The importance of maternal BMI on infant's birth weight in four BMI groups for the period 1978-2001.

OBJECTIVE: To study whether increased maternal weight and other factors of importance is associated with higher birth weights of the children over a period of almost 25 years.

DESIGN: Retrospective cohort study.

SETTING: Delivery wards in southeast Sweden.

SAMPLE: A total of 4,330 delivered women and their children from the years 1978, 1986, 1992, 1997, and 2001.

METHODS: Analysis of covariance was used to evaluate the importance of the mother's body mass index (BMI) on the children's birth weights during the study years and smoking, parity, employment, gestational age, and the age of the mothers were adjusted for.

MAIN OUTCOME MEASURES: Weight of the offspring in relation to maternal BMI and possible confounders such as smoking, parity, employment, gestational age, and the age of the mother.

RESULTS: Between 1978 and 1992, there was an increase in birth weight in each of the four BMI categories (i.e. BMI<20, 20-24.9, 25-29.9 and > or =30, respectively) even after adjustments were made for relevant background characteristics (p<0.001). However, between 1992 and 2001, the birth weight for children whose mothers had a BMI of less than 20 or between 20 and 24.9 decreased (p<0.001). For almost every study year, the mothers' BMI was of significant influence on the children's birth weights. However, the proportion of variance explained by the models (i.e. the adjusted R(2)) was not substantially altered when the mother's BMI was excluded from the models.

CONCLUSION: Maternal BMI is of significance to explain trends in infants' birth weight over time, but not of sole importance.