AJCN Cancer Health Disparities Conference
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McCann, J. C
Right arrow Articles by Ames, B. N
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McCann, J. C
Right arrow Articles by Ames, B. N
Agricola
Right arrow Articles by McCann, J. C
Right arrow Articles by Ames, B. N
American Journal of Clinical Nutrition, Vol. 82, No. 2, 281-295, August 2005
© 2005 American Society for Clinical Nutrition

REVIEW ARTICLE

Is docosahexaenoic acid, an n–3 long-chain polyunsaturated fatty acid, required for development of normal brain function? An overview of evidence from cognitive and behavioral tests in humans and animals1,2,3

Joyce C McCann and Bruce N Ames

1 From the Nutrition, Metabolism and Genomics Center, Children’s Hospital Oakland Research Institute, Oakland, CA (JCM and BNA)

2 Supported by the Bruce and Giovanna Ames Foundation (to JCM), by grant no. P60-MD00222-01 from the National Center on Minority Health and Health Disparities (to BNA), and by grant no. K05 AT001323-01 from the National Center of Complementary and Alternative Medicine (to BNA).

3 Reprints not available. Address correspondence to JC McCann, Children’s Hospital Oakland Research Institute, 5700 Martin Luther King Jr Way, Oakland, CA 94609. E-mail: jmccann{at}chori.org.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 HUMAN STUDIES
 STUDIES IN ANIMALS
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
This review is part of a series intended for nonspecialists that will summarize evidence relevant to the question of whether causal relations exist between micronutrient deficiencies and brain function. Here, we focus on experiments that used cognitive or behavioral tests as outcome measures in experimental designs that were known to or were likely to result in altered brain concentrations of the n–3 fatty acid docosahexaenoic acid (DHA) during the perinatal period of "brain growth spurt." Experimental designs reviewed include observational breastfeeding studies and randomized controlled trials in humans and studies in rodents and nonhuman primates. This review is based on a large number of expert reviews and commentaries and on some 50 recent studies in humans and animals that have not yet been included in published reviews. Expert opinion regarding the strengths and weaknesses of the major experimental systems and uncertainties associated with interpreting results is summarized. On the basis of our reading of this literature, we conclude that evidence from several types of studies, particularly studies in animals, suggests that, within the context of specific experimental designs, changes in brain concentrations of DHA are positively associated with changes in cognitive or behavioral performance. Additional experimental information required to conclude that a causal association exists is discussed, as are uncertainties associated with applying results from specific experimental designs to the question of whether infant formula should be supplemented with DHA.

Key Words: Long-chain polyunsaturated fatty acids • LCPUFAs • docosahexaenoic acid • DHA • linolenic acid • cognition • learning • memory • perinatal period • breastfeeding • formula feeding • essential fatty acids • n–3 fatty acids • brain • neurology • infants • childhood • rodent studies • nonhuman primate studies


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
HUMAN STUDIES
 STUDIES IN ANIMALS
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
During the last trimester of fetal life and the first 2 y of childhood, the brain undergoes a period of rapid growth termed the "brain growth spurt" (1). A large and dispersed literature suggests that an inadequate supply of any of a number of essential micronutrients during this period, as well as throughout life, can compromise brain function (2–9). In addition to the enormous literature examining the effects of individual micronutrients, some studies have examined the effects of multivitamin and mineral supplementation on cognitive function (10, 11).

If a causal relation between micronutrient deficiencies and suboptimal brain function exists, it has major public health implications. Large segments of the world population (including the United States), particularly the poor, are known to be undernourished in a number of micronutrients (12–19). A major effort to address micronutrient undernutrition, as an adjunct to the various programs underway to improve dietary habits, particularly among the poor, will be well justified. One of us has discussed such an approach as a relatively inexpensive and efficacious adjunct to current public health programs (20, 21). This review is the first of a series intended to provide the nonspecialist with critical summaries of the available experimental evidence pertinent to a discussion of this important public health question (20, 21).

The goal of this review is to provide an overview of human and animal in vivo experiments that link the availability during development of the long-chain polyunsaturated fatty acids (LCPUFAs), particularly the n–3 LCPUFA docosahexaenoic acid (DHA, 22:6n–3), to performance on cognitive or behavioral tests. Primary resources were recent research reports not yet reviewed, key earlier studies, and a large number of expert reviews and commentaries. We searched the literature by using a combination of techniques including key word and author searches of the National Library of Medicine’s PUBMED database and the Science Citation Index Cited References database. We also surveyed citations included in recent research and review articles. Abstracts were not included.

General background
LCPUFAs, particularly DHA and arachidonic acid (AA), are highly concentrated in cell membranes of the retina and brain (22–26), and they accumulate rapidly during the brain growth spurt (27, 28). Whereas AA is known to be involved in cell signaling pathways and to be a precursor of eicosanoids that are important for a number of cellular processes (29–31), specific functional roles for DHA in the brain are less well elucidated. Examples of research and reviews of this important topic are available in many publications (31–59).

Reduced visual acuity has consistently been observed in primate and rodent offspring subjected to dietary conditions during gestation that result in significant reductions in retinal concentrations of DHA (60–65). Some formula-comparison studies in humans, particularly in very-low-birth-weight (VLBW) and preterm infants, suggest that greater visual acuity in infancy is associated with increased intake of LCPUFAs (66–68), and these results are supported by several meta-analyses (69–71). The conclusions of systematic reviews differ somewhat, however, as to the overall strength of evidence for a positive effect of LCPUFA-supplemented infant formula on visual acuity (30, 69–74). Numerous publications provide additional review (4, 30, 57, 75–80). Effects of DHA availability during development on visual acuity are not a focus of this review but are discussed as a potential confounder of results in some cognitive and behavioral tests.

DHA is not widely distributed in the diet but is present in some foods, particularly fatty fish (30, 81–84). DHA can also be synthesized in the body from the essential fatty acid (EFA) {alpha}-linolenic acid (ALA). Although infants are able to convert ALA to DHA (85–88), DHA is known to be supplied directly to the fetus through the placenta and to the postnatal infant via breast milk (89–93). Uncertainty as to whether endogenous synthesis by the infant is sufficient to supply enough DHA during the postnatal phase of perinatal growth has stimulated a great deal of interest in whether infant formula should be supplemented with LCPUFAs. This general issue has been discussed in many reviews and commentaries (4, 30, 54, 79, 94–100). We do not take up this issue directly in this review, but we do comment on the relevance to this important question of the results from the experimental systems we reviewed.

Basic experimental designs
Three types of protocols were most commonly used in the studies reviewed. Each of these protocols differently manipulates the dietary supply of EFAs or LCPUFAs during development, and all result, or are likely to result, in different brain concentrations of DHA in comparison groups, which makes them relevant to a discussion of whether DHA availability in the brain is linked to cognitive or behavioral function.

Most observational breastfeeding studies in humans compare children who were either breastfed or formula-fed or those who were breastfed for different periods of time. Most randomized controlled trials (RCTs) in humans compare groups fed infant formulas that are either unsupplemented or supplemented with LCPUFAs. Most studies in animals compare offspring whose supply of ALA, the precursor of DHA, has been severely limited for different periods of time.

