The Influence of Head Growth in Fetal Life, Infancy, and Childhood on Intelligence at the Ages of 4 and 8 Years
OBJECTIVE. We investigated the effects of head growth prenatally, during infancy, and during later periods of development on cognitive function at the ages of 4 and 8 years.
METHODS. We studied 633 term-born children from the Avon Longitudinal Study of Parents and Children cohort whose head circumference was measured at birth and at regular intervals thereafter. Their cognitive function was assessed with the Wechsler Preschool and Primary Scale of Intelligence at the age of 4 years and with the Wechsler Intelligence Scale for Children at the age of 8 years. Linear regression analysis was used to calculate postnatal head growth between successive time points, conditional on previous size, and to examine the relationship between head growth during different periods of development and later IQ.
RESULTS. When the influence of head growth was distinguished for different periods, only prenatal growth and growth during infancy were associated with subsequent IQ. At 4 years, after adjustment for parental characteristics, full-scale IQ increased an average of 2.41 points for each 1-SD increase in head circumference at birth and 1.97 points for each 1-SD increase in head growth during infancy, conditional on head size at birth. At 8 years, head circumference at birth was no longer associated with IQ, but head growth during infancy remained a significant predictor, with full-scale IQ increasing an average of 1.56 points for each 1-SD increase in growth.
CONCLUSION. The brain volume a child achieves by the age of 1 year helps determine later intelligence. Growth in brain volume after infancy may not compensate for poorer earlier growth.
Several studies in children have shown that those with larger brains, measured with MRI or as head circumference, tend to score higher on tests of cognitive function.1–4 Similar associations have been found in adults.5,6 Maximal brain volume is usually achieved between the ages of 5 and 10 years,7 but rates of brain growth are highest in the last part of gestation and the first 1 year of life. There is evidence that impaired brain growth in utero8 and in infancy1 may lead to poorer cognitive function in childhood, but less is known about the effect of rates of brain growth after infancy. Findings from studies of very low birth weight infants suggest that the critical period for catch-up growth, in terms of later intelligence, may be confined to the first 1 year of life,9,10 but whether infancy is the most important period of postnatal brain growth for children of normal birth weight is unclear. We investigated the effects of brain growth prenatally, during infancy, and during later periods of postnatal development on cognitive function at the ages of 4 and 8 years among term-born members of the Children in Focus subset of the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort whose head circumference was measured at birth and at regular intervals thereafter.
The ALSPAC is a prospective study of 14541 pregnancies recruited from all pregnancies in 3 Bristol-based district health authorities with expected dates of delivery between April 1991 and December 1992. Full details of the methods have been published.11 The Children in Focus cohort is a 10% random selection from the last 6 months of ALSPAC births and includes 1366 term-born children. The cohort was invited to attend clinics for examination at 4, 8, 12, 18, 25, 31, 37, 43, 49, 61, and 96 months of age. Children had their cognitive function assessed at the ages of 4 and 8 years. Informed consent was obtained from mothers, and ethical approval was obtained from local ethics committees.
Cognitive Function Testing
Cognitive function was assessed at 4 years of age with the Wechsler Preschool and Primary Scale of Intelligence, Revised United Kingdom Edition,12 and at 8 years of age with the Wechsler Intelligence Scale for Children.13
Occipitofrontal circumference was measured soon after birth and during each visit to the clinic, by passing a tape measure around the widest horizontal circumference of the head. The tape measure was kept taut, and measurements were made to the nearest 0.1 cm. Standing heights at the ages of 4 and 8 years were measured in the clinics with a Leicester height measure (Cranlea; Birmingham, United Kingdom).
