|dc.description.abstract||Aim: Studies that investigate the main predictors of breastfeeding duration are important in realising the WHO recommendation of breastfeeding for two years and beyond. The main aim of the study was to model the impact of socioeconomic and demographic factors on the rate and duration of breastfeeding, while controlling for health sector variables. A secondary analysis of the 2007 Swaziland Demographic and Health Survey (SDHS) was conducted to achieve this aim.
Sample: A total sample of 2,812 women aged 15-49 reported the birth of at least one child born within five years of the survey. However only 2,601 women out of the 2,812 had information on the duration of breastfeeding of their last born children, and these were used as a sample for analysis in this study. Thus 211 from the 2,812 were considered missing.
Method: To establish the effects of the socioeconomic and demographic variables on the breastfeeding rate (ever breastfed) while controlling for health sector variables, a multivariate logistic regression was used. Furthermore, to model the impact of socioeconomic and demographic variables on the duration of breastfeeding, three logistic regression models were fitted to the data for children breastfed for at least six, 12 and 24 months of age.
Results: The rate of ever breastfeeding was found to be 94.3 % in Swaziland. In the multivariate logistic regression, highly educated women, women residing in urban areas, women residing in the Manzini region, delivery through cesarean section, and non-use of modern contraceptives were associated with lower odds of ever breastfeeding.
For the duration of breastfeeding three binary logistic models were fitted at six, 12 and 24 months. Of the mothers with a child aged 6 to 11 months at the time of the survey, 17.5% had breastfed for at least six months. In the multivariate analysis, highly educated mothers, those older than 20 years, mothers residing in urban areas, and mothers who delivered at a health facility were more likely to have breastfed their last born child for at least six months.
Among mothers with a last born child aged 12 to 23 months at the time of the survey, 39.8% reported to breastfeeding for at least 12 months. In the multivariate analysis less educated and uneducated mothers, mothers aged 35 and above, mothers residing in rural areas, mothers with a parity of one child, mothers using modern contraceptives, and mothers who delivered in a health facility had higher odds of breastfeeding for at least 12 months.
At the time of the survey only 13.1% of the mothers with a last born child aged 24 and above reported breastfeeding for at least 24 months. In the multivariate analysis mothers from the poorest households had the lowest odds of breastfeeding for at least 24 months compared to the poorer and richest households. Mothers older than 20 years, those who were married, mothers who delivered in a health facility, and were using modern contraceptives were more likely to report have breastfed their last born child at 24 months during the time of the survey.
Conclusion: The study found that the rate of ever breastfeeding is common in Swaziland. However, more attention need to be focused on the duration of breastfeeding as it is below the WHO recommendation with only 13.1% of children breastfed up to two years and beyond. The study found that SES, demographic and health sector variables were significant predictors of breastfeeding.||en_US