In most rodent ALA-restriction studies, brain concentrations of DHA are dramatically reduced in test animals, by as much as 85% (101), and decreases of 75% in nonhuman primates have been reported (64). It is important, however, that some studies in rodents also used dietary conditions that resulted in smaller decreases in brain DHA (40, 102–105). Human autopsy studies reported significant differences of {approx}11% to 40% in DHA concentrations in brain gray matter between breastfed and unsupplemented formula-fed infants (106–110); two further references provide review (4, 111). Direct autopsy evidence that compares brain DHA concentrations in human infants fed unsupplemented and LCPUFA-supplemented formulas is not available. However, a recent autopsy study in nonhuman primates reported {approx}30% lower concentrations of DHA in the visual cortex of preterm infants fed unsupplemented formula than in those fed LCPUFA-supplemented formula (112). In humans, significant differences in plasma concentrations of DHA between unsupplemented and supplemented formula comparison groups are well documented (113, 114).

It is important to note that, in all of these common experimental designs that affect DHA status, DHA is not the only variable. Even in RCTs, formulas typically are supplemented with other LCPUFAs, particularly AA, in addition to DHA. Thus, to different degrees, these designs have an inherent lack of specificity that is important to take into account in evaluating the significance of results relative specifically to DHA.

Performance tests
Many different kinds of tests have been used in human and animal studies. In this review, we focus primarily on tests aimed at assessing cognition or related mental functions, but a few tests assessing other correlates of functional development of the nervous system, such as neuromotor activity (115, 116) and sleep patterns (117), are also included. Several reviews address this important topic (4, 48, 57, 118–127). Tests in human studies include standardized global tests that screen broadly for cognitive-related functions, such as the Bayley Scales of Infant Development (128) and the Kaufman Assessment Battery for Children (129), and tests that target more specific functions, such as the development of language and communication skills [eg, the MacArthur Communicative Development Inventory (130)], visual recognition [eg, the Fagan Test of Infant Intelligence (131)], or problem solving (132, 133). As previously reviewed (57, 75, 118), performance tests in nonhuman primate studies have included assessments of behavior patterns (134, 135) and some tests that have also been used in human studies, most notably "look duration" tests of visual attention (136). Commonly used in rodent studies are tests that measure aspects of spatial learning such as the radial maze (137, 138) and Morris water maze (139–141), the elevated plus maze (a test that measures anxiety; 142, 143), and tests that measure stimulus-response type learning, such as brightness discrimination tests (144, 145). Several reviews provide further discussion of performance tests used in rodent studies (119, 144, 146, 147).

The fact that different tests measure different aspects of cognition or behavior must be taken into account in the interpretation and comparison of results across different tests. For example, in humans, it has been suggested that different results might be expected in global tests (such as the Bayley Scales of Infant Development) and in tests targeted to more specific neural domains (such as look duration tests), which may not equally measure specific neural functions that may be affected by LCPUFA availability (57, 124). In rodent tests, it has been suggested that a version of the Morris water maze that measures working memory may be more sensitive in detecting performance deficits in n–3–restricted animals than is the more commonly used "place" version of the test (148). Evaluation of results in performance tests must also consider potential confounders. For example, most of the tests listed above rely on vision or motor activity, and thus possible effects of DHA on visual or motor development could theoretically influence the outcome.


   HUMAN STUDIES
TOP
 ABSTRACT
 INTRODUCTION
 HUMAN STUDIES
STUDIES IN ANIMALS
 SUMMARY AND CONCLUSIONS
 REFERENCES
 
Breastfeeding studies
Information provided by observational breastfeeding studies, although relevant to the question of causality, is limited. As has frequently been noted (4, 80, 91, 126, 127, 149–153), the decision to breastfeed is associated with a number of factors that potentially confound positive results in these studies, such as socioeconomic status, home environment, and maternal intelligence quotient (IQ). Moreover, experimental designs involving breastfeeding are inherently limited in their ability to identify which among a number of potentially active ingredients in breast milk might be responsible for enhancement effects (4, 91, 127, 154, 155).

Five systematic reviews published since 1999 critically evaluated partially overlapping subsets of breastfeeding studies spanning >20 y (4, 152, 153, 156, 157). Most of the studies included in these reviews compared the performance of children who were breastfed or formula-fed. Before adjustment for covariables, most of these studies reported higher scores on performance tests for children who were breastfed. Reviewers were in agreement that a significant number of potentially confounding variables complicated the analysis and interpretation of these results. However, they agreed less as to whether there was some residual enhancement in test performance after adjustment for covariables. For example, one can compare the conclusions of the studies by Anderson et al (156) and Drane and Logemann (157) and of several critical commentaries (149, 150, 158).

Areas that reviewers suggested for future research included improvement in the overall quality of studies; comprehensive confounder analysis (153, 157); examination of longer durations of breastfeeding, with duration of breastfeeding as a dose-response variable (157); and more studies comparing breast milk–fed and formula-fed premature infants (152).

Breastfeeding studies published in the past 5 y (since 1999), along with formula-feeding studies, are listed in Table 1Go. The table includes an indication of the type of experimental design, the performance tests used, and conclusions of the investigators. Reviews that have cited these studies are also indicated in the table. The range of experimental approaches is greater in these newer studies than in those reviewed earlier, which is consistent with suggestions of earlier reviewers. Thus, whereas several of these studies compared breastfed and formula-fed children (160–167), 12 studies compared the performance of children who were breastfed for different periods of time (159, 168–178).


View this table:
[in this window]
[in a new window]
  		TABLE 1 Human studies (1999–2004) that compared effects of breastfeeding or long-chain polyunsaturated fatty acid (LCPUFA)–supplemented or unsupplemented formula feeding on cognitive or behavioral function1

 
The great majority of these studies reinforce conclusions of earlier work. Most studies that compared breastfed and formula-fed groups and those that examined groups that were breastfed for different periods of time observed a weak association between scores on performance tests and longer periods of breastfeeding, even after adjustment for at least some potential confounders. As also indicated in Table 1Go, positive results were reported across a wide range of different performance tests.

It is important to note, however, that these studies do not adjust for all possible confounders. As an example, one of the strongest potential confounders logically is—and has been shown experimentally to be—maternal IQ. As discussed previously (150), in the 1999 Anderson et al (156) meta-analysis of earlier studies, only 6 of the 20 breastfeeding studies cited included maternal IQ as a potential confounder. Among the more recent observational breastfeeding studies listed in Table 1Go, only 8 included maternal IQ as a potential confounder (159, 160, 169, 174, 177–180). Among those 8 studies, only 2 (159, 177) reported that better test performance by breastfed youngsters remained significantly different after adjustment for multiple confounders. However, a commentary on 1 of these 2 positive studies, that of Rao et al (177), suggested that additional confounders still might account for the residual positive results (149). Specifically, the commentators suggested the inclusion of a verbal measure of maternal IQ (ie, the Peabody Picture Vocabulary Test) and of a measure of parenting quality (such as the Home Observation Measurement of Environment). These measures were significant confounders in a study reported by the commentators (160) that lost statistical significance with adjustment (Table 1Go). It may be of interest that only one other study (178) among those discussed here used a verbal measure of maternal IQ [all other studies used the Raven’s Progressive Matrices (203, 204)]. That study also reported, as discussed above, that adjustment almost completely removed significance (178).

