Breastfeeding status and duration are simulated in the model after delivery, and are used to determine if women experience protective effects of lactational amenorrhea on risk of pregnancy.
We obtained data compiled by UNICEF on ‘any breastfeeding’ for 131 countries, and ‘exclusive breastfeeding’ of infants 0-5 months for 134 countries. Globally, it is estimated that 42% of infants 0-5 months of age were exclusively breastfed in 2018, with variation by region.
|Region||Exclusive Breastfeeding (%)|
|Eastern and Southern Africa||55|
|Latin America and the Caribbean||38|
|West and Central Africa||34|
|Eastern Europe and Central Asia||33|
|East Asia and the Pacific||30|
|Middle East and North Africa||30|
We estimated hierarchical logistic regression models for the probability of any breastfeeding, and of exclusive breastfeeding for infants 0-5 months based on the compiled UNICEF data. We also set priors for the probability of breastfeeding for 6 months (as opposed to 9 months, the maximum duration of LAM efficacy we model) conditional on any breastfeeding. We allowed the sampled parameters to vary by subgroup in the model.
Breastfeeding status is simulated after live birth delivery in the model. Conditional on any breastfeeding, a status (exclusive/non-exclusive) and duration are assigned. We assume that women who breastfeed exclusively for any duration will then continue to breastfeed non-exclusively for an additional 3 months. If a breastfeeding child dies, we assume that the mother stops breastfeeding and no longer experiences any protective efficacy from LAM. (Note: If a mother had twins, then LAM protective efficacy continues if at least 1 breastfeeding child is still alive.)