N-3 fatty acids

In a study with mother-infant pairs (n = 39), human milk was obtained at one, three, and six months postpartum. ^{Reference Zielinska, Hamulka, Grabowicz-Chądrzyńska, Brys and Wesolowska16} ALA, DHA, and total n-3 long chain PUFA content at one and three months were positively correlated with infant motor development at 6.6 months of age using the Child Development Scale. The correlation was still statistically significant after adjusting for infant age, sex, birthweight, maternal age, education, psychological status, and parity. ^{Reference Zielinska, Hamulka, Grabowicz-Chądrzyńska, Brys and Wesolowska16} In another study, the Brazelton Neonatal Behavioral Assessment Scale (NBAS) was administered ^{Reference Hart, Boylan and Carroll17} at nine days of age to assess motor development. A positive correlation (r = 0.57) was found between relative DHA content in human milk and range of cluster score on the NBAS. ^{Reference Hart, Boylan and Carroll17} This finding suggests that infants receiving higher levels of DHA are less likely to frequently change state of arousal.

In a controlled intervention study, women were randomized to 2.2 g of DHA (n = 98) from week 20 gestation until birth or 4.0 g of olive oil (n = 46) daily ^{Reference Dunstan, Mitoulas and Dixon18} and human milk (n = 78) was collected at three days and six months postpartum. A positive correlation was found between relative DHA content in milk at three days postpartum for six out of seven subscales of Griffiths Mental Development Scales at 2.5 years of age: locomotor (r = 0.27), speech and hearing (r = 0.288), eye and hand coordination (r = 0.41), performance (r = 0.32), practical reasoning, and general quotient. ^{Reference Dunstan, Mitoulas and Dixon18} There were no correlations between health outcomes and DHA concentration in milk in the 6 month postpartum samples. ^{Reference Dunstan, Mitoulas and Dixon18}

Mother-infant pairs (n = 709) participated in a study aimed to identify pre- and postnatal determinants on child development and health outcomes. At two years of age, motor development assessed using the French Psychomotor Developmental Scale for Early Childhood of Brunet-Lézine and MacArthur Communicative Development Inventory (CDI) was not related to milk DHA content. At three years of age, there was no association observed between n-3 fatty acid content in milk and cognitive function assessed using the second French edition of the Ages and Stages Questionnaire (ASQ-3). ^{Reference Bernard, Armand and Garcia19} Furthermore, milk DHA content at four months was not related to cognitive development assessed by the Mullen Scales of Early Learning at 12 months, ^{Reference Keim, Daniels, Siega-Riz, Herring, Dole and Scheidt20} and n-3 fatty acid content of colostrum and milk at one, two, and four months of lactation was not related to visual evoked potential at 2.5 or 7.5 months, or outcomes from the Bayley Scales of Infant Development (BSID)-II at 12 months. ^{Reference Hurtado, Iznaola and Pena21}

N-6 fatty acids

The relation between neurodevelopmental outcomes and n-6 fatty acid content was investigated in a study of human milk (n = 78) obtained at three days, six weeks, and six months postpartum. Infants were tested by a psychologist at 2.5 years for receptive language skills using the Peabody Picture Vocabulary Test IIIA. ^{Reference Dunstan, Mitoulas and Dixon18} Increased relative AA content in milk six months postpartum was associated with reduction in vocabulary skills including the average length of phrases used (r = −0.533) and number of words used (r = −0.371). ^{Reference Dunstan, Mitoulas and Dixon18} Relative LA content in milk one-week postpartum was negatively correlated to Motor-2 at two years assessed using the CDI and the ASQ-3 at three years of age. ^{Reference Bernard, Armand and Garcia19} Furthermore, milk AA content at four months was not related to cognitive development assessed by the Mullen Scales of Early Learning at 12 months, ^{Reference Keim, Daniels, Siega-Riz, Herring, Dole and Scheidt20} and n-6 fatty acid content of colostrum and milk at one, two, and four months of lactation was not related to visual evoked potential at 2.5 or 7.5 months, or outcomes from the BSID-II at 12 months. ^{Reference Hurtado, Iznaola and Pena21}

N-3/n-6 fatty acid ratio

Fatty acids were measured in colostrum (n = 319) obtained 48−96 hours postpartum and correlated with the BSID-I at 14 months of age. The ALA/LA, EPA/AA, and DHA/AA ratios were positively correlated to mental development using the BSID-I at 14 months of age. ^{Reference Guxens, Mendez and Molto-Puigmarti22} In a cohort study of infant-mother pairs (n = 709), cognitive function was assessed using the second French edition of the ASQ-3 relative to milk fatty acid content obtained one-week postpartum. Consistent with studies relating n-3 and n-6 fatty acid individually to neurodevelopment outcomes, a positive association between relative total n-3/n-6 ratio and ASQ-3 score at three years was observed. ^{Reference Bernard, Armand and Garcia19} Mothers with rs2397142 CC genotype in ELOVL5 conferring higher DHA/AA ratio showed improved child cognitive index assessed by the McCarthy Scales of Children’s Abilities at 14 months compared to G allele carriers. ^{Reference Morales, Bustamante and Gonzalez23}