Randomized controlled trials
Studies comparing children fed differently supplemented formulas
Randomized controlled trials (RCTs) offer much greater opportunity than do observational studies for the control of experimental variables, including the quantity and composition of LCPUFA supplements. In addition, this design affords the opportunity to avoid many of the potential confounders that complicate the interpretation of observational breastfeeding studies. RCTs, however, are also subject to confounding and to other sources of uncertainty that can complicate the comparison of results in different trials, such as the use of different formulas or oils or the presence of other sources of variability in the DHA status of infant groups. A general review is available in numerous publications (4, 48, 54, 77, 122, 151, 154, 181, 182).

Results of RCTs have been discussed by many investigators and reviewers. We focused on several systematic reviews and meta-analyses involving term infants (69, 71, 74, 122), preterm infants (70, 73), or both (4, 72, 78–80, 182, 183). Although some of these reviews focused only on visual acuity, the most commonly measured outcome in human RCTs, most reviews also critically examined cognitive or behavioral outcomes (4, 72–74, 78–80, 182, 183). Conclusions varied somewhat among these reviewers as to the strength of the evidence for effects of LCPUFA supplementation on performance in cognitive or behavioral tests, and, as indicated in the Introduction, on visual acuity. The 2 Cochrane reviews (73, 74) pointed to 1 positive problem-solving study in term infants (205) and to 2 studies in preterm infants (206, 207) that reported shorter look durations, which are considered to indicate greater development of attentional processing. However, the Cochrane reviewers concluded overall that available evidence did not support suggestions that LCPUFA supplementation benefited neural development. Other reviewers acknowledged that results in RCTs were mixed, particularly in term infant studies (57, 72, 79) but also pointed out that tests measuring visual attention (206-208) provided consistently positive results in human RCTs in both preterm and term infants, which paralleled similar results in nonhuman primate studies (136).

As shown in Table 1Go, only 3 of the RCTs listed that compared LCPUFA-supplemented and -unsupplemented formula-fed children were not discussed in previous reviews. Auestad et al (163), in a 39-mo follow-up study of term infants, did not observe a significant difference in performance on several mental and motor development tests (or in visual acuity). Bouwstra et al (164), also in a study of term infants, reported positive effects (P < 0.05) of LCPUFA supplementation at 3 mo of age by using a test that assessed the quality of general movements considered to be an indicator of brain function. And Fewtrell et al (184) reported weakly significant (P = 0.04) enhancement effects in 18-mo-old children fed fish oil–supplemented formula as preterm infants in the Mental Development Index component of the Bayley Scales.

Other experimental designs
The study of Colombo et al (185) listed in Table 1Go adds to the series of positive results obtained in look duration tests. This study reported a positive correlation between maternal erythrocyte phospholipid DHA at birth and performance in look duration tests over the first 18 mo of life. Other studies listed in Table 1Go that were not previously discussed by reviewers include a negative study that examined possible correlations between performance in several mental development tests at 4 and 7 y of age and LCPUFA blood concentrations at birth or at the time of testing (179, 180), a study reporting a positive correlation between maternal plasma phospholipid DHA concentrations and the maturity of sleep patterning in infants (186), and a study, some results of which were significant, that compared performance in several mental and motor development tests of {approx}1-y-old term infants fed infant formulas differing in ALA content (187).


   STUDIES IN ANIMALS
TOP
 ABSTRACT
 INTRODUCTION
 HUMAN STUDIES
 STUDIES IN ANIMALS
SUMMARY AND CONCLUSIONS
 REFERENCES
 
The great strength of animal studies is that they afford the opportunity for more flexibility in design and in the ability to control experimental variables than can be achieved in human studies. The most commonly used experimental design in animals involves limiting the dietary supply of ALA, the precursor of DHA, which is not feasible in human studies because ALA is an EFA. Previous reviewers of animal studies (49, 57, 119, 146, 147) concluded that reproducible effects of dietary restriction of ALA on look duration in nonhuman primates suggested possible involvement of DHA in cortical pathways associated with visual attention (57) and also concluded that effects on cognition in rodent studies were shown in tests that involved sensory pathways other than the visual (31, 49, 209-211).

Most animal studies have been conducted by using rodents. Although a detailed comprehensive methodologic review in 1992 (146) discussed early rodent studies, and more recent reviews have critically discussed some subsequent work (49, 57, 119, 147), there is now quite a large database of rodent behavioral studies that have not been critically reviewed. More than 30 rodent studies (102–105, 148, 212–238) have been published since the 1992 comprehensive methodologic review of Wainwright (146). More than 20 of these reports were published since 1999 (Table 2Go). (Investigations that focused solely on visual function are not included here.) The results reported in these studies are briefly summarized below.


View this table:
[in this window]
[in a new window]
  		TABLE 2 Animal studies (1999–2004) that tested effects of diets differing in n–3 fatty acid content during development on behavior1

 
First, almost all of the 20 rodent studies listed in Table 2Go reported significant deficits in test performance by subjects compared with controls on at least some tasks in at least some n–3–restricted groups. These results support earlier studies, previously reviewed (49, 57, 119, 146, 147), that suggested an association between a diet severely restricted in n–3 FAs during development and poorer performance of offspring in tests designed to measure cognitive or behavioral ability. This important basic conclusion is tempered by the fact that performance differences are typically not large, are observed on some tasks and not others, often are not observed across multiple test sessions, and have not been observed in all reports (further discussed below).

Second, performance deficits were reported in a variety of tests, which suggests that the effect on performance is not test-specific. For example, among studies listed in Table 2Go, tests for which effects were reported included the Morris water maze (105, 212, 215-217, 219, 220, 232), 8-arm radial maze (229, 230), elevated plus maze (220, 234), active avoidance (227, 235), olfactory-cued (214, 216), and brightness discrimination (224, 227) tests.

Third, 8 of the studies in Table 2Go supplemented n–3–restricted animals with DHA, DHA+AA, DHA-rich oils, or DHA and additional n–6 FAs and compared the animals’ performance with that of n–3–restricted controls (40, 212, 217, 224, 229, 230, 234, 235). All of these studies reported that performance was significantly enhanced in the supplemented groups.

Fourth, all but 5 studies (218, 219, 233-235) listed in Table 2Go also measured brain concentrations of LCPUFAs. In almost all cases, brain DHA concentrations were significantly lower in n–3–restricted groups than in controls, and several investigations pointed to the importance of AA as well (224, 227, 228). In addition to conducting performance tests and determining brain concentrations of LCPUFAs, several studies recorded changes in biochemical indicators of brain function in n–3–restricted and control groups (226, 229, 235).

