THE INDIRECT ASSOCIATION OF LACTATION WITH SUBSEQUENT PERIMENOPAUSAL BODY-WEIGHT

Objective: We aimed to estimate the relationship of prior breastfeedin g to perimenopausal body mass index (BMI) (kg/m(2)). While most long-t erm studies of women's body weight after reproductive experience have found a negative relationship between lactation and body weight, most short-term studies found either no impact of lactation on body weight, or decelerated postpartum weight loss among breastfeeding women. Desi gn: We performed a retrospective cohort study. Setting: The study was done in The Netherlands between 1987 and 1990. Subjects: The 1067 sing leton females who were born around the time of the Dutch famine (1 Aug ust 1944 through 15 April 1946) at The University of Amsterdam Teachin g Hospital were all traced and accounted for. This analysis refers to the 671 who were interviewed and for whom all data elements were avail able. Analysis: We related parity and number of children breastfed at least 1 week to BMI at age 45 and at the time of marriage, both calcul ated from respondents: recall of weight and height, by linear multiple regression analysis, adjusting for potentially confounding variables. Results: Among parous women, each additional live birth was associate d with a BMI increase of 0.69 kg/m(2) (CI = 0.24 to 1.13, P = 0.003), and for each additional child breastfed, the BMI was lower by 0.41 kg m(2) (CI = -0.77 to -0.04, P = 0.03) at age 45. However, BMI before an y reproductive experience was 0.39 kg/m(2) greater (CI = 0.08 to 0.71, P = 0.01) with each additional subsequent birth, and for each additio nal infant breastfed, 0.34 kg/m(2) lower (CI = -0.60 to -0.09, P = 0.0 08). Conclusions: The negative relationship of prior breastfeeding to perimenopausal weight was probably not causally related to breastfeedi ng, since the relationship was present before any reproductive experie nce. The results could also have arisen in part because of biased reca ll of weight; however, the relationship of parity to perimenopausal BM I was far less attenuated by adjusting for prior BMI than was the asso ciation with breastfeeding, lending support to an indirect association of breastfeeding with later BMI.