Gender, gestational age, maternal age, parental education, social class, parenting behavior, duration of breastfeeding, number of older siblings, and presence of postnatal depression were identified as potential confounding factors. Social class was defined by using the occupational coding of the United Kingdom Registrar General. Parenting behavior was assessed with a self-completed questionnaire when each child was 2 years of age. Mothers were asked about the frequency with which they did the following: let the child play with paints, mud, or messy objects, let the child use objects to build towers or other creations, sing to the child, read the child stories, praise the child, kiss or cuddle the child, shout at the child, slap the child, go to a park or playground with the child, have a meal with the child, and let the child make a lot of noise (eg, singing or banging saucepans). The 5 response options were every day (score of 1), several times per week (score of 2), approximately once per week (score of 3), rarely (score of 4), and never (score of 5). All activities except shouting and slapping were recoded subsequently (1 = 5 and 5 = 1), so that parents who reported, for example, that they read to their child every day received a parenting score of 5 for this activity, whereas those who slapped their child every day received a score of 1. A total parenting score was calculated by adding the scores for the 11 activities. The presence of postnatal depression was defined as a score of >12 on the Edinburgh Postnatal Depression Scale when the child was 8 weeks of age.14
To consider the effects of different periods of head growth (prenatal and postnatal) on cognitive function, we used 4 head circumference measurements for each child, namely, at birth, 1 year, 4 years, and 8 years. Head circumference from birth onward was, on average, smaller for girls than for boys. Head circumference at birth and 1 year of age tended to increase with gestational age at birth. We used multivariate linear regression to create gender- and gestation-adjusted head circumference measurements. To facilitate comparisons of the influence of head circumference at different ages, we converted these gender- and gestation-adjusted head circumference measurements at birth, 1 year, 4 years, and 8 years into SD scores (SDSs) or z scores, by subtracting the mean head circumference measurement from each individual's measurement and then dividing the difference by the SD. Postnatal head growth between successive time points, conditional on previous size, was calculated by saving the residuals from linear regression models of head circumference SDS at each successive time point versus head circumference SDS at all earlier time points. For example, head growth between birth and 1 year of age, conditional on head size at birth, was calculated by saving the residuals from a linear regression model of head circumference SDS at 1 year of age versus head circumference SDS at birth. Head growth between 1 and 4 years of age, conditional on growth up to 1 year of age, was calculated by saving the residuals from a linear regression model of head circumference SDS at 4 years of age versus head circumference SDS at birth and head circumference SDS at 1 year of age, and so on. The residuals obtained from these models indicate whether head growth by a particular age is greater or less than would be predicted from previous head circumference measurements; they have the statistical property that they are mutually uncorrelated (r = 0.00; P = 1.00) and so can be included in a model simultaneously, to distinguish the effects of head growth during different periods.
We used t or χ2 tests, as appropriate, to compare the characteristics of the children with and without complete data on head circumference and IQ. Multivariate linear regression, with adjustment for parental and family factors, was used to examine the influence of achieved head circumference and head growth during different periods of development on cognitive function at 4 and 8 years of age.
Of the 1366 term-born children in the cohort, 633 (46.3%) had complete data on head circumference at birth, 1 year, 4 years, and 8 years and IQ at 4 years and 8 years. These children were similar to the 733 children who did not take part in all of the follow-up examinations with respect to gender, birth weight, number of older siblings, and paternal social class, but they were more likely to have been breastfed and to have parents who had A levels (General Certificate of Education advanced level exams, normally taken at age 18 years) or a college degree. Their mothers were slightly older and were more likely to come from higher social classes than were the mothers of the children who had incomplete data (Table 1). The analyses that follow are based on results for the 633 children with complete data.
Head circumference (mean ± SD) increased from 34.9 ± 1.1 cm at birth to 46.6 ± 1.3 cm by 1 year of age. Growth after infancy was slower, with mean head circumference increasing to 50.9 ± 1.4 cm by 4 years and to 53.4 ± 1.4 cm by 8 years. At the age of 4 years, full-scale IQ (mean ± SD) was 106.3 ± 13.7, verbal IQ was 101.7 ± 13.2, and performance IQ was 109.8 ± 14.0. At the age of 8 years, full-scale IQ (mean ± SD) was 105.6 ± 15.8, verbal IQ was 109.0 ± 16.4, and performance IQ was 100.4 ± 16.4. There were no statistically significant differences in IQ between the genders.