Fifth, limited evidence of a dose-response effect is reported in 2 studies (215, 240). We also note several reports in which performance deficits were observed in some groups by using dietary protocols that resulted in significantly less reduction in brain DHA than is seen in most multigeneration ALA-restriction studies (38, 40, 105, 148).

Factors involved in assessing and interpreting studies in rodents
An in-depth methodologic review of all of the studies discussed above is beyond the scope of this review. We note, however, that a thorough critical evaluation of results should take into account a number of methodologic and other issues, many of which have been discussed by previous reviewers (49, 51, 52, 55, 57, 118, 119, 146, 147, 209, 241, 242). Several of these issues are briefly summarized below.

Potential confounders not usually evaluated cannot be excluded as alternative explanations for results in many experiments, as is discussed in several reviews (4, 57, 119, 243, 244). Most tests measure a combination of performance and behavioral characteristics in addition to cognition per se; Wainwright (119) and Wainwright et al (148) provide further discussion of this topic. For example, in addition to learning ability, commonly used water and radial maze tests involve locomotor ability and visual recognition of cues, and brightness discrimination tests require visual ability. As discussed above, chronic dietary restriction of ALA adversely affects the development of visual function, which must be considered to potentially influence the results of these tests. Other potential confounders are long-term severe imbalance in the ratio of n–6 to n–3 FAs or other effects of chronic ALA restriction. In addition to its role as precursor of DHA, ALA is an important source of energy, is a precursor for the synthesis of saturated and monounsaturated FAs (51, 245)—that comprise some 75% (by wt) of the FAs present in brain tissue (4), including myelin (246)—and is believed to be required for the development of an adequate arterial supply (245, 247). In addition, the very high n–6:n–3 used to generate n–3–deficient groups in the rat experiments may result in an unnaturally high level of inflammatory activity (248, 249), which could affect the general health or vigor of the animals and thus influence their ability to perform in cognitive and behavioral tests.

Statistical considerations not taken into account can affect conclusions about the positivity (or negativity) of experiments, as well as the interpretation of positive results. Wainwright (250) has discussed several design and analysis issues with respect to multigenerational studies in multiparious species, and she emphasizes that weak or marginal results must be viewed with caution. With reference to the studies included in this review, we point particularly to 3 important statistical considerations. First, as Wainwright discussed (250), the inflation of sample size by using the number of pups instead of the number of litters as the unit of measure results in an increased potential for false positives unless appropriate cluster design adjustments are made. Second, in correlational analyses, if group means for the measures being correlated are significantly different, a simple correlation across groups does not necessarily indicate a correlation for individual animals. Third, because "learning" implies a change over time, it is inappropriate to conclude that there is a learning difference between compared groups unless the statistical analysis indicates a group x time interaction.

As discussed by many reviewers (4, 119, 147), test results obtained in experiments involving severe depletion of brain DHA have uncertain relevance to effects that might be expected under less severe conditions. In nonhuman primates (64) and piglets (38), DHA concentrations in the brain are dramatically reduced after the maternal dietary supply of ALA is limited during one gestational period. In rodents, a longer period of chronic ALA restriction is required to achieve a comparable reduction because of athe more efficient biosynthesis of DHA from ALA in rodents (251). Thus, most rodent studies have been conducted with animals whose mothers or even mothers and grandmothers were raised on diets limited in ALA. In animals whose supply of DHA has been limited in this way, during the perinatal period (when rapid accumulation of brain DHA would normally occur), the place of DHA is taken primarily by another FA, docosapentaenoic acid (DPA; 22:5n–6), which differs from DHA in just one double bond—the n–3 double bond (101, 252).

Performance of n–3–restricted rats and mice in the Morris water maze
The Morris water maze is the most commonly used learning and memory test in the rodent studies reviewed (102–105, 148, 212, 215-217, 219, 220, 223, 232, 236). Below, we briefly discuss the subset of these studies that included a determination of brain concentrations of DHA and compared an n–3–restricted group to "n–3–adequate" controls. It is noted that the 10 studies selected (103–105, 148, 212, 215-217, 220, 223) included additional outcome measures or other comparisons not discussed in this example. Some of these studies have been reviewed elsewhere (49, 119, 241).

In rats, one laboratory reported reproducibly better performance of controls than of severely n–3–restricted animals in the "place" version of the test (212, 215, 216). Using the same version of the test, a different laboratory reported a nonsignificant trend (P > 0.08) in one study (148) and a negative result for a "saturated fat" group in a larger study (105). The latter 2 studies also used the "delayed-matching-to-place" or "working memory" version of the test, and both reported significant positive results [a diet x trials interaction, P < 0.01 (148) or P < 0.05 in a planned comparison (105)].

In mice, a positive result (P < 0.05) was reported for a saturated fat group, but not for an n–3–restricted group in the same study (103). Positive results were also reported in 2 other studies (217, 223). However, in one of those studies (217), significance relied on a post hoc comparison of results in a single session conducted on a data set after a nonsignificant ANOVA, and, in the other study (223), significance was marginal (P = 0.044) and relied on the use of a one-tailed test. Two other mouse studies reported negative results (104, 220).

Although rats are known to perform better than mice in the water maze (253–255), it may be of interest that, among the studies discussed above, positive results were obtained in both species in the place version of the test when reductions in brain concentrations of DHA in the restricted group exceeded 80% (103, 212, 215, 216). Results were negative or marginal when brain reductions were less, ie, 25% (217), 41% (220), 50% (103), 60% (223), and 53% (104). It may also be of interest that positive results were obtained in rats with less severe depletion of brain DHA, ie, 38% (105) and 51% (148), in the delayed-matching-to-place version of the water maze, which, as discussed previously (148), may be more sensitive for detection of effects of n–3 restriction than is the place version.

Docosahexaenoic acid supplementation studies
Only a few studies in animals have examined effects of LCPUFA supplementation of normal diets during development on performance in cognitive or behavioral tests. Among the rodent studies listed in Table 2Go, only one (105) used a study design analogous to that of formula comparison studies in humans. That study did not find that supplementation with DHA or AA (or both) improved subsequent performance on 2 versions of the Morris water maze, despite the fact that feeding diets with normal concentrations of EFAs supplemented with different amounts of AA and DHA to artificially reared pups was found to result in a range of different concentrations of AA and DHA in forebrain phospholipids (256). One other study listed in Table 2Go included dietary groups for which n–6:n–3 was within the range considered roughly normal (ie, {approx}5–10) or enriched for n–3 FAs (219). That study reported some changes in performance on the Morris water maze in animals fed the n–3–enriched diet but also reported adverse effects in older animals fed the n–3–enriched diet for extended periods. A study in rhesus macaque infants (239) reported that those fed LCPUFA-supplemented formula had stronger orienting and motor skills than did those fed standard formula.


