Abstract
Background: Docosahexanoic acid (DHA) is an important constituent of the brain. Evidence from well-designed intervention trials of the long-term benefits of increasing DHA intake during pregnancy has been sparse.
Objective: We evaluated global cognition, behavior, and attention at age 5 y in the offspring of Mexican women who participated in a randomized controlled trial of prenatal DHA supplementation.
Design: A total of 1094 women were randomly assigned to receive 400 mg of either DHA or placebo/d from 18 to 22 wk of pregnancy until delivery. We assessed cognitive development and behavioral and executive functioning, including attention, in 797 offspring at age 5 y (82% of 973 live births) with the use of the McCarthy Scales of Children’s Abilities (MSCA), the parental scale of the Behavioral Assessment System for Children, Second Edition (BASC-2), and the Conners’ Kiddie Continuous Performance Test (K-CPT). We compared the groups on raw scores, T-scores, and standardized scores, as appropriate. We examined heterogeneity by the quality of the home environment, maternal intelligence, and socioeconomic status.
Results: There were no group differences for MSCA scores (P > 0.05), but the positive effect of the home environment at 12 mo on general cognitive abilities was attenuated in the DHA group compared with in the placebo group (P-interaction < 0.05). There were no differences between groups on the BASC-2. On the K-CPT, offspring in the DHA group showed improved mean ± SD T-scores compared with those of the placebo group for omissions (DHA: 47.6 ± 10.3; placebo: 49.6 ± 11.2; P < 0.01) with no differences (P > 0.05) for the other K-CPT scores or of the proportion who were clinically at risk of attention deficit hyperactivity disorders after Bonferroni correction for multiple comparisons.
Conclusion: Prenatal exposure to DHA may contribute to improved sustained attention in preschool children. This trial was registered at clinicaltrials.gov as NCT00646360.
Keywords: attention, child development, fatty acids, prenatal supplementation, omega-3 fatty acids, preschool cognitive development
INTRODUCTION
Interventions and policies to improve maternal and child nutritional status during the first 1000 d of life are effective means to improve short- and long-term cognitive performance, schooling, and other cognitive-related outcomes (1, 2), thereby contributing to economic progress and development (3–5). Several nutrients, especially trace elements such as iron and zinc and, more recently, n–3 fatty acids are critical for the development of the human brain and neurotransmitter function. In particular, the n–3 fatty acid DHA (22:6n–3) has been identified in animal and human studies as being crucial for the development of the central nervous system. Biological evidence has supported the importance of DHA on global cognition and behavioral and executive functioning, particularly attention, through its role in frontal cortex functions (6–8). In animal models, induced n–3 deficiency has caused the activity and habituation patterns that are consistent with attention-deficit hyperactivity disorder (ADHD)9 (9, 10) In human-infants studies, maternal DHA status at delivery has been inversely correlated with distractibility (11). In school-age children with ADHD, supplementation with DHA improved working memory (12), and higher blood concentrations of DHA and the n–3 fatty acid EPA were associated with better attention and literacy outcomes (13). Attention is an important component of cognitive performance and helps in optimizing learning outcomes; improved attention during the first decade of life has been associated with better school performance in the long term (14).
Infant DHA stores are established early beginning in utero through the first year of life. Therefore, an adequate maternal DHA supply during both pregnancy and lactation is key to establishing adequate offspring stores (15). The main dietary sources of DHA are cold-water fish and seafood. Although the human body can convert the more-common α-linolenic n–3 fatty acid into DHA, dietary sources and enzymatic conversion may not be sufficient to ensure an adequate supply of preformed DHA during the first 1000 d of life, and supplementation may be required (16).
Despite all the biological and empirical evidence that supports the potential benefits of DHA supplementation during pregnancy on the cognitive and behavioral functioning of offspring, to our knowledge, no randomized controlled trials (RCTs) have adequately assessed the impact on offspring cognitive and behavioral functioning beyond infancy. Consequently, the objective of this study was to determine whether prenatal DHA supplementation improves global cognitive function, parent-reported behavioral functioning, and performance on an objective measure of attention at 5 y of age in Mexican children whose mothers participated in an RCT of the effects of prenatal supplementation with DHA on child growth and development in the POSGRAD (Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development) study.
METHODS
We followed up with offspring who were born to women who participated in the POSGRAD study. The detailed methods of the original trial (clinicaltrials.gov; NCT00646360) have been described elsewhere (17). Briefly, between 2005 and 2007, pregnant women who were planning to deliver at the Instituto Mexicano del Seguro Social General Hospital 1 (which provides medical services to an employed, middle-income population) in Cuernavaca, Mexico, were randomly assigned (n = 1094) to receive either an algal DHA supplement (400 mg/d; produced by Martek Biosciences and distributed by Mead Johnson) or a corn- and soy-oil–based placebo from midgestation (18–22 wk) until delivery. Detailed information on birth outcomes was obtained, and all live births (n = 973) were followed prospectively (Figure 1). All study protocols were approved by the Research, Ethics and Biosecurity Commissions of the Instituto Nacional de Salud Pública, Mexico, and by the Institutional Review Board of Emory University. The study procedures were described to subjects, and written informed consent and verbal assent were obtained from the mother (or primary caregiver) and the children, respectively.
FIGURE 1.
Sample flowchart for the assessment of the impact of prenatal supplementation with DHA on cognitive outcomes at 5 y of age.
