However, we emphasise that in our study we aimed to assess whether any intrauterine effects on bone mineral accrual persist into childhood. The authors refer to differences between maternal and paternal associations in our complete case analysis as potentially representing an intrauterine effect. We point out that in our complete case, as in our multiple imputation analysis, there were similar-sized associations of maternal and paternal smoking with offspring total body BMC, bone area and bone mineral density in confounder-adjusted models; parental differences were apparent only in models for spinal bone outcomes. Furthermore, in both analyses, any associations of maternal smoking with offspring bone measures were positive, and thus in the opposite direction to those found with BMC at birth [1, 3, 4], and completely explained by the child’s height and, in particular, weight. Our findings, therefore, are not consistent with an adverse intrauterine effect of maternal smoking on bone development which persists into later childhood. In our other study referenced by the authors, the positive relationship of maternal BMI with offspring bone mass at age 10 years [5] is consistent with that reported at birth [1]. In our complete case, but not in our multiple imputation analysis, maternal effects were slightly stronger than paternal effects, so an intrauterine influence is plausible. However, detailed analyses suggested that these associations were completely driven by the association of maternal BMI with child’s BMI, rather than direct effects on bone development. Previous studies investigating maternal and paternal–offspring BMI associations have had inconsistent findings [6–9], although similar associations of maternal and paternal BMI with offspring BMI demonstrated in ALSPAC [6] and in the largest study to date [9] give support to our findings from multiple imputation of similarsized parental BMI associations with offspring bone mass. Finally, the authors suggest that maternal smoking status and BMI during/pre-pregnancy will be highly correlated with those post-pregnancy. We would expect that this is also true for fathers, suggesting that effects of parental exposures would be similar unless there was a specific effect of the maternal exposure during the intrauterine period [10]. Furthermore, we do not agree that an association of birth weight with later bone density is proof of an intrauterine effect. Smoking is the strongest environmental determinant of birth weight and strongly related to socioeconomic position (and largely explains the association of socioeconomic position with birth weight [11]); thus, familial socioeconomic position and associated lifestyles would C. Macdonald-Wallis (*) :G. Davey Smith :D. A. Lawlor MRC Centre for Causal Analyses in Translational Epidemiology, School of Social and Community Medicine, University of Bristol, Oakfield House, Oakfield Grove, Bristol BS8 2BN, UK e-mail: C.Macdonald-Wallis@bristol.ac.uk