Smoothed breastfeeding durations and waiting time to conception.

ABSTRACT

PIP: Clinical evidence supports the relationship between breast feeding and postpartum amenorrhea (PPA). The effect of heaping errors (digit preference at months 3, 6, 9, 12 of breast feeding and PPA) in studying the relationship between breast feeding and PPA in the context of time-dependent hazard models is examined, with a B-spline smoothing technique to adjust for the heaping errors in the data. The relationship among PPA, breast feeding after menses resume, and waiting time to conception are also examined with and without adjustments. A retrospective survey entitled Breastfeeding and its effect on fertility conducted in 1987 by Banaras Hindu University, Varanasi, India, provided the basic data. Data (age at marriage, children ever born, lactation, PPA) were collected from 1100 urban and 900 rural households of Varanasi. Women were classified into low, medium, and high social groups. Univariate life table technique was used to calculate total breast feeding, for which median duration was 16.25 months and 13.35 months, respectively, with and without smoothing. Similarly, the median duration of PPA was 7.82 and 5.82 months, respectively. For closed (uncensored) intervals the time of conception was estimated by subtracting 9 months from the month of delivery. The overall median duration of waiting time to conception after menstruation was more than 15 months. The heaping of breast feeding PPA showed reduced effect on the chance of conception after menstruation. Without smoothing adjustment the rate of conception after menstruation was reduced by 47% if the woman was breast feeding at that time. With adjustment the reduction increased to 63%. Younger mothers had significantly higher rates of conception once menses returned, while literate women had significantly less chance of conception than illiterate ones after menses. For every additional month of breast feeding there was a 1.8% (3.3%) reduction in the risk of conception after return of menses without (with) smoothing breast feeding and PPA. Breast feeding had a significant effect on the duration of PPA.