   SUMMARY AND CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 HUMAN STUDIES
 STUDIES IN ANIMALS
 SUMMARY AND CONCLUSIONS
REFERENCES
 
Each of the 3 major experimental designs reviewed contributes some evidence that is relevant to a discussion of possible causal linkages between altered brain concentrations of DHA during the perinatal period and subsequent cognitive or behavioral performance. (As indicated above, effects of DHA availability during development on visual function are not a subject of this review.) Evidence from chronic dietary restriction rodent studies that is most relevant to the issue of causality shows that the addition of DHA to diets of animals whose brain concentrations of DHA have been severely reduced restored control performance levels (40, 212, 217, 224, 229, 230, 234, 235). As discussed in this review, the relevance of these results to effects that might be expected in humans under less severe conditions is uncertain. Formula comparison and maternal supplementation studies in humans and ALA dietary restriction studies in nonhuman primates both link the availability of n–3 LCPUFAs to the development of visual attention (136, 185, 206-208), although, as discussed, it is difficult to exclude as an alternative explanation for these results the possible confounding due to effects of DHA on visual function. We also point to formula supplementation (164, 239) or maternal plasma correlational (186) studies in humans or monkeys that suggest enhanced neuromotor development in infants with higher DHA status. Positive results in breastfeeding studies are seriously eroded by adjustment for multiple covariates, and residual positive effects cannot be attributed to LCPUFAs because of other potentially active constituents in breast milk. Nevertheless, it is of interest that these studies consistently showed a positive association between breastfeeding and performance across a wide range of different tests, even after adjustment for at least some potential confounders.

Clearly, the experiments most capable of providing definitive evidence that is directly relevant to human exposure conditions are RCTs in humans. As reviewed above, these trials have often not shown an effect of LCPUFA supplementation on cognitive or behavioral performance, and some reviewers have considered that, overall, the evidence was insufficient to conclude that LCPUFA supplementation benefited development (73, 74). Understanding why most RCTs have yielded mixed results, particularly in light of more consistently positive effects observed in rodent tests and human breastfeeding studies, is of obvious importance and has been discussed by a number of reviewers and investigators (4, 49, 77, 79, 122). The suggestion of previous reviewers (57, 124) that differences in the sensitivity of global tests, such as the Bayley Scales, and of tests targeted at more specific neural domains, such as look duration or problem-solving tests, might account for mixed results in RCTs was discussed earlier in this review. Other possible explanations include inadequate supplementation of DHA in formulas (49, 257), poor study quality (77), the ability of term infants fed unsupplemented formulas to synthesize their own DHA (73), an absence of cognitive deficits when differences in brain concentrations of DHA are small due to brain plasticity (ie, the ability of the brain to adapt), or the inability of performance tests to detect subtle differences in performance that result from relatively small differences in brain concentrations of DHA (105, 210).

The relative merits of hypotheses that invoke brain plasticity or difficult-to-detect weak effects due to small decreases in brain DHA are important to follow up. The former hypothesis suggests that there may be no adverse consequences of relatively small reductions in brain DHA, whereas the latter suggests that relatively small reductions could result in subtle performance deficits that are difficult to detect. If experts can agree on a reliable marker, constructing a dose-response curve that relates the degree of reduction in brain DHA to effects on neural function would be of great value. It appears to us from our reading of the literature that cognitive and behavioral tests may not be sensitive or consistent enough to permit such an analysis, but some biochemical markers, such as dopamine inducibility (36, 38), may be useful.

Most of the explanations discussed above focus on the possibility that negative results in RCTs may be false-negative. Following some discussion of this point in the literature (146), we note that it is equally important to consider the possibility that positive results are false positives. In addition to the need to better identify confounders and methodologic sources of possible bias, as discussed, we note the value of independent replication to confirm positive results.

It is of interest to examine the degree to which evidence from the 3 major experimental systems reviewed satisfies conditions of causality. Causal criteria [slightly adapted from the original formulation (258)] require evidence of 1) a consistent association, 2) a plausible biological rationale, 3) an ability to experimentally manipulate the effect, 4) a dose-response relation, and 5) specificity of cause and effect. Whereas some of these 5 criteria are satisfied by some experimental results reviewed, others are not. First, as detailed above in this review, all 3 major experimental systems reviewed (ie, human breastfeeding studies, human formula comparison RCTs, and animal ALA dietary restriction studies) show consistent associations under certain conditions. Second, mechanistic studies are not discussed in this review, but plausible biological rationales have been suggested, although not proven; recent references and reviews are cited in the Introduction. Third, experiments in rodents that restored performance by supplementing severely ALA-restricted animals with DHA provided some evidence of ability to manipulate the effect. Fourth, as indicated above in this review, limited evidence suggesting a possible dose-response effect was provided in 2 rodent studies (215, 240) and in a study that correlated concentrations of DHA in the blood of breastfed infants with enhanced performance (188). The only studies reviewed that provided consistent evidence of a dose-response effect are breastfeeding studies that correlate duration of breastfeeding with improvements in test performance (see Table 1Go). Fifth and finally, as discussed above, the interpretation of results in all 3 experimental designs is complicated by uncertainty as to the specificity of cause and effect. With respect to "cause," DHA is not specifically pointed to in breastfeeding studies because there are numerous factors in breast milk besides DHA that could affect the outcome. DHA is also not specifically implicated in ALA-restriction studies in animals because of FA imbalance or other possible effects of severe and chronic ALA restriction. Human formula–comparison studies are the least uncertain with regard to the specificity criterion. However, most of these studies do not supplement formulas only with DHA. With respect to "effect," possible effects of DHA on endpoints other than cognition or behavior, such as visual function, could bias the results of numerous tests, as indicated above. Thus, on the basis of these 5 causal criteria, the experimental results reviewed here supply some evidence that is not inconsistent with a causal connection between DHA availability and cognitive function, but they do not show causality.

Finally, we comment briefly on our view of the relevance of results reviewed here to the question of whether infant formula should be supplemented with DHA. The experiments most directly relevant to this question obviously are human RCTs that compare performance in children fed LCPUFA-supplemented or -unsupplemented formulas. As discussed, results from these RCTs are mixed and have not consistently shown a positive effect of supplementation on cognitive or behavioral function other than intriguing results from a relatively small number of studies. Results of human breastfeeding studies, though seriously confounded, are not inconsistent with a need for supplementation, but the studies do not provide direct or clear evidence. We consider that animal studies provide the most convincing and consistent evidence linking a decrease in brain concentrations of DHA to altered performance on cognitive or behavioral tests. However, effects are not large, despite the fact that the studies were conducted under severe dietary conditions, and results are difficult to extrapolate to the human situation. In our view, the main contribution of these animal studies to the discussion of infant formula supplementation is that they suggest the possibility, as discussed by others, eg, Wainwright (119), that small differences in brain concentrations of DHA, such as most likely occur between infants fed supplemented or unsupplemented formulas, may result in subtle effects that currently are difficult to detect but could be significant.


   ACKNOWLEDGMENTS
 
We thank Kathleen Boyle for assembling a reference collection on micronutrients and cognition and Teresa Klask for help with the manuscript. We are grateful for comments on an earlier version of this manuscript from Walter Willett, Francesco Visioli, Jane Higdon, Harold Helbock, and Lynn Wallock. We are especially grateful to Francesco Visioli for his cheerful support and helpful input on manuscript revisions, to Mark Hudes for statistical help, and to the AJCN reviewers whose constructive comments improved the quality of the review. We also thank James Ntambi for help with perinatal long-chain polyunsaturated fatty acid physiology and biochemistry.

