Determinants of maternal and infant omega-3 status at 3 months postpartum: findings from the APrON longitudinal cohort study

Abstract

Background: Omega-3 long-chain-polyunsaturated fatty acids (LCPUFAs) are important dietary components for maternal and infant health during pregnancy and lactation.

Objectives: This study investigated determinants of maternal and infant LCPUFAs status at 3 mo postpartum and the relationship between maternal serum, maternal milk, and infant LCPUFAs.

Methods: This cross-sectional study included mothers (n = 1481) and their offspring (n = 526) at 3 mo postpartum from the Alberta Pregnancy Outcomes and Nutrition (APrON) cohort. Maternal dietary intake (24-h recall), blood samples from mothers and infants, and maternal milk were collected. Fatty acid composition (relative % of total fatty acids) was determined by gas-liquid chromatography. Linear regression analyses explored associations between diet, sociodemographic factors, and fatty acid status.

Results: In a multivariable-adjusted analysis, maternal total dietary intake (supplement + food) was positively associated with the percentage of DHA (standardized ß [Sβ] = 0.158; ß = 0.394; 95% [confidence interval] CI: 0.192, 0.558; P < 0.001) in maternal serum phospholipids. Similar associations were found for DHA and eicosapentaenoic acid in maternal milk and plasma phospholipids of infants. Prepregnancy body mass index (BMI) was negatively associated with DHA (Sß = -0.073; ß = -0.003; 95% CI: -0.006, -0.001; P = 0.008) and positively associated with total saturated fatty acids (Sß = 0.086; ß = 0.111; 95% CI: 0.042, 0.180; P = 0.002) in maternal milk. Infants receiving formula combination with maternal milk had lower percentage of DHA (Sß = -0.177; ß = -0.390; 95% CI: -0.604, -0.175; P < 0.001) and arachidonic acid (Sß = -0.106; ß = -0.595; 95% CI: -1.122, -0.067; P = 0.027) in their plasma phospholipids compared with those who fed exclusively maternal milk.

Conclusions: Maternal total dietary intake and prepregnancy BMI are independently associated with their serum fatty acid status during lactation, whereas maternal diet, milk fatty acid composition, and lactation status are important determinants of infant n-3 LCPUFAs fatty acid status. Future research should investigate the impact of these differences in fatty acid status on infant health outcomes.

Keywords: dietary intake; docosahexaenoic acid; fatty acids; human milk; lactation.

FIGURE 1

Description of the number of participants included in each analysis. ∗Not all participants have calculated data for all nutrients assessed. The specific number of participants is identified in each table of the results section.

FIGURE 2

Determinants of relative percentage of fatty acid composition in maternal serum phospholipids, human milk total lipids, and infant plasma phospholipids. The sample size for each analysis is as follows: n = 1481 for human milk, n = 1066 for maternal serum phospholipids, and n = 526 for infant plasma phospholipids. Standardized β estimates are reported for univariate and multivariate analyses. Standardized ß should be interpretated as the changes in standard deviation units of the dependent variable by 1 standard deviation increase or decrease of the independent variable, holding all the other variables constant in the model. The following continuous independent variables were log transformed: dietary and total intake of DHA, EPA, DHA+EPA, and ARA. It is noted that following log transformation of the independent variable, the ß coefficients may differ in interpretability compared to those of untransformed variables. Dietary intake does not include the quantity of supplement intake and those who were taking supplements. Supplement intake was converted to a dichotomous variable (no or yes). Total dietary intake refers to the dietary intake + supplement intake of DHA and EPA. For continuous predictor variables, a negative sign in the regression coefficient suggests an inverse association, meaning as the predictor increases, the outcome decreases. For categorical predictor variables, a negative sign indicates that the outcome is lower in the second category compared to the first. Univariable linear regression was performed for each covariate to determine its individual contribution. Variables with P < 0.1 were included in the multivariable model. Supplemental Tables 3–8 describes the list of covariates used for each outcome and the sample size, as some variables had missing data. There are known associations between sociodemographic variables and fatty acid composition. Maternal age and prepregnancy BMI were retained in the multivariable analysis for all fatty acids. Infant sex was also included in the multivariable model for infants’ plasma phospholipids. Cells with a white background indicate a β lower than 0.0004 in the univariable model or that the variable was not included in the multivariable model due to lack of associations. ∗ Indicates statistical significance. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Fatty acids: Total n–3 = sum of 20:4n–3, 20:5n–3, 22:5n–3, 22:6n–3; 2 Total n–6 = sum of 18:2n–6, 18:3n–6, 20:2n–6, 20:3n–6, 20:4n–6, 22:4n–6, 22:5n–6; PUFAs = sum of 18:2n–6, 18:3n–6, 20:2n–6, 20:3n–6, 20:4n–6, 22:4n–6, 22:5n–6, 18:3n–3, 20:4n–3, 20:5n–3, 22:5n–3, 22:6n–3; MUFAs = sum of 16:1n–7, 18:1 c11, 18:1n–9, 18:1n–7, 24:1n–9; SFAs = sum of 14:0, 16:0, 18:0, 20:0, 24:0. Abbreviations: ARA, arachidonic acid; BMI, body mass index; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; SFA, saturated fatty acid.