N-3 fatty acids

Milk samples were obtained from mothers (n = 208) at six weeks and four months postpartum in the Impact of Nutritional Fatty acids during pregnancy and lactation for early human Adipose Tissue development (INFAT) study. ^{Reference Much, Brunner and Volhardt24} There was a positive correlation between relative EPA content (r = 0.20), DHA content, and total n-3 long chain PUFA in the six week milk with skinfold thickness (SFT) at 12 months (r = 0.16). Total n-3 long chain PUFA was also positively correlated to the ratio of subcutaneous to preperitoneal fat (r = 0.12) at six weeks of age, and with ponderal index (PI) at 12 months of age. ^{Reference Much, Brunner and Volhardt24} At one year postpartum, PI and body mass index (BMI) were related to DHA content (r = 0.15), but only PI was related to EPA content. ^{Reference Much, Brunner and Volhardt24} A follow-up to this study ^{Reference Much, Brunner and Volhardt24} was conducted in children (n = 169) who participated in the original study. Relative total n-3 from milk six-week postpartum was positively correlated to weight, BMI, SFT, and body fat at two years of age. Only the association between total n-3 fatty acids and lean body mass (LBM) remained statistically significant until five years of age. In the milk four months postpartum, total n-3 long chain PUFA concentration was negatively correlated to preperitoneal fat mass at two but not at three, four, or five years of age. ^{Reference Meyer, Brei, Stecher, Much, Brunner and Hauner25}

N-6 fatty acids

Body composition outcomes were related to n-6 fatty acids in the INFAT study. ^{Reference Much, Brunner and Volhardt24} There were negative correlations between total relative n-6 long chain PUFA concentration in milk six weeks postpartum and BMI, LBM, PI, and weight at four months of age (r = −0.41,−0.36,−0.31,−0.36 respectively). N-6 long chain PUFA content was negatively correlated to fat mass at six weeks of age (r = −0.29), sum of SFT at 12 months of age (r = −0.19), and body fat percentage at six weeks of age (r = −0.18) but not at 12 months of age. ^{Reference Much, Brunner and Volhardt24} A negative correlation was observed between relative AA content in milk at six weeks postpartum and LBM, BMI, PI, and weight at four months of age (r = −0.33, −0.35, −0.26, −0.32 respectively). AA content of milk was negatively correlated to fat mass at six weeks of age (r = −0.23), but not at 12 months. ^{Reference Much, Brunner and Volhardt24} In another study, mothers (n = 614) provided a two week milk expression between four and eight months postpartum until 100 mL was collected. Absolute linoleic acid content was negatively related to SFT at 12 months of age, suggesting that n-6 fatty acid content of milk is negatively related to indices for body fat. ^{Reference Prentice, Ong and Schoemaker26}

N-3/n-6 fatty acid ratio

Body weight, BMI, SFT, and LBM were measured in children (n = 169) at two, three, four, and five years of age. ^{Reference Meyer, Brei, Stecher, Much, Brunner and Hauner25} The n-3/n-6 long chain PUFA ratio from milk six weeks postpartum was positively associated with weight and BMI at two years of age, and negatively with LBM at four and five years of age. ^{Reference Meyer, Brei, Stecher, Much, Brunner and Hauner25} In another study, the fatty acid content of milk at two weeks postpartum from mothers with BMI ≥ 25 (n = 56) was assessed as a predictor of child body composition and anthropometry at two weeks, two months and six months of age. In children receiving formula and human milk, the n-3/n-6 ratio was positively correlated with weight-for-age z-score, weight-for-length z-score, and BMI z-score change from two weeks to six months. ^{Reference Ellsworth, Perng, Harman, Das, Pennathur and Gregg27} The clinical relevance of this finding is obscured because the n-3/n-6 ratio was not related to anthropometry in babies fed human milk exclusively. ^{Reference Ellsworth, Perng, Harman, Das, Pennathur and Gregg27}

N-3 fatty acids

A study of human milk (n = 315) at one month postpartum assessed eczema, atopic dermatitis, and allergic sensitization at ages seven, 12, and 24 months. Higher total n-3 long chain PUFA concentrations were associated with lower risk for atopic dermatitis (OR = 0.24, 0.07–0.91) at two years of age and risk of allergic sensitization at one year of age (OR = 0.17, 0.04–0.77). ^{Reference Thijs, Muller and Rist28} Another study examined the relation between incidence of atopic disease and relative n-3 concentration in human milk (n = 263) collected at six weeks and six months postpartum. ^{Reference Oddy, Pal and Kusel29} Infants whom developed non-atopic eczema at six months of age had lower total n-3 PUFA in the milk six weeks postpartum, but not in the milk six months postpartum. ^{Reference Oddy, Pal and Kusel29}

