MATERNAL SMOKING DURING PREGNANCY AND BIRTH OUTCOMES WITH WEIGHT-GAINADJUSTMENTS VIA VARYING-COEFFICIENT MODELS

There is considerable interest in the impact of maternal exposures dur ing pregnancy on birth outcomes, Clearly, exposures associated with po or birth outcomes need modification or avoidance. However, arriving at such estimates of association is made challenging by a number of feat ures characteristic of the relevant data. First, exposures may be time varying (for example, cigarette and alcohol consumption) so that, to relate them to birth outcomes, one needs to model them and then extrac t derived parameters, Secondly, there are likely to be unequal numbers and spacings of exposure determinations during pregnancy. Thirdly, on e needs to account for a variety of additional covariates. Finally, th e variability and non-linearities inherent in birth outcomes mandate f lexible modelling approaches. Here we use data from a cohort of East B oston mothers to assess the impact of smoking during pregnancy on birt h weights. We emphasize modelling of, and then adjusting for, maternal weight gain during pregnancy and a proxy measure for pre-pregnancy we ight, so as to obtain better estimates of the smoking effect. Througho ut, our analysis is guided by appropriate graphics. The adjustment fea tures an interesting application of varying-coefficient models. Result s indicate that smoking related deficits in birth weights depend on th e mode of adjustment, and that previously observed deficits of approxi mately 200 g are best recaptured with use of varying-coefficient model s. (C) 1997 by John Wiley & Sons, Ltd.