OBJECTIVE: Docosahexaenoic acid (DHA) and arachidonic acid (ARA) are long-chain polyunsaturated fatty acids found in breast milk and recently added to infant formulas. Their importance in infant nutrition was recognized by the rapid accretion of these fatty acids in the brain during the first postnatal year, reports of enhanced intellectual development in breastfed children, and recognition of the physiologic importance of DHA in visual and neural systems from studies in animal models. These considerations led to clinical trials to evaluate whether infant formulas that are supplemented with DHA or both DHA and ARA would enhance visual and cognitive development or whether conversion of linoleic acid and alpha-linolenic acid, the essential fatty acid precursors of ARA and DHA, respectively, at the levels found in infant formulas is sufficient to support adequately visual and cognitive development. Visual and cognitive development were not different with supplementation in some studies, whereas other studies reported benefits of adding DHA or both DHA and ARA to formula. One of the first trials with term infants that were fed formula supplemented with DHA or both DHA and ARA evaluated growth, visual acuity (Visual Evoked Potential; Acuity Card Procedure), mental and motor development (Bayley Scales of Infant Development), and early language development (MacArthur Communicative Developmental Inventories). Growth, visual acuity, and mental and motor development were not different among the 3 formula groups or between the breastfed and formula-fed infants in the first year of life. At 14 months of age, infants who were fed the formula with DHA but no ARA had lower vocabulary production and comprehension scores than infants who were fed the unsupplemented control formula or who were breastfed, respectively. The present follow-up study evaluated IQ, receptive and expressive vocabulary, visual-motor function, and visual acuity of children from the original trial when they reached 39 months of age.

METHODS: Infants were randomized within 1 week after birth and fed a control formula (n = 65), one containing DHA (n = 65), or one containing both ARA and DHA (n = 66) to 1 year of age. A comparison group (n = 80) was exclusively breastfed for at least 3 months after which the infants continued to be exclusively breastfed or were supplemented with and/or weaned to infant formula. At 39 months, standard tests of IQ (Stanford Binet IQ), receptive vocabulary (Peabody Picture Vocabulary Test-Revised), expressive vocabulary (mean length of utterance), visual-motor function (Beery Visual-Motor Index), and visual acuity (Acuity Card Procedure) were administered. Growth, red blood cell fatty acid levels, and morbidity also were evaluated.

RESULTS: Results were analyzed using analysis of variance or linear regression models. The regression model for IQ, receptive and expressive language, and the visual-motor index controlled for site, birth weight, sex, maternal education, maternal age, and the child's age at testing. The regression model for visual acuity controlled for site only. A variable selection model also identified which of 22 potentially prognostic variables among different categories (feeding groups, the child and family demographics, indicators of illness since birth, and environment) were most influential for IQ and expressive vocabulary. A total of 157 (80%) of the 197 infants studied at 12 months participated in this follow-up study. Characteristics of the families were representative of US families with children up to 5 years of age, and there were no differences in the demographic or family characteristics among the randomized formula groups. As expected, the formula and breastfed groups differed in ethnicity, marital status, parental education, and the prevalence of smoking. Sex, ethnicity, gestational age at birth, and birth weight for those who participated at 39 months did not differ from those who did not. The 12-month Bayley mental and motor scores and 14-month vocabulary scores of the children who participated also were were not different from those who did not. At 39 months, IQ, receptive and expressive language, visual-motor function, and visual acuity were not different among the 3 randomized formula groups or between the breastfed and formula groups. The adjusted means for the control, ARA+DHA, DHA, and breastfed groups were as follows: IQ scores, 104, 101, 100, 106; Peabody Picture Vocabulary Test, 99.2, 97.2, 95.1, 97.4; mean length of utterance, 3.64, 3.75, 3.93, 4.08; the visual-motor index, 2.26, 2.24, 2.05, 2.40; and visual acuity (cycles/degree), 30.4, 27.9, 27.5, 28.6, respectively. IQ was positively associated with female sex and maternal education and negatively associated with the number of siblings and exposure to cigarette smoking in utero and/or postnatally. Expressive language also was positively associated with maternal education and negatively associated with the average hours in child care per week and hospitalizations since birth but only when the breastfed group was included in the analysis. The associations between maternal education and child IQ scores are consistent with previous reports as are the associations between prenatal exposure to cigarette smoke and IQ and early language development. Approximately one third of the variance for IQ was explained by sex, maternal education, the number of siblings, and exposure to cigarette smoke. Growth achievement, red blood cell fatty acid levels, and morbidity did not differ among groups.

CONCLUSIONS: We reported previously that infants who were fed an unsupplemented formula or one with DHA or with both DHA and ARA through 12 months or were breastfed showed no differences in mental and motor development, but those who were fed DHA without ARA had lower vocabulary scores on a standardized, parent-report instrument at 14 months of age when compared with infants who were fed the unsupplemented formula or who were breastfed. When the infants were reassessed at 39 months using age-appropriate tests of receptive and expressive language as well as IQ, visual-motor function and visual acuity, no differences among the formula groups or between the formula and breastfed groups were found. The 14-month observation thus may have been a transient effect of DHA (without ARA) supplementation on early vocabulary development or may have occurred by chance. The absence of differences in growth achievement adds to the evidence that DHA with or without ARA supports normal growth in full-term infants.

In conclusion, adding both DHA and ARA when supplementing infant formulas with long-chain polyunsaturated fatty acids supports visual and cognitive development through 39 months.