Women at risk of having a child with allergy were randomized to receive 1.6 g of EPA and 1.1 g DHA or soya bean placebo from 25 weeks of gestation to three months postpartum (n = 145). ^{Reference Warstedt, Furuhjelm, Falth-Magnusson, Fageras and Duchen30} Milk samples were collected at two-four days, one month, and three months postpartum. At two years of age, IgE-associated disease was lower in children consuming higher levels of EPA and DHA in milk from two-four days and one month postpartum. ^{Reference Warstedt, Furuhjelm, Falth-Magnusson, Fageras and Duchen30} In another study of mother-infant pairs (n = 34), the relation between atopic disease and fatty acid composition of milk obtained at one month postpartum was investigated. ^{Reference Hoppu, Rinne, Lampi and Isolauri31} Lower content of EPA and total n-3 PUFA was observed in milk from mothers with infants who developed atopic dermatitis within the first year of life. ^{Reference Hoppu, Rinne, Lampi and Isolauri31} In another cohort (n = 245), the risk for developing allergic sensitization up to 24 months of age was lower in children consuming milk with higher total n-3 content. ^{Reference Rosenlund, Fagerstedt, Alm and Mie32}

N-6 fatty acids

Human milk (n = 265) was obtained from mothers with or without asthma, hay fever, allergy for pets, or allergy for house dust/dust mites at three months postpartum. ^{Reference Wijga, Van Houwelingen and Kerkhof33} The International Study on Asthma and Allergy in Children questionnaire was completed by parents at one and four years of child age. Indicators of allergic disease were determined by eczema at one year, and asthma and sensitization at four years. Lower AA content in milk was a risk factor for all allergic symptoms at one and four years of age only in children of mothers with allergy. ^{Reference Wijga, Van Houwelingen and Kerkhof33} In addition, the total n-6 relative concentrations in milk six weeks postpartum were higher in a study of infants (n = 263) whom developed non-atopic eczema at six months of age. ^{Reference Oddy, Pal and Kusel29}

N-3/n-6 fatty acid ratio

Infants with at least one parent with a history of allergy defined by hay fever, eczema, and/or asthma had skin conditions reported at six weeks, six months, 12 months, and annually until five years of age. ^{Reference Oddy, Pal and Kusel29} Decreased n-3/n-6 ratio in milk six weeks postpartum was associated with increased risk of non-atopic eczema at six weeks and six months but not at one or five years of age. ^{Reference Oddy, Pal and Kusel29} High n-3/n-6 ratio was associated with lower incidence of eczema at one year, asthma at four years, and persistence of symptoms assessed by a questionnaire based on the International Study on Asthma and Allergy in Children. ^{Reference Wijga, Van Houwelingen and Kerkhof33} Blood samples were collected from infants (n = 245) at six, 12, and 24 months of age and infants were assessed for sensitization to any of hen’s egg, cow’s milk, peanut, cat, dog, birch and timothy, determined as IgE ≥ 0.35 kUA/l. Increased risk of sensitization up to 24 months was associated with decreased EPA/AA ratio from milk two months postpartum. ^{Reference Rosenlund, Fagerstedt, Alm and Mie32} Lower incidence of IgE-associated disease was observed in infants at six, 12, and 24 months of age with higher EPA/AA in colostrum, but not with fatty acids in milk from one and three months postpartum. ^{Reference Warstedt, Furuhjelm, Falth-Magnusson, Fageras and Duchen30}

Another study indicated no association between n-3/n-6 ratio in milk with atopic dermatitis. ^{Reference Hua, Su and Kuo34} Colostrum and milk two months postpartum were obtained from mothers (n = 146) with a history of allergy. Infants were assessed at two, four, six, eight, 12, and 24 months of age for atopic dermatitis. There were no statistically significant findings observed for fatty acid concentration in milk and infant atopic dermatitis when compared to healthy control groups with and without controlling for potential confounders including maternal age and allergy, cesarean section, parity, pets in the household, tobacco smoking within the last year, and use of antibiotics/probiotics during sample collection. ^{Reference Hua, Su and Kuo34}

Quality assessment

Risk of bias was present in studies where analyses were conducted post hoc or with convenience sampling. Furthermore, bias was present when a 24 hr expression of milk was not collected or reported. Inconsistency and imprecision were considered negligible as the fatty acid predictors were directionally and uniformly associated with the health outcomes in the three domains of interest in 20 of 21 studies assessed. Indirectness was present for each health outcome assessed. For example, neurodevelopmental outcome domains included visual and cognitive indices across a range of data acquisition tools. Body composition was assessed by various metrics including weight, PI, or BMI. Finally, allergic outcomes included serum IgE antibody titers, and skin problems, atopy, eczema, and asthma which were often self-reported.

The average certainty of evidence scores for neurodevelopment, body composition, and allergy, skin & eczema were 8.0, 8.5, and 7.6 respectively. There was evidence of moderate certainty from the systematic review of observational studies demonstrating an association between n-3 and n-6 fatty acids and neurodevelopmental outcomes. Assessment of the observational studies on body composition outcomes yielded evidence of moderate certainty from the systematic review for the association to n-3 and n-6 fatty acids. Imprecision and indirectness were evident mostly in studies of allergic outcomes. There was evidence of moderate certainty from the systematic review of observational studies demonstrating an association between n-3 and n-6 and allergy outcomes.