Measures of cognition
Child cognitive functioning was measured at 5 y of age with the use of the Spanish language version of the McCarthy Scales of Children’s Abilities (MSCA) (18). The MSCA is a comprehensive test battery that measures a variety of cognitive and motor behaviors in children aged 2.5 through 8.5 y and includes 18 subtests that constitute the following 6 scales: Verbal, Perceptual-Performance, Quantitative, Memory, Motor, and General Cognitive. The General Cognitive scale is derived from the Verbal, Perceptual-Performance, and Quantitative scales.
Study psychologists (n = 3) who were trained by the lead psychologist (LS) administered the MSCA to the study children in a quiet and private setting within the hospital; the entire test battery took ∼1 h. The administration of the MSCA was supervised through direct random observations and a review of the completed MSCA forms. The original data were entered on site. Raw scores were computed for each scale and converted into scale indexes with the use of the Scale Index Equivalents of Composite Raw Scores Table, which is standardized for the child’s age at the time of test administration as described in the MSCA manual (19). The Verbal, Perceptual-Performance, Quantitative, Memory, and Motor scale indexes have a mean ± SD score of 50 ± 10. The General Cognitive scale index has a mean ± SD score of 100 ± 15.
Measures of behavior
The parent-report version of the Behavioral Assessment System for Children, Second Edition (BASC-2), was used. This tool was developed as a “multimethod, multidimensional system to evaluate the perceived behavior of children and young adults between 2 and 25 years of age” (20, 21). The tool assesses adaptive and maladaptive behaviors. For this assessment, trained psychologists used the Spanish version of the BASC-2 Parent Rating Scale to interview caregivers about their perception of their children’s behaviors. An initial clinical history was completed to better understand each child’s context and was followed by a list of 134 statements with Likert-scale–type responses (i.e., never, sometimes, often, almost always). The BASC-2 scale provides scores on the following 3 broad domains: externalizing problems (hyperactivity and aggression), internalizing problems (anxiety, depression, and somatization), and adaptive skills; an overall composite of problem behaviors is given by the Behavioral Symptoms Index. Critical items include aggression, anxiety, attention problems, atypicality, conduct problems, depression, hyperactivity, somatization, and withdrawal. Optional scales include anger control, bullying, developmental and social disorders, emotional self-control, executive functioning, negative emotionality, and resiliency. Percentiles were calculated relative to the BASC-2 normative sample, which included 5892 children from different ethnicities and was designed to “resemble the U.S. population with respect to sex, socio-economic status, race/ethnicity, geographic region, and special education classification” (20, 21).
Measures of attention
We measured attention with the use of the Conners’ Kiddie Continuous Performance Test (K-CPT), which is an objective measure of attention that was administered directly to the child at the follow-up visit at age 5 y. The K-CPT was developed as a tool to screen for attention disorders but is also sensitive to small changes of measures of attention along the entire spectrum of the distribution (not only in the extremes) (22–24). The test takes ∼8 min, and children are asked to respond to target stimuli that appear on a screen and to refrain from responding to nontarget stimulus. The interstimuli intervals (ISIs) range between 1500–3000 ms, and the display time lasts 500 ms. The test includes 5 different blocks, each of which is divided into 2 sub-blocks of 20 trials each (one sub-block with 1500-ms ISIs, and the other subblock with 3000-ms ISIs). The information obtained overall includes the number of omissions (nonresponses to targets), commissions (responses to nontargets), Hit response time (“average speed of correct responses for the entire test”), and a hit-response SE (response speed consistency). More-specific indicators are the variability of a SE (“within respondent speed consistency in relation to his own overall standard error”), detectability (“value of the difference between signal and noise distributions”), perseverations (“any reaction that is less than 100 ms is considered a perseveration”), and hit reaction time and SE by ISI and block change (intrastimuli time) (22).
The K-CPT was directly administered to the children with the use of a computer at the study headquarters. We compared the results to the standardization sample from the K-CPT validation study, which included 454 children aged 4 and 5 y who were selected to match the US Census on a number of key demographic variables including sex, region, race or ethnicity, and parental education. Also, 100 children in this sample had previously been diagnosed with ADHD and were included to develop a clinical norm but did not contribute to the standard nonclinical norm (22). The K-CPT software (MHS, Second Edition) generates T-scores and percentiles for each of the indicators relative to this distribution. The K-CPT indexes are presented in raw and T-score formats. The T-score approach allows for the comparison of a performance level to existing validated norms and was, for that reason, used as the outcome of interest. A T-score of 50 represents the mean for the comparison group. These T-scores have an SD of 10 and, and a T-score ≥1 SD is considered to be high and indicates greater risk of an attention problem (22). T-scores were used to assess inattention, impulsivity, and vigilance (22). We also compared individual child profiles to the clinical and nonclinical distributions and computed the probability of being a clinical case of an attention disorder.
Other variables
Maternal characteristics obtained at recruitment included obstetric history, dietary intake, anthropometric measurements, socioeconomic status, and performance on the Raven’s Progressive Matrices test; offspring characteristics obtained at birth included size, sex, and gestational age as described previously (17, 25). Details of feeding practices were obtained during the first year of life, and a trained study psychologist also assessed the quality of the home environment with the use of the Spanish version of the Home Observation for Measurement of the Environment (HOME) (26) during a home visit at 12 mo of age and again at 5 y of age. Similarly, at the follow-up at age 5 y, trained staff also collected details of health histories and child-care practices by interviewing the mothers or primary caregivers and obtained anthropometric measurements for all offspring.