As expected, IQ scores were associated strongly with parental characteristics. In univariate analyses, IQ scores were, on average, higher for children whose parents were more educated or from more advantaged social classes, those whose mothers were older or had breastfed them for ≥3 months, and those whose mothers scored higher on the parenting questionnaire. Having older siblings was associated with lower verbal and performance IQ scores. Children whose mothers had been depressed postnatally tended to have slightly lower IQ scores, but these differences were not statistically significant. We adjusted for all of these factors in the multivariate analyses.
Children who had a larger head circumference when their cognitive function was measured tended to have a higher IQ, compared with children whose head growth was poorer. Full-scale IQ at 4 years of age increased an average of 2.38 points (95% confidence interval [CI]: 1.32–3.44 points), verbal IQ increased an average of 1.74 points (95% CI: 0.71–2.77 points), and performance IQ increased an average of 2.36 points (95% CI: 1.22–3.40 points) for each 1-SD increase in current head circumference, after adjustment for gender, gestational age, and parental factors. Similar trends were seen between full-scale and verbal IQ scores at 8 years of age and current head circumference. Full-scale IQ increased an average of 1.27 points (95% CI: 0.08–2.47 points) and verbal IQ increased an average of 1.45 points (95% CI: 0.21–2.69 points) for each 1-SD increase in current head circumference, after multivariate adjustment. The relationship between performance IQ and current head circumference was weaker and not statistically significant after multivariate adjustment. All of these relationships were little affected by adjustment for the child's birth weight or current height, neither of which was associated significantly with IQ independent of head circumference.
We checked whether the association between larger current head circumference and higher IQ was also present for children who had been excluded from full analysis because of missing data (n = 733), by examining the relationship in a subset of 306 children who had information on head circumference and IQ at 4 years of age. For these children, full-scale IQ increased an average of 2.14 points (95% CI: 0.57–3.71 points) per 1-SD increase in head circumference at 4 years of age, an association similar to that seen for the children with complete data.
The serial measurements of head circumference made for the children between birth and 8 years of age were highly correlated. Whether the association between current head circumference and IQ at 4 or 8 years of age reflects the effect of head growth achieved by that age or the influence of growth during an earlier period was therefore unclear. To disentangle which periods of head growth were important for cognitive function, we calculated the extent of growth during 3 periods (ie, birth to 1 year of age, 1–4 years of age, and 4–8 years of age), conditional on size at all previous time points. Each conditional measure represents growth between 2 time points that is greater or less than would be expected on the basis of the child's head size at earlier time points. Because head circumference at birth and the conditional head growth measures are all mutually uncorrelated and therefore independent, they can be used in a regression model simultaneously.
Table 2 shows the relationship between IQ at 4 years of age and the head growth variables considered simultaneously. Full-scale, verbal, and performance IQ scores at 4 years of age were associated positively with head circumference at birth and with head growth during infancy, conditional on head size at birth. These associations were independent of each other and persisted after adjustment for parental factors. Therefore, the highest IQ scores at 4 years of age were seen for children whose head circumference had grown large prenatally and whose head circumference during infancy had grown larger than expected, given its size at birth. Full-scale IQ, for example, increased an average of 2.41 points (95% CI: 1.31–3.50 points) for each 1-SD increase in head circumference at birth and by an average of 1.97 points (95% CI: 0.68–3.26 points) for each 1-SD increase in head growth during infancy. There were similar associations with verbal and performance IQ scores (Table 2). Greater-than-expected head growth between 1 and 4 years of age was not associated with a higher IQ at 4 years of age. The relationships between head growth variables and IQ at 4 years of age were little changed when additional adjustment was made for birth weight (data not shown).
