Published online by Cambridge University Press: 31 July 2008
The effects of initiation of solid and liquid supplementation on resumption of post-partum menstruation are examined, 0using data from a 2-year prospective study of birth interval dynamics from central Java, Indonesia. The sample analysed consisted of 444 women who experienced resumption of menses while breast-feeding, women who were breast-feeding and amenorrhoeic at the end of the study, or women who resumed menstruation or were censored after infant mortality and weaning which preceded the resumption of menses. Multivariate hazard model analysis was used to assess the significance of supplementation, various breast-feeding covariates, and age and parity of the dependent variable. Because the timing of supplementation varies, the supplementation variables were introduced into the analysis as time-varying covariates. For the mothers in the sample, solid and liquid supplementation was initiated at medians of 2·1 and 8·0 months respectively. The former had a significant effect on resumption of menses, while the latter was only marginally significant. Earlier supplementation meant shorter durations of amenorrhoea for the majority of women. However, the effect was not consistent across all categories of women. For the small group of mothers who were low intensity breast-feeders (≤6 minutes per nursing episode) or as low frequency day-time breast-feeders (≤6 nursing episodes per day-time), earlier supplementation had no additional effect on their rate of resumption of menses post-partum.