Data analysis
We generated descriptive statistics for all variables previously described and assessed the normality of continuous outcomes. We evaluated if there were any differences by treatment group in the final analytic sample for selected maternal characteristics obtained at the time of enrollment and offspring characteristics at birth and also tested for a potential selection bias by comparing the final analytic sample to those subjects who were lost to follow-up (prenatal and postnatal). We used Student’s t tests to compare group means for continuous variables and chi-square tests to examine differences in proportions for categorical variables. Because there were no differences by treatment group (37) for any examined baseline characteristics or age at follow-up, no further adjustments were made in the primary comparisons. We tested for heterogeneity by selected characteristics that were determined a priori by examining interaction terms between treatment and 1) maternal intelligence scores, 2) socioeconomic status, 3) sex, and 4) the HOME score at 12 and 60 mo of age. Bonferroni correction for multiple testing, in which the resulting P value was multiplied by the number of tests that measured similar constructs, was used whenever there were significant associations and both unadjusted and corrected P values were reported.
All analyses were done with SAS 9.3 software (SAS Institute Inc.), and P < 0.05 was considered statistically significant. The analytic sample had >95% power to detect effect sizes as small as 0.3 SD for intent-to-treat comparisons.
RESULTS
We measured cognitive outcomes, behavioral functioning, and attention in 797 children (396 children in the placebo arm and 401 children in the DHA arm), who represented 82% of the birth cohort (n = 973) (Figure 1). There were no significant differences between intervention and control groups across a range of maternal and child characteristics (Table 1), and there were not any significant differences between children who were studied at age 5 y and those who did not attend the examination (results not shown). The main reasons for a loss to follow-up were migration and a lack of interest in continued participation in the study; these reasons did not differ by treatment group.
TABLE 1.
Selected characteristics of 797 children who were followed up at age 5 y as part of the POSGRAD study by initial treatment assignment1
| Placebo (n = 396) | DHA (n = 401) | P | |
| Maternal characteristics at random assignment | |||
| Age, y | 26.3 ± 4.72 | 26.3 ± 4.9 | 0.94 |
| Gestational age, wk | 20.5 ± 2.1 | 20.5 ± 2.0 | 0.84 |
| Socioeconomic status3 | 0.04 ± 1.0 | 0.02 ± 1.0 | 0.73 |
| Schooling (completed high school or more), % | 60.7 | 55.8 | 0.16 |
| Time in school, y | 12.1 ± 3.6 | 11.8 ± 3.5 | 0.31 |
| Ravens score | 41.1 ± 9.3 | 40.4 ± 9.1 | 0.34 |
| Primigravida, % | 38.7 | 34.9 | 0.27 |
| Weight, kg | 63.6 ± 11.0 | 62.4 ± 11.6 | 0.12 |
| Height, cm | 155.4 ± 5.6 | 154.8 ± 5.7 | 0.10 |
| BMI, kg/m2 | 26.3 ± 4.3 | 26.0 ± 4.3 | 0.29 |
| Offspring characteristics | |||
| Birth weight, g | 3,214 ± 467 | 3,216 ± 451 | 0.95 |
| Birth length, cm | 50.4 ± 2.6 | 50.3 ± 2.3 | 0.77 |
| Head circumference at birth, cm | 34.3 ± 1.8 | 34.4 ± 1.6 | 0.34 |
| Low birth weight (<2500 g), % | 4.6 | 5.5 | 0.54 |
| Gestational age, wk | 39.1 ± 1.7 | 39.0 ± 1.9 | 0.74 |
| Preterm birth,4 % | 8.1 | 10.0 | 0.36 |
| Sex (M), % | 54.3 | 53.6 | 0.85 |
| Intrauterine growth retardation, % | 10.7 | 11.0 | 0.88 |
| Breastfeeding at 3 mo, % | 23.7 | 24.9 | 0.69 |
| Duration of breastfeeding, mo | 9.5 ± 7.6 | 9.4 ± 8.1 | 0.84 |
| HOME score5 | 36.8 ± 4.4 | 36.6 ± 4.4 | 0.59 |
| More than one caregiver at age 5 y, % | 61.5 | 57.5 | 0.28 |
| Height at age 5 y, cm | 108.3 ± 4.4 | 108.3 ± 4.4 | 0.88 |
Sample sizes vary slightly for individual variables because of item-specific missing values. P values were determined with the use of the t test for comparisons of means and the chi-square test for comparisons of proportions. HOME, Home Observation for Measurement of the Environment; POSGRAD, Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development.
Mean ± SD (all such values).
Measured on the basis of a score considering assets and computed with the use of a principal components analysis.
Defined as <37 wk of gestation
Measurement of home environment at 12 mo of age.
Effects of DHA on cognitive functioning
The intent-to-treat analysis revealed no significant differences by treatment group in the mean General Cognitive score or 6 subscales of the MSCA (Table 2). All scores were within normal ranges. However, we showed evidence of effect modification by HOME scores at 12 mo of age for the General Cognitive score on the MSCA (P-interaction < 0.05). The magnitude of the positive association between the quality of home stimulation during infancy and later cognitive functioning was decreased in subjects who were exposed to DHA (β = 0.71; 95% CI: 0.13, 1.29) in utero compared with in the placebo group (β = 1.71; 95% CI: 1.09, 2.33) (Figure 2). Similar relations were shown for the Verbal, Perceptual-Performance, and Memory domains (P < 0.10). There was no significant heterogeneity by sex, maternal intelligence scores, socioeconomic status, or HOME scores at 60 mo of age.