By the age of 8 years, head circumference at birth had ceased to be a statistically significant predictor of IQ but greater-than-expected head growth during infancy was still associated with higher full-scale and verbal IQ scores (Table 3). Full-scale IQ increased an average of 1.56 points (95% CI: 0.11–3.01 points) and verbal IQ increased an average of 1.55 points (95% CI: 0.04–3.06 points) for each 1-SD increase in conditional head growth during infancy, after multivariate adjustment. Performance IQ tended to be higher for children who experienced greater-than-expected head growth during infancy, but this relationship became nonsignificant after adjustment for parental factors. There were no statistically significant associations between head growth after infancy and IQ at 8 years of age. Children whose head circumference had grown more than expected between the ages of 1 and 4 years tended to have higher verbal IQ scores at 8 years of age, but this relationship was of borderline statistical significance.
The finding that IQ scores at 4 and 8 years of age were associated more strongly with head growth up to 1 year of age than with subsequent head growth raises the question of whether the effects of poorer head growth during infancy on later IQ might be overcome by improved growth at a later age. We explored this by performing additional linear regression analyses of IQ scores at 4 and 8 years of age, in which we examined whether there was evidence of an interaction between head circumference at 1 year of age and head growth between 1 and 4 years of age. The interaction between head circumference at 1 year of age and head growth between 1 and 4 years of age was not statistically significant in any of the models (P > .5), which suggests that greater-than-expected growth after 1 year of age is not associated with improvements in subsequent IQ scores if head growth in infancy is poor.
In this population-based study of term-born children, we found that, although IQ scores at 4 and 8 years of age tended to be higher for children who had a larger current head size, the extent of prenatal head growth and, particularly, head growth in infancy seemed to be most important for subsequent intelligence, when we distinguished the effects of growth in different periods. Head growth after infancy was not associated with later IQ scores and did not seem to compensate for poorer growth in the first 1 year of life.
One limitation of this study was that complete data on head circumference measurements and cognitive function were available for only 46.3% of the Children in Focus cohort. There was evidence that the mothers of the children who had complete data tended to be more educated and of higher social class than the mothers of children with incomplete data. However, the exclusion of children who did not take part in all of the follow-up studies at 1, 4, and 8 years of age is unlikely to have biased our results, unless the relationship between head growth and cognitive function was different for the children we excluded. We can be fairly certain that this was not the case because, among the 306 children who had incomplete data but did have information on head circumference and IQ at 4 years of age, the relationship between head circumference and IQ was similar to that found for children with complete data.
Among the strengths of the study was the availability of cognitive function data at 4 and 8 years of age, which provided the opportunity to test the consistency of the associations we found. We were also able to adjust for potentially confounding factors known to influence intelligence, such as parental social class and education, parenting behavior, number of older siblings, maternal age, duration of breastfeeding, and history of postnatal depression. Perhaps the main advantage, however, was that the serial measurements of head size allowed us to examine how cognitive function was influenced by brain growth both prenatally and during different periods of postnatal development. Head circumference has been shown to correlate closely with brain volume in neonates, children, and young adults5,15,16 and has been widely used as a proxy for brain volume in studies of cognitive function.6 Imaging studies of children showed that maximal brain volume is usually achieved between the ages of 5 and 10 years. Increases in myelination and refinement of neuronal connections result in changes in gray and white matter volumes during childhood and adolescence,3,17 but the relationship of these structural brain changes to cognitive development is not yet clear.
