Breast milk retinol and plasma retinol-binding protein concentrations provide similar estimates of vitamin A deficiency prevalence and identify similar risk groups among women in Cameroon but breast milk retinol underestimates the prevalence of deficiency among young children.

Breast milk vitamin A (BMVA) has been proposed as an indicator of population vitamin A status but has rarely been applied in large-scale surveys or compared with conventional vitamin A biomarkers. We assessed the prevalence of, and risk factors for, low BMVA and its relation to vitamin A intake, plasma retinol-binding protein (pRBP), and markers of inflammation in a national survey in Cameroon. We randomly selected 30 clusters in each of 3 strata (South, North, and Cities). Casual milk samples were collected from approximately 5 women per cluster (n = 440). pRBP, plasma C-reactive protein (pCRP), plasma α1-acid glycoprotein (pAGP), and 24-h vitamin A intake were assessed in 10 women aged 15-49 y and 10 children aged 12-59 mo per cluster, including a subset of lactating women (n = 253). Low BMVA was infrequent: 7.2% (95% CI: 4.7, 9.8) of values were <1.05 μmol/L, and 9.3% (95% CI: 5.8, 12.7) were <8 μg/g fat, consistent with the low prevalence of pRBP <0.78 μmol/L among women (< 5%) but lower than the prevalence of pRBP <0.83 μmol/L among children (35%). Risk factors for both low BMVA and pRBP included living in the North and low maternal education. BMVA was positively associated with inflammation-adjusted pRBP among women in the lowest vitamin A intake tertile [<115 μg retinol activity equivalents (RAEs)/d, P < 0.01] but not in the highest tertile (>644 μg RAEs/d, P > 0.4). Controlling for milk fat, BMVA was negatively associated with pCRP (P < 0.02) but not pAGP (P > 0.5). BMVA and pRBP provide similar estimates of vitamin A deficiency prevalence and identify the same risk groups among women in Cameroon, but BMVA underestimates the prevalence of vitamin A deficiency among young children.