Maternal dietary patterns are associated with human milk composition in Chinese lactating women.

OBJECTIVES: Dietary patterns are a useful tool to study the impact of overall maternal diet on human milk (HM) composition beyond single foods or nutrients. The present study aimed to identify dietary patterns among Chinese lactating women and assess their associations with HM macronutrient composition. METHODS: Dietary intake data and HM samples were collected from 122 Chinese mothers at three to five study visits during the first 52 d postpartum. Dietary patterns were derived using principal component analysis. Cross-sectional associations of dietary patterns and HM macronutrients were assessed using multivariable linear regression models adjusted for total energy intake. All analyses were done separately for colostrum (postpartum days 0-7) and mature milk (postpartum days 8-52). RESULTS: Four dietary patterns were identified: high-in-animal-foods, high-in-eggs, high-in-plant-foods, and high-in-fruits. Compared with the lowest tertile (T1), participants in the highest tertile (T3) of the high-in-animal-foods and high-in-plant-foods patterns had lower protein (respectively, T3 - T1 = -1.09 g/100 mL, Ptrend = 0.002; T3 - T1 = -0.54 g/100 mL, Ptrend = 0.001) and higher fat (respectively, T3 - T1 = 0.86 g/100 mL, Ptrend = 0.040; T3 - T1 = 0.40 g/100 mL, Ptrend = 0.004) concentrations in colostrum. In contrast, in mature milk the high-in-animal-foods pattern was positively associated with carbohydrates (T3 - T1 = 0.53 g/100 mL, Ptrend = 0.008) and the high-in-plant-foods pattern was negatively associated with fat (T3 - T1 = -0.64 g/100 mL, Ptrend = 0.002). The high-in-eggs pattern was weakly positively associated with protein concentration in mature milk (T3 - T1 = 0.10 g/100 mL, Ptrend = 0.023). CONCLUSIONS: Maternal dietary patterns with high proportions of animal and plant-based foods were associated with higher fat and lower protein concentrations in colostrum. Different associations were found in mature milk. Dietary-pattern analysis provides an opportunity to characterize total diet and may be more predictive of HM composition than single foods or nutrients.