JCM searched and analyzed the literature and wrote the review. BNA suggested the topic, gave advice, and edited the review. Neither of the authors had a personal or financial conflict of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 HUMAN STUDIES
 STUDIES IN ANIMALS
 SUMMARY AND CONCLUSIONS
 REFERENCES
 

  1. Kolb B, Whishaw IQ. An introduction to brain and behavior. New York: Worth Publishers, 2001.
  2. Beard JL, Connor JR. Iron status and neural functioning. Annu Rev Nutr 2003;23:41–58.[Medline]
  3. Black MM. The evidence linking zinc deficiency with children’s cognitive and motor functioning. J Nutr 2003;133:1473S–6S.[Abstract/Free Full Text]
  4. Lauritzen L, Hansen HS, Jorgensen MH, Michaelsen KF. The essentiality of long-chain n–3 fatty acids in relation to development and function of the brain and retina. Prog Lipid Res 2001;40:1–94.[Medline]
  5. Mattson MP, Shea TB. Folate and homocysteine metabolism in neural plasticity and neurodegenerative disorders. Trends Neurosci 2003;26:137–46.[Medline]
  6. Morris MS. Homocysteine and Alzheimer’s disease. Lancet Neurol 2003;2:425–8.[Medline]
  7. Zeisel SH. Choline: needed for normal development of memory. J Am Coll Nutr 2000;19:528S–31S.[Abstract/Free Full Text]
  8. Meck WH, Williams CL. Metabolic imprinting of choline by its availability during gestation: implications for memory and attentional processing across the lifespan. Neurosci Biobehav Rev 2003;27:385–99.[Medline]
  9. Giacobini E. Cholinergic function and Alzheimer’s disease. Int J Geriatr Psychiatry 2003;18:S1–5.[Medline]
  10. Benton D. Micro-nutrient supplementation and the intelligence of children. Neurosci Biobehav Rev 2001;25:297–309.[Medline]
  11. Schoenthaler SJ, Bier ID, Young K, Nichols D, Jansenns S. The effect of vitamin-mineral supplementation on the intelligence of American schoolchildren: a randomized, double-blind placebo-controlled trial. J Altern Complement Med 2000;6:19–29.[Medline]
  12. Bodnar LM, Cogswell ME, Scanlon KS. Low income postpartum women are at risk of iron deficiency. J Nutr 2002;132:2298–302.[Abstract/Free Full Text]
  13. Bogen DL, Krause JP, Serwint JR. Outcome of children identified as anemic by routine screening in an inner-city clinic. Arch Pediatr Adolesc Med 2001;155:366–71.[Abstract/Free Full Text]
  14. Siega-Riz AM, Popkin BM. Dietary trends among low socioeconomic status women of childbearing age in the United States from 1977 to 1996: a comparison among ethnic groups. J Am Med Womens Assoc 2001;56:44–8, 72.
  15. Lee JS, Frongillo EA Jr. Nutritional and health consequences are associated with food insecurity among U.S. elderly persons. J Nutr 2001;131:1503–9.[Abstract/Free Full Text]
  16. Dixon LB, Winkleby MA, Radimer KL. Dietary intakes and serum nutrients differ between adults from food-insufficient and food-sufficient families: third National Health and Nutrition Examination Survey, 1988–1994. J Nutr 2001;131:1232–46.[Abstract/Free Full Text]
  17. Kumanyika S, Krebs-Smith S. Preventive nutrition issues in ethnic and socioeconomic groups in the United States. In: Bendich A, Deckelbaum RJ, eds. Primary and secondary preventive nutrition. Totowa, NJ: Humana Press, 2001:325–56.
  18. Eden AN, Mir MA. Iron deficiency in 1- to 3-year-old children. A pediatric failure? Arch Pediatr Adolesc Med 1997;151:986–8.[Abstract]
  19. Liu T, Soong SJ, Wang X, Wilson NP, Craig CB. African American and white differences in nutritional status among low-income women attending public health clinics. J Health Care Poor Underserved 1996;7:323–37.[Medline]
  20. Ames BN. A role for supplements in optimizing health: the metabolic tune-up. Arch Biochem Biophys 2004;423:227–34.[Medline]
  21. Ames BN. The metabolic tune-up: metabolic harmony and disease prevention. J Nutr 2003;133(suppl):15445–85.
  22. Svennerholm L. Distribution and fatty acid composition of phosphoglycerides in normal human brain. J Lipid Res 1968;9:570–9.[Abstract]
  23. Anderson RE, Maude MB, Zimmerman W. Lipids of ocular tissues—X. Lipid composition of subcellular fractions of bovine retina. Vision Res 1975;15:1087–90.[Medline]
  24. Martin RE, Bazan NG. Changing fatty acid content of growth cone lipids prior to synaptogenesis. J Neurochem 1992;59:318–25.[Medline]
  25. Sastry PS. Lipids of nervous tissue: composition and metabolism. Prog Lipid Res 1985;24:69–176.[Medline]
  26. Giusto NM, Pasquare SJ, Salvador GA, Castagnet PI, Roque ME, Ilincheta de Boschero MG. Lipid metabolism in vertebrate retinal rod outer segments. Prog Lipid Res 2000;39:315–91.[Medline]
  27. Martinez M. Tissue levels of polyunsaturated fatty acids during early human development. J Pediatr 1992;120:S129–38.[Medline]
  28. Clandinin MT, Chappell JE, Leong S, Heim T, Swyer PR, Chance GW. Intrauterine fatty acid accretion rates in human brain: implications for fatty acid requirements. Early Hum Dev 1980;4:121–9.[Medline]
  29. Bhattacharya SK, Dasgupta G, Sen AP. Prostaglandins modulate central serotonergic neurotransmission. Indian J Exp Biol 1989;27:393–8.[Medline]
  30. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids (macronutrients). Washington, DC: National Academy Press, 2002.
  31. Wainwright P. Essential fatty acids and behaviour: is there a role for eicosanoids? In: Mostofsky SYD, ed. Handbook of essential fatty acid biology: biochemistry, physiology and behavioural neurobiology. Totowa, NJ: Humana Press, 1997:299–341.
  32. Delion S, Chalon S, Guilloteau D, Besnard JC, Durand G. alpha-Linolenic acid dietary deficiency alters age-related changes of dopaminergic and serotoninergic neurotransmission in the rat frontal cortex. J Neurochem 1996;66:1582–91.[Medline]
  33. Aid S, Vancassel S, Poumes-Ballihaut C, Chalon S, Guesnet P, Lavialle M. Effect of a diet-induced n–3 PUFA depletion on cholinergic parameters in the rat hippocampus. J Lipid Res 2003;44:1545–51.[Abstract/Free Full Text]
  34. Kodas E, Galineau L, Bodard S, et al. Serotoninergic neurotransmission is affected by n–3 polyunsaturated fatty acids in the rat. J Neurochem 2004;89:695–702.[Medline]