TABLE 2.
Comparison of measures of cognitive development with the use of McCarthy Scales of Children’s Abilities at 5 y of age in children born to women who participated in a trial of 400 mg DHA/d during pregnancy by intervention group1
| Variable | Placebo (n = 396) | DHA (n = 401) | P |
| Raw McCarthy score | |||
| Verbal | 50.9 ± 11.5 | 50.6 ± 11.1 | 0.67 |
| Perceptual-performance | 46.7 ± 9.3 | 46.9 ± 9.3 | 0.73 |
| Quantitative | 20.1 ± 6.4 | 19.9 ± 6.3 | 0.60 |
| Memory | 25.3 ± 7.6 | 25.3 ± 7.6 | 0.98 |
| Motor | 37.8 ± 7.0 | 38.1 ± 6.8 | 0.53 |
| General cognitive | 117.9 ± 23.3 | 117.7 ± 22.0 | 0.94 |
| Scale index McCarthy score | |||
| Verbal | 43.5 ± 8.7 | 43.3 ± 8.5 | 0.69 |
| Perceptual-performance | 48.7 ± 8.3 | 49.0 ± 8.3 | 0.60 |
| Quantitative | 45.3 ± 9.7 | 45.1 ± 9.6 | 0.70 |
| Memory | 42.1 ± 8.7 | 42.2 ± 9.0 | 0.92 |
| Motor | 47.9 ± 8.8 | 48.3 ± 8.9 | 0.54 |
| General cognitive | 92.9 ± 13.3 | 92.4 ± 13.4 | 0.64 |
All values are means ± SDs. P values were determined with the use of the t test for comparison of means.
FIGURE 2.
Relation between HOME score and Composite McCarthy Composite Raw Score at 5 y of age by intervention group. Lighter lines represent 95% CIs. Placebo: n = 396; DHA: n = 401. HOME, Home Observation for Measurement of the Environment.
Effects of DHA on behavior
There were no significant differences between DHA and placebo groups on any aspects of child behavioral functioning (either on the composite or individual scales of the BASC-2) as reported by the caregivers (Table 3). There was also no evidence of heterogeneity by sex, maternal intelligence scores, socioeconomic status, or HOME scores at 12 or 60 mo of age.
TABLE 3.
Behaviors reported by parents with the use of BASC-2 of a sample of Mexican children whose mothers participated in POSGRAD results by treatment group1
| Placebo (n = 396) | DHA (n = 401) | |
| Externalizing problems composite, % | 53.9 ± 27.4 | 51.4 ± 29.2 |
| Hyperactivity | 59.5 ± 25.5 | 57.9 ± 27.8 |
| Aggression | 48.7 ± 28.3 | 45.8 ± 30.0 |
| Internalizing problems composite, % | 63.0 ± 26.7 | 62.5 ± 26.6 |
| Anxiety | 74.3 ± 22.0 | 72.5 ± 22.5 |
| Depression | 51.2 ± 29.8 | 51.1 ± 31.0 |
| Somatization | 52.3 ± 28.9 | 54.4 ± 27.9 |
| Adaptive skills composite, % | 52.5 ± 27.8 | 54.3 ± 28.8 |
| Withdrawal | 44.8 ± 29.8 | 43.6 ± 30.5 |
| Adaptability | 54.8 ± 28.9 | 55.7 ± 29.1 |
| Activities in daily life | 58.3 ± 29.5 | 60.8 ± 29.7 |
| Atypicality | 54.4 ± 29.7 | 53.4 ± 30.1 |
| Attention problems | 48.0 ± 30.1 | 51.2 ± 30.8 |
| Functional communication | 42.6 ± 23.1 | 42.1 ± 24.3 |
| Social skills | 51.1 ± 29.5 | 53.3 ± 29.6 |
| Behavioral Symptoms Index, % | 51.5 ± 28.4 | 51.0 ± 29.8 |
| Anger control | 51.4 ± 27.5 | 49.6 ± 28.9 |
| Bullying | 53.9 ± 26.7 | 52.0 ± 29.5 |
| Developmental and social disorders | 53.6 ± 27.9 | 53.9 ± 27.9 |
| Emotional self-control | 56.0 ± 29.7 | 53.5 ± 31.9 |
| Executive functioning | 55.7 ± 29.2 | 54.8 ± 30.9 |
| Negative emotionality | 54.7 ± 30.2 | 52.0 ± 31.3 |
| Resiliency | 51.8 ± 28.6 | 52.1 ± 29.8 |
All values are means ± SDs. Percentiles were determined relative to the standard population. None of the comparisons were significant. BASC-2, Behavioral Assessment System for Children, Second Edition; POSGRAD, Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development.
Effects of DHA on attention
Children in the DHA group had significantly fewer omissions on the K-CPT than did children in the control group with no differences between groups for commissions or Hit response times (Table 4). The Hit response time Block Change and Hit SE by Block Change, which are the 2 components of vigilance, were also significantly better in the DHA group, but there were no significant differences between groups on any of the impulsivity components. However, these associations were attenuated after Bonferroni correction for multiple comparisons. Compared with controls, a higher proportion of children in the DHA group had a T-score <40 for omissions, and fewer children had a T-score >60. These differences in distribution remained significant after Bonferroni correction for multiple testing (Table 5). There were no significant differences in the response style or proportion of children who were classified at clinical risk of suffering from a disorder (>70th percentile) such as ADHD (22). There was no heterogeneity by sex, maternal intelligence scores, socioeconomic status, or HOME scores at 12 and 60 mo of age.
TABLE 4.
K-CPT main-component scores by intervention group in Mexican children whose mothers participated in the POSGRAD study1
| Placebo (n = 396) | DHA (n = 401) | P | P after Bonferroni correction | |
| Overall | ||||
| Omissions | 49.5 ± 11.2 | 47.6 ± 10.3* | 0.01 | 0.11 |
| Commissions | 51.1 ± 10.2 | 51.2 ± 11.0 | 0.86 | — |
| Hit response time | 56.1 ± 10.6 | 55.5 ± 10.5 | 0.36 | — |
| Inattention | ||||
| Omissions2 | 49.5 ± 11.2 | 47.6 ± 10.3* | 0.01 | 0.11 |
| Commissions2 | 51.1 ± 10.2 | 51.2 ± 11.0 | 0.86 | — |
| Hit response time2 | 56.1 ± 10.6 | 55.5 ± 10.5 | 0.36 | — |
| Hit response time, SE | 51.7 ± 9.6 | 51.2 ± 9.5 | 0.42 | — |
| Variability | 50.7 ± 8.8 | 50.2 ± 8.3 | 0.40 | — |
| Detectability | 50.1 ± 10.7 | 49.9 ± 10.3 | 0.72 | — |
| Hit response time ISI change | 49.9 ± 13.4 | 50.6 ± 12.7 | 0.47 | — |
| Hit SE ISI change | 49.2 ± 9.8 | 50.1 ± 10.0 | 0.90 | — |
| Impulsivity | ||||
| Commissions3 | 51.1 ± 10.2 | 51.2 ± 11.0 | 0.86 | — |
| Hit response time3 | 56.1 ± 10.6 | 55.5 ± 10.5 | 0.36 | — |
| Perseverations | 44.8 ± 7.1 | 45.6 ± 9.1 | 0.13 | — |
| Vigilance | ||||
| Hit response time block change | 52.0 ± 12.5 | 50.1 ± 12.2* | 0.04 | 0.44 |
| Hit SE by block change | 49.9 ± 9.4 | 48.2 ± 10.0* | 0.01 | 0.11 |
All values are means ± SDs. T-scores were determined relative to the standard population (higher values represent poorer performance). For Bonferroni correction, P values were multiplied by the 11 different outcomes measured in the test. Adjusted P values are only reported for significant (P < 0.05) results. *P < 0.05 for difference between treatment groups before Bonferroni correction for multiple testing. ISI, interstimuli interval; K-CPT, Kiddie Continuous Performance Test; POSGRAD, Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development.
Same variable as previously reported in the overall category. The variable was repeated to show it is as a component of inattention.
Same variable as previously reported in the overall category. The variable was repeated to show it is as a component of impulsivity.
TABLE 5.
Distribution of K-CPT scores relative to the standard sample by intervention group in POSGRAD1
| Placebo (n = 396) | DHA (n = 401) | P | P after Bonferroni correction | |
| Omissions score, % of children | ||||
| <40 | 14.4 | 25.7 | <0.0001 | <0.0001 |
| 40–60 | 69.4 | 64.6 | — | — |
| >60 | 16.2 | 9.7 | — | — |
| Commissions score, % of children | ||||
| <40 | 12.1 | 16.2 | 0.08 | — |
| 40–60 | 67.2 | 59.9 | — | — |
| >60 | 20.7 | 23.9 | — | — |
| Hit response time score, % of children | ||||
| <40 | 3.3 | 6.7 | 0.08 | — |
| 40–60 | 64.9 | 61.9 | — | — |
| >60 | 31.8 | 31.4 | — | — |
| Response style (β) score, % of children | ||||
| <40 | 9.6 | 12.5 | 0.35 | — |
| 40–60 | 68.4 | 68.1 | — | — |
| >60 | 22.0 | 19.5 | — | — |
| Overall score, % of children | ||||
| >70 | 8.1 | 7.2 | 0.62 | — |
Values were determined for children of the POSGRAD group within the specified T-score relative to the standard population. For Bonferroni correction, P values were multiplied by the 11 different outcomes measured in the test. Adjusted P values are only reported for significant (P < 0.05) results. K-CPT, Kiddie Continuous Performance Test; POSGRAD, Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development.
DISCUSSION
Prenatal supplementation with DHA had no overall effect on global measures of cognitive functioning at age 5 y, but the effect of the quality of the home environment during infancy on later cognition was significantly reduced in offspring who were exposed in utero to DHA than in the control group who received a placebo. These findings are similar and consistent with our earlier findings in this cohort whereby we showed evidence of selective effects on measures of motor development at 18 mo of age (28). This interaction can be interpreted 2 ways. On one side, it is possible that DHA supplementation was a substitute of early stimulation for children with less-stimulating home environments; if this was the case, DHA could be particularly important for cognitive functioning for kids with less-stimulating environments. In contrast, it is less likely but also possible that prenatal DHA decreases the effect of early stimulation on development in which case better targeted interventions would be necessary to prevent this unintended consequence.