  35. Zimmer L, Delion-Vancassel S, Durand G, et al. Modification of dopamine neurotransmission in the nucleus accumbens of rats deficient in n–3 polyunsaturated fatty acids. J Lipid Res 2000;41:32–40.[Abstract/Free Full Text]
  36. Zimmer L, Vancassel S, Cantagrel S, et al. The dopamine mesocorticolimbic pathway is affected by deficiency in n–3 polyunsaturated fatty acids. Am J Clin Nutr 2002;75:662–7.[Abstract/Free Full Text]
  37. de la Presa Owens S, Innis SM. Docosahexaenoic and arachidonic acid prevent a decrease in dopaminergic and serotoninergic neurotransmitters in frontal cortex caused by a linoleic and alpha-linolenic acid deficient diet in formula-fed piglets. J Nutr 1999;129:2088–93.[Abstract/Free Full Text]
  38. Ng KF, Innis SM. Behavioral responses are altered in piglets with decreased frontal cortex docosahexaenoic acid. J Nutr 2003;133:3222–7.[Abstract/Free Full Text]
  39. Lesa GM, Palfreyman M, Hall DH, et al. Long chain polyunsaturated fatty acids are required for efficient neurotransmission in C. elegans. J Cell Sci 2003;116:4965–75.[Abstract/Free Full Text]
  40. Levant B, Radel JD, Carlson SE. Decreased brain docosahexaenoic acid during development alters dopamine-related behaviors in adult rats that are differentially affected by dietary remediation. Behav Brain Res 2004;152:49–57.[Medline]
  41. Ikemoto A, Nitta A, Furukawa S, et al. Dietary n–3 fatty acid deficiency decreases nerve growth factor content in rat hippocampus. Neurosci Lett 2000;285:99–102.[Medline]
  42. Barcelo-Coblijn G, Kitajka K, Puskas LG, et al. Gene expression and molecular composition of phospholipids in rat brain in relation to dietary n–6 to n–3 fatty acid ratio. Biochim Biophys Acta 2003;1632:72–9.[Medline]
  43. Mitchell DC, Niu SL, Litman BJ. DHA-rich phospholipids optimize G-protein-coupled signaling. J Pediatr 2003;143:S80–6.[Medline]
  44. Bowen RA, Clandinin MT. Dietary low linolenic acid compared with docosahexaenoic acid alter synaptic plasma membrane phospholipid fatty acid composition and sodium-potassium ATPase kinetics in developing rats. J Neurochem 2002;83:764–74.[Medline]
  45. Ahmad A, Murthy M, Greiner RS, Moriguchi T, Salem N Jr. A decrease in cell size accompanies a loss of docosahexaenoate in the rat hippocampus. Nutr Neurosci 2002;5:103–13.[Medline]
  46. Ahmad A, Moriguchi T, Salem N. Decrease in neuron size in docosahexaenoic acid-deficient brain. Pediatr Neurol 2002;26:210–8.[Medline]
  47. Kitajka K, Sinclair AJ, Weisinger RS, et al. Effects of dietary omega-3 polyunsaturated fatty acids on brain gene expression. Proc Natl Acad Sci U S A 2004;101:10931–6.[Abstract/Free Full Text]
  48. Innis SM. The role of dietary n–6 and n–3 fatty acids in the developing brain. Dev Neurosci 2000;22:474–80.[Medline]
  49. Salem N Jr, Litman B, Kim HY, Gawrisch K. Mechanisms of action of docosahexaenoic acid in the nervous system. Lipids 2001;36:945–59.[Medline]
  50. Salvati S, Attorri L, Avellino C, Di Biase A, Sanchez M. Diet, lipids and brain development. Dev Neurosci 2000;22:481–7.[Medline]
  51. Sinclair AJ, Attar-Bashi NM, Li D. What is the role of alpha-linolenic acid for mammals? Lipids 2002;37:1113–23.[Medline]
  52. Innis SM. Perinatal biochemistry and physiology of long-chain polyunsaturated fatty acids. J Pediatr 2003;143:S1–8.[Medline]
  53. Kurlak LO, Stephenson TJ. Plausible explanations for effects of long-chain polyunsaturated fatty acids (LCPUFA) on neonates. Arch Dis Child Fetal Neonatal Ed 1999;80:F148–54.[Free Full Text]
  54. Gibson RA, Makrides M. Long-chain polyunsaturated fatty acids in breast milk: are they essential? Adv Exp Med Biol 2001;501:375–83.[Medline]
  55. Das UN. Long-chain polyunsaturated fatty acids in the growth and development of the brain and memory. Nutrition 2003;19:62–5.[Medline]
  56. Chalon S, Vancassel S, Zimmer L, Guilloteau D, Durand G. Polyunsaturated fatty acids and cerebral function: focus on monoaminergic neurotransmission. Lipids 2001;36:937–44.[Medline]
  57. Carlson SE, Neuringer M. Polyunsaturated fatty acid status and neurodevelopment: a summary and critical analysis of the literature. Lipids 1999;34:171–8.[Medline]
  58. Tanabe Y, Hashimoto M, Sugioka K, et al. Improvement of spatial cognition with dietary docosahexaenoic acid is associated with an increase in Fos expression in rat CA1 hippocampus. Clin Exp Pharmacol Physiol 2004;31:700–3.[Medline]
  59. Calderon F, Kim HY. Docosahexaenoic acid promotes neurite growth in hippocampal neurons. J Neurochem 2004;90:979–88.[Medline]
  60. Anderson R, Benolken R, Dudley P, Landis D, Wheeler T. Polyunsaturated fatty acids of photoreceptor membranes. Exp Eye Res 1974;18:205–13.[Medline]
  61. Benolken RM, Anderson RE, Wheeler TG. Membrane fatty acids associated with the electrical response in visual excitation. Science 1973;182:1253–4.[Abstract/Free Full Text]
  62. Bourre JM, Francois M, Youyou A, et al. The effects of dietary alpha-linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, resistance to poisons and performance of learning tasks in rats. J Nutr 1989;119:1880–92.
  63. Neuringer M, Connor WE, Van Petten C, Barstad L. Dietary omega-3 fatty acid deficiency and visual loss in infant rhesus monkeys. J Clin Invest 1984;73:272–6.
  64. Neuringer M, Connor WE, Lin DS, Barstad L, Luck S. Biochemical and functional effects of prenatal and postnatal omega 3 fatty acid deficiency on retina and brain in rhesus monkeys. Proc Natl Acad Sci U S A 1986;83:4021–5.[Abstract/Free Full Text]
  65. Wheeler TG, Benolken RM, Anderson RE. Visual membranes: specificity of fatty acid precursors for the electrical response to illumination. Science 1975;188:1312–4.[Abstract/Free Full Text]
  66. Birch DG, Birch EE, Hoffman DR, Uauy RD. Retinal development in very-low-birth-weight infants fed diets differing in omega-3 fatty acids. Invest Ophthalmol Vis Sci 1992;33:2365–76.[Abstract/Free Full Text]
  67. Birch EE, Birch DG, Hoffman DR, Uauy R. Dietary essential fatty acid supply and visual acuity development. Invest Ophthalmol Vis Sci 1992;33:3242–53.[Abstract/Free Full Text]
  68. Carlson SE, Werkman SH, Rhodes PG, Tolley EA. Visual-acuity development in healthy preterm infants: effect of marine-oil supplementation. Am J Clin Nutr 1993;58:35–42.[Abstract/Free Full Text]
  69. Uauy R, Hoffman DR, Mena P, Llanos A, Birch EE. Term infant studies of DHA and ARA supplementation on neurodevelopment: results of randomized controlled trials. J Pediatr 2003;143:S17–25.[Medline]
  70. SanGiovanni JP, Parra-Cabrera S, Colditz GA, Berkey CS, Dwyer JT. Meta-analysis of dietary essential fatty acids and long-chain polyunsaturated fatty acids as they relate to visual resolution acuity in healthy preterm infants. Pediatrics 2000;105:1292–8.[Abstract/Free Full Text]
  71. San Giovanni JP, Berkey CS, Dwyer JT, Colditz GA. Dietary essential fatty acids, long-chain polyunsaturated fatty acids, and visual resolution acuity in healthy fullterm infants: a systematic review. Early Hum Dev 2000;57:165–88.[Medline]
  72. Uauy R, Hoffman DR, Peirano P, Birch DG, Birch EE. Essential fatty acids in visual and brain development. Lipids 2001;36:885–95.[Medline]
  73. Simmer K, Patole S. Longchain polyunsaturated fatty acid supplementation in preterm infants. Cochrane Database Syst Rev 2004:CD000375.
  74. Simmer K. Longchain polyunsaturated fatty acid supplementation in infants born at term (Cochrane Review). Cochrane Database Syst Rev 2004:1–25.
  75. Neuringer M, Jeffrey BG. Visual development: neural basis and new assessment methods. J Pediatr 2003;143:S87–95.[Medline]
  76. Neuringer M. Infant vision and retinal function in studies of dietary long-chain polyunsaturated fatty acids: methods, results, and implications. Am J Clin Nutr 2000;71:256S–67S.[Abstract/Free Full Text]
  77. Gibson RA, Makrides M. Polyunsaturated fatty acids and infant visual development: a critical appraisal of randomized clinical trials. Lipids 1999;34:179–84.[Medline]
  78. Gibson RA, Chen W, Makrides M. Randomized trials with polyunsaturated fatty acid interventions in preterm and term infants: functional and clinical outcomes. Lipids 2001;36:873–83.[Medline]
  79. Koletzko B, Agostoni C, Carlson SE, et al. Long chain polyunsaturated fatty acids (LC-PUFA) and perinatal development. Acta Paediatr 2001;90:460–4.[Medline]
  80. Cockburn F. Role of infant dietary long-chain polyunsaturated fatty acids, liposoluble vitamins, cholesterol and lecithin on psychomotor development. Acta Paediatr Suppl 2003;92:19–33.
  81. Visioli F, Rise P, Barassi MC, Marangoni F, Galli C. Dietary intake of fish vs. formulations leads to higher plasma concentrations of n–3 fatty acids. Lipids 2003;38:415–8.[Medline]
  82. Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the food chain in the United States. Am J Clin Nutr 2000;71:179S–88S.[Abstract/Free Full Text]
  83. Zimmerman M. Burgerstein’s handbook of nutrition: micronutrients in the prevention and therapy of disease. New York: Thieme, 2001.
  84. Simopoulos AP. Essential fatty acids in health and chronic disease. Am J Clin Nutr 1999;70(suppl):560S–9S.[Abstract/Free Full Text]
  85. Carnielli VP, Wattimena DJ, Luijendijk IH, Boerlage A, Degenhart HJ, Sauer PJ. The very low birth weight premature infant is capable of synthesizing arachidonic and docosahexaenoic acids from linoleic and linolenic acids. Pediatr Res 1996;40:169–74.[Medline]
  86. Salem N Jr, Wegher B, Mena P, Uauy R. Arachidonic and docosahexaenoic acids are biosynthesized from their 18-carbon precursors in human infants. Proc Natl Acad Sci U S A 1996;93:49–54.[Abstract/Free Full Text]
  87. Sauerwald TU, Hachey DL, Jensen CL, Heird WC. New insights into the metabolism of long chain polyunsaturated fatty acids during infancy. Eur J Med Res 1997;2:88–92.[Medline]
  88. Uauy R, Mena P, Wegher B, Nieto S, Salem N Jr. Long chain polyunsaturated fatty acid formation in neonates: effect of gestational age and intrauterine growth. Pediatr Res 2000;47:127–35.[Medline]
  89. Dutta-Roy AK. Transport mechanisms for long-chain polyunsaturated fatty acids in the human placenta. Am J Clin Nutr 2000;71(suppl):315S–22S.[Abstract/Free Full Text]
  90. Putnam JC, Carlson SE, DeVoe PW, Barness LA. The effect of variations in dietary fatty acids on the fatty acid composition of erythrocyte phosphatidylcholine and phosphatidylethanolamine in human infants. Am J Clin Nutr 1982;36:106–14.[Abstract/Free Full Text]
  91. Uauy R, Peirano P. Breast is best: human milk is the optimal food for brain development. Am J Clin Nutr 1999;70:433–4.[Free Full Text]
  92. Haggarty P. Effect of placental function on fatty acid requirements during pregnancy. Eur J Clin Nutr 2004;58:1559–70.[Medline]
  93. Crawford M. Placental delivery of arachidonic and docosahexaenoic acids: implications for the lipid nutrition of preterm infants. Am J Clin Nutr 2000;71(suppl):275S–84S.[Abstract/Free Full Text]
  94. Innis SM. Polyunsaturated fatty acids in human milk: an essential role in infant development. Adv Exp Med Biol 2004;554:27–43.[Medline]
  95. Muskiet FA, Fokkema MR, Schaafsma A, Boersma ER, Crawford MA. Is docosahexaenoic acid (DHA) essential? Lessons from DHA status regulation, our ancient diet, epidemiology and randomized controlled trials. J Nutr 2004;134:183–6.[Free Full Text]
  96. Carlson SE. Docosahexaenoic acid and arachidonic acid in infant development. Semin Neonatol 2001;6:437–49.[Medline]
  97. Cunnane SC. Problems with essential fatty acids: time for a new paradigm? Prog Lipid Res 2003;42:544–68.[Medline]
  98. Hornstra G. Essential fatty acids in mothers and their neonates. Am J Clin Nutr 2000;71(suppl):1262S–9S.[Abstract/Free Full Text]
  99. Simopoulos AP, Leaf A, Salem N Jr. Workshop on the essentiality of and recommended dietary intakes for omega-6 and omega-3 fatty acids. J Am Coll Nutr 1999;18:487–9.[Free Full Text]
  100. Koletzko B, Sinclair A. Long-chain polyunsaturated fatty acids in diets for infants: choices for recommending and regulating bodies and for manufacturers of dietary products. Lipids 1999;34:215–20.[Medline]
  101. Greiner RS, Catalan JN, Morigu