More importantly, we present evidence of a beneficial effect of prenatal supplementation with DHA on objective measures of attention and executive function at preschool age. Children whose mothers received DHA during pregnancy committed fewer omissions than those in the placebo group. These results are consistent with evidence from animal models and biological evidence linking DHA to functions of the prefrontal cortex such as responsiveness and sustained attention. Similar clinical trials that have assessed the effect of DHA supplementation on measures of attention during childhood have yielded mixed results. The DHA for Maternal and Infant Outcomes trial, which provided 800 mg DHA and 100 mg EPA/d during the last 2 trimesters of pregnancy to Australian women showed no effect of the intervention on offspring attention, working memory, and inhibitory control at 2 y of age. The only difference reported by the authors was that children in the intervention group, compared with the controls, were able to look away from target toys fewer times when distracted by other toys (29). However, Jensen et al. (30) reported improvements in the attention of offspring at age 5 y in an RCT of postnatal supplementation in which lactating women received 200 mg of either DHA or placebo/d. Consistent with our findings, the only effect of the intervention was on the Sustained Attention Scale of the Leiter International Performance Scale.
We showed no group differences of any parent-reported behaviors on the BASC-2, including attention problems, in contrast with the objective results obtained from the K-CPT. The more-subjective nature of the BASC-2 could have accounted for these differences. Also, our results for the K-CPT show no differences in the clinical range (in the number of children with clinical problems) but only significant differences within the normal population. Hence, it is possible that the BASC-2, which is designed to detect attention problems, missed these more subtle differences.
This trial has important characteristics that distinguish it from previous studies. First, supplementation was done exclusively with DHA that was obtained from an algal source that was provided only during pregnancy as opposed to a combination of n–3 fatty acids. Second, the study population was from an urban area in Mexico, which is a setting that is representative of many low- to middle-income countries that have experienced the nutrition transition with very low intakes of n–3 fatty acids, especially of preformed DHA, and high intakes of n–6 fatty acids primarily from vegetable oils (31). Third, and most importantly, we used the MSCA and K-CPT, which provided objective measures of cognitive functioning and attention, in combination with the parent-reported BASC-2, which facilitated a comprehensive assessment of the effect of prenatal supplementation with DHA on cognitive and behavioral functioning during the preschool years; moreover, all of these measures were previously used in Mexican children (32). Other strengths of the study include the low attrition rate through age 5 y, the large sample size, the availability of data on potential determinants of cognitive and behavioral outcomes including the quality of the home environment during early life and caregiver characteristics, and that the cohort remained balanced for several maternal and child characteristics.
Potential limitations of the study include the lack of measures of DHA status and details of the quality of the child’s learning environment, especially of the school quality at the time of follow-up. Other factors that potentially affected sustained attention at 5 y of age included maternal first- or second-hand smoking and alcohol consumption during pregnancy (33), exposure to environmental pollutants prenatally and across the life span (34, 35), sleep patterns (36), and the presence of an illness or distress at the time that the test was administered. Children were not tested if they were unwell in the current study, and none of the women smoked during pregnancy. We do not have data for these other variables. However, the randomized design allowed us to expect that both observed and unobserved characteristics were balanced across treatment groups, and although we could not confirm it, we expect that random assignment minimized potential biases that are related to these other factors. Similarly, we did not observe systematic differences in baseline characteristics by attrition status.
An important consideration is the clinical significance of the observed differences in measures of attention shown in this study (fewer omissions). Children in the study were generally healthy, and the prevalence of clinical cases was low (7.6%; n = 61). Fewer than 10% of children in the DHA group had a T-score that was >60 for omissions (>1 SD from the mean), which could be interpreted as a “high mean” or even “elevated” risk of developing a clinical condition (22), compared with >16% of controls. Conversely, >25% of the children in the DHA group had a T-score <40 [which is interpreted as having “lower than average” risk (22)] compared with 14% of controls. Razza et al. (37) showed a longitudinal association between attentional performance at 5 y of age (measured with the use of the Leiter International Performance Scale) and academic achievement at 9 y of age in a sample of >2500 low-income children from 20 cities across the United States. Follow-up of the children in the current study will be important to assess if differences measured by the K-CPT at 5 y of age translate into better academic performance or other functional outcomes during the school years.
In conclusion, our results show the potential of DHA supplementation in the second half of pregnancy to improve the sustained attention of offspring beyond infancy. No differences were shown regarding parent-reported measures of attention problems, which are less-objective measures. The current study confirms the observed differences between treatment groups in the interaction between prenatal DHA and the quality of the home environment in early life on cognitive abilities seen previously in our cohort (27), suggesting that these effects may persist through early childhood. There is evidence that the benefits of improving nutrition during the first 1000 d of life on cognitive outcomes may not become evident until children start school and are more challenged by the environment (38, 39). The long-term significance of our current findings and impact of the POSGRAD intervention on cognitive and behavioral functioning during the school years remains to be determined.
Acknowledgments
We thank the dedicated efforts of Raquel Garcia Feregrino, who has been responsible for the field coordination and supervision of the study since its inception.
The authors’ responsibilities were as follows—UR: had primary responsibility for the final content of the manuscript; UR and IG-C: wrote the manuscript; UR, LS, AD, LMN, JAR, ADS, and RM: participated in the design of the study; UR, LS, JAR, and ADS: conducted the research and participated in the data analysis; IG-C, ADQ, BCP, and WH: analyzed the data and contributed to the interpretation of the results; and all authors: contributed to the interpretation and discussion of the results and read and approved the final version of the manuscript. None of the authors reported a conflict of interest related to the study.
Footnotes
Abbreviations used: ADHD, attention-deficit hyperactivity disorder; BASC-2, Behavioral Assessment System for Children, Second Edition; HOME, Home Observation for Measurement of the Environment; ISI, interstimuli interval; K-CPT, Kiddie Continuous Performance Test; POSGRAD, Prenatal Omega-3 Fatty Acid Supplementation and Child Growth and Development; RCT, randomized controlled trial.
REFERENCES
- 1.Glewwe P. The impact of child health and nutrition on education in developing countries: theory, econometric issues, and recent empirical evidence. Food Nutr Bull 2005;26(2 Suppl 2):S235–50. [DOI] [PubMed] [Google Scholar]
- 2.Jukes M. The long-term impact of preschool health and nutrition on education. Food Nutr Bull 2005;26(2 Suppl 2):S193–201. [DOI] [PubMed] [Google Scholar]
- 3.Stenberg K, Axelson H, Sheehan P, Anderson I, Gulmezoglu AM, Temmerman M, Mason E, Friedman HS, Bhutta ZA, Lawn JE, et al. Advancing social and economic development by investing in women’s and children’s health: a new Global Investment Framework. Lancet 2014;383:1333–54. [DOI] [PubMed] [Google Scholar]
- 4.Martorell R, Melgar P, Maluccio JA, Stein AD, Rivera JA. The nutrition intervention improved adult human capital and economic productivity. J Nutr 2010;140:411–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hoddinott J, Maluccio JA, Behrman JR, Flores R, Martorell R. Effect of a nutrition intervention during early childhood on economic productivity in Guatemalan adults. Lancet 2008;371:411–6. [DOI] [PubMed] [Google Scholar]
- 6.Campoy C, Escolano-Margarit MV, Anjos T, Szajewska H, Uauy R. Omega 3 fatty acids on child growth, visual acuity and neurodevelopment. Br J Nutr 2012;107:S85–106. [DOI] [PubMed] [Google Scholar]
- 7.Huffman SL, Harika RK, Eilander A, Osendarp SJ. Essential fats: how do they affect growth and development of infants and young children in developing countries? A literature review. Matern Child Nutr 2011;7(Suppl 3):44–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Martinez M. Tissue levels of polyunsaturated fatty acids during early human development. J Pediatr 1992;120:S129–38. [DOI] [PubMed] [Google Scholar]
- 9.Vancassel S, Blondeau C, Lallemand S, Cador M, Linard A, Lavialle M, Dellu-Hagedorn F. Hyperactivity in the rat is associated with spontaneous low level of n-3 polyunsaturated fatty acids in the frontal cortex. Behav Brain Res 2007;180:119–26. [DOI] [PubMed] [Google Scholar]
- 10.Levant B, Zarcone TJ, Fowler SC. Developmental effects of dietary n-3 fatty acids on activity and response to novelty. Physiol Behav 2010;101:176–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Colombo J, Kannass KN, Shaddy DJ, Kundurthi S, Maikranz JM, Anderson CJ, Blaga OM, Carlson SE. Maternal DHA and the development of attention in infancy and toddlerhood. Child Dev 2004;75:1254–67. [DOI] [PubMed] [Google Scholar]
- 12.Widenhorn-Müller K, Schwanda S, Scholz E, Spitzer M, Bode H. Effect of supplementation with long-chain omega-3 polyunsaturated fatty acids on behavior and cognition in children with attention deficit/hyperactivity disorder (ADHD): a randomized placebo-controlled intervention trial. Prostaglandins Leukot Essent Fatty Acids 2014;91:49–60. [DOI] [PubMed] [Google Scholar]
- 13.Milte CM, Parletta N, Buckley JD, Coates AM, Young RM, Howe PR. Increased erythrocyte eicosapentaenoic acid and docosahexaenoic acid are associated with improved attention and behavior in children with ADHD in a randomized controlled three-way crossover trial. J Atten Disord 2015;19:954–64. [DOI] [PubMed] [Google Scholar]
- 14.Washbrook E, Propper C, Sayal K. Pre-school hyperactivity/attention problems and educational outcomes in adolescence: prospective longitudinal study. Br J Psychiatry 2013;203:265–71. [DOI] [PubMed] [Google Scholar]
- 15.Brenna JT. Animal studies of the functional consequences of suboptimal polyunsaturated fatty acid status during pregnancy, lactation and early post-natal life. Matern Child Nutr 2011;7(Suppl 2):59–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Brenna JT, Carlson SE. Docosahexaenoic acid and human brain development: evidence that a dietary supply is needed for optimal development. J Hum Evol 2014;;77:99–106. [DOI] [PubMed] [Google Scholar]
- 17.Ramakrishnan U, Stein AD, Parra-Cabrera S, Wang M, Imhoff-Kunsch B, Juarez-Marquez S, Rivera J, Martorell R. Effects of docosahexaenoic acid supplementation during pregnancy on gestational age and size at birth: randomized, double-blind, placebo-controlled trial in Mexico. Food Nutr Bull 2010;31(2 Suppl):S108–16. [DOI] [PubMed] [Google Scholar]
- 18.Schnaas L, Rothenberg SJ, Perroni E, Martínez S, Hernández C, Hernández RM.. Temporal pattern in the effect of postnatal blood lead level on intellectual development of young children. Neurotoxicol Teratol 2000;22:805–10. [DOI] [PubMed] [Google Scholar]
- 19.McCarthy D. Manual for the McCarthy Scales of Children’s Abilities. New York: The Psychological Corporation; 1972. [Google Scholar]
- 20.Reynolds CR, Kamphaus RW. BASC: Behavior Assessment System for Children: manual. Circle Pines (MN): American Guidance Service; 1998. [Google Scholar]
- 21.Reynolds CR, Kamphaus RW. The clinician’s guide to the behavior assessment system for children. New York: The Guilford Press; 2002. [Google Scholar]
- 22.Conners C. Conners’Continuous Performance Test II: Computer program for Windows technical guide and software manual. Tonawanda(NY): Multi-Health Systems Inc; 2000. [Google Scholar]
- 23.Pollak Y, Shomaly HB, Weiss PL, Rizzo AA, Gross-Tsur V. Methylphenidate effect in children with ADHD can be measured by an ecologically valid continuous performance test embedded in virtual reality. CNS Spectr 2010;15:125–30. [DOI] [PubMed] [Google Scholar]
- 24.Pollak Y, Weiss PL, Rizzo AA, Weizer M, Shriki L, Shalev RS, Gross-Tsur V. The utility of a continuous performance test embedded in virtual reality in measuring ADHD-related deficits. Journal of developmental and behavioral pediatrics. J Dev Behav Pediatr 2009;30:2–6. [DOI] [PubMed] [Google Scholar]
- 25.Stein AD, Wang M, Rivera JA, Martorell R, Ramakrishnan U. Auditory- and visual-evoked potentials in Mexican infants are not affected by maternal supplementation with 400 mg/d docosahexaenoic acid in the second half of pregnancy. J Nutr 2012;142:1577–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Caldwell BM, Bradley RH. Home Observation for Measurement of the Environment: administration manual. Tempe (AZ): Family and Human Dynamics Research Institute, Arizona State University; 2003. [Google Scholar]
- 27.Gonzalez-Casanova I, Stein AD, Hao W, Garcia-Feregrino R, Barraza-Villarreal A, Romieu I, Rivera JA, Martorell R, Ramakrishnan U. Prenatal supplementation with docosahexaenoic acid has no effect on growth through 60 months of age. J Nutr 2015;145:1330–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ramakrishnan U, Stinger A, DiGirolamo AM, Martorell R, Neufeld LM, Rivera JA, Schnaas L, Stein AD, Wang M. Prenatal docosahexaenoic acid supplementation and offspring development at 18 months: randomized controlled trial. PLoS One 2015;10:e0120065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gould JF, Makrides M, Colombo J, Smithers LG. Randomized controlled trial of maternal omega-3 long-chain PUFA supplementation during pregnancy and early childhood development of attention, working memory, and inhibitory control. Am J Clin Nutr 2014;99:851–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Jensen CL, Voigt RG, Llorente AM, Peters SU, Prager TC, Zou YL, Rozelle JC, Turcich MR, Fraley JK, Anderson RE, et al. Effects of early maternal docosahexaenoic acid intake on neuropsychological status and visual acuity at five years of age of breast-fed term infants. J Pediatr 2010;157:900–5. [DOI] [PubMed] [Google Scholar]
- 31.Parra-Cabrera S, Stein AD, Wang M, Martorell R, Rivera J, Ramakrishnan U. Dietary intakes of polyunsaturated fatty acids among pregnant Mexican women. Matern Child Nutr 2011;7:140–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Fortenberry GZ, Meeker JD, Sanchez BN, Bellinger D, Peterson K, Schnaas L, Solano-Gonzalez M, Ettinger AS, Hernandez-Avila M, Hu H, et al. Paraoxonase I polymorphisms and attention/hyperactivity in school-age children from Mexico City, Mexico. Environ Res 2014;132:342–9. [DOI] [PubMed] [Google Scholar]
- 33.Han JY, Kwon HJ, Ha M, Paik KC, Lim MH, Gyu Lee S, Yoo SJ, Kim EJ. The effects of prenatal exposure to alcohol and environmental tobacco smoke on risk for ADHD: a large population-based study. Psychiatry Res 2015;225:164–8. [DOI] [PubMed] [Google Scholar]
- 34.Lee DH, Jacobs DR, Porta M. Association of serum concentrations of persistent organic pollutants with the prevalence of learning disability and attention deficit disorder. J Epidemiol Community Health 2007;61:591–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Perera FP, Chang HW, Tang D, Roen EL, Herbstman J, Margolis A, Huang TJ, Miller RL, Wang S, Rauh V. Early-life exposure to polycyclic aromatic hydrocarbons and ADHD behavior problems. PLoS One 2014;9:e111670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Lufi D. The relationship between sleep patterns and attention levels. Neurocase 2014;20:591–8. [DOI] [PubMed] [Google Scholar]
- 37.Razza RA, Martin A, Brooks-Gunn J. The implications of early attentional regulation for school success among low-income children. J Appl Dev Psychol 2012;33:311–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Freeman HE, Klein RE, Townsend JW, Lechtig A. Nutrition and cognitive development among rural Guatemalan children. Am J Public Health 1980;70:1277–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Freeman HE, Klein RE, Kagan J, Yarbrough C. Relations between nutrition and cognition in rural Guatemala. Am J Public Health 1977;67:233–9. [DOI] [PMC free article] [PubMed] [Google Scholar]