Evidence that head circumference at birth can predict cognitive outcomes in childhood has been inconsistent. Some studies that categorized participants according to whether they were microcephalic at birth reported significant differences in intelligence. In a study of 118 low birth weight infants, for example, those whose head circumference was in <10th percentile at birth had a mean IQ at 5 years of age that was 17 points lower than that of those whose head circumference was in >50th percentile.18 Among >248000 young men, the risk of poor performance on a test of intelligence was 28% higher for those whose head circumference at birth had been >2 SDs below the mean, compared with those whose head circumference was within the normal range.8 However, in a study of >14000 children born at term whose head circumference at birth had been in >10th percentile for gestational age, there were no differences in IQ scores at 4 or 7 years of age between children whose head circumference had been small relative to their birth weight (relative microcephaly) and the rest of the study participants.19 A study that examined the relationship between head circumference at birth, as a continuous variable, and IQ among 221 children at 9 years of age found no statistically significant association.1 In the current investigation of 633 children, we found that head circumference at birth was a significant predictor of full-scale, verbal, and performance IQ scores at the age of 4 years but was outweighed in importance as a predictor of IQ at the age of 8 years by postnatal head growth. Variation in brain growth prenatally, although associated with intelligence in early childhood, does not seem to have a lasting influence on cognitive function, at least for children whose head circumference was within the normal range at birth.
Evidence for the importance of postnatal brain growth in cognitive development has come from studies of children between 7 and 9 years of age, showing associations between higher IQ scores and larger current head circumferences.4,20 However, whether these associations reflect the cumulative effect of brain growth achieved by that age or the influence of critical periods of growth earlier in life has been unclear. Our findings, derived by distinguishing the effects of brain growth during different periods of development, suggest that these associations between head circumference and intelligence in later childhood owe more to the influence of brain growth during infancy than they do to growth after infancy; the findings fit with observations in a cohort of 2023 children that those with IQ scores in the superior range (scores of ≥120) at the age of 7 years had larger head circumferences at 1 year of age than did children whose IQ scores were average (scores of 80–119).2 In a recent study, higher IQ scores at the age of 9 years were associated with increased head growth both in the first 9 months of life and between 9 months and 9 years of age, regardless of head size at the start of these periods, but detailed information on growth after 9 months of age was not available.1 Results from some small studies of very low birth weight infants have indicated that the critical period for catch-up brain growth, in terms of later intelligence, may be confined to the first 1 year of life.9,10 Our findings provide additional evidence that infancy is the most important period of postnatal brain growth for determining later intelligence. Brain growth after infancy, at least in terms of brain volume, is unlikely to compensate for poor growth in the first year of life.
Brain growth in early life may be important in determining not only the level of peak cognitive function attained but also whether such function is preserved in old age. Older people with a larger head circumference tend to perform better on tests of cognitive function and may have reduced risks of cognitive decline and of Alzheimer disease.21–23 Twin and family studies have shown that genetic factors play a major role in determining maximal brain volume,24 but there is also evidence that environmental factors influence brain growth. Postnatal head growth tends to be greater for children whose mothers are better educated or who come from higher social classes.1,9 The extent to which this reflects the quality of cognitive stimulation, nutrition, style of parenting, or other influences is unclear. An understanding of which environmental factors influence brain growth, particularly during infancy, may point to ways to maximize an individual's genetic potential for cognitive performance and to reduce his or her risk of cognitive decline in later life.
- Accepted May 23, 2006.
- Address correspondence to Catharine R. Gale, PhD, Medical Research Council Epidemiology Resource Centre, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, United Kingdom. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Gale CR, O'Callaghan FJ, Godfrey KM, Law CM, Martyn CN. Critical periods of brain growth and cognitive function in children. Brain.2004;127 :321– 329
- ↵Reiss AL, Abrams MT, Singer HS, Ross JL, Denckla MB. Brain development, gender and IQ in children. Brain.1996;119 :1763– 1774
- ↵Vernon PA, Wickett JC, Bazana PG, Stelmark RM. The neuropsychology and psychophysiology of human intelligence. In: Sternberg RJ, ed. Handbook of Intelligence. Cambridge, United Kingdom: Cambridge University Press; 2000:245–264
- ↵Lemire RJ, Loeser J, Leech R, Alvord E. Normal and Abnormal Development of the Human Nervous System. Hagerstown, MD: Harper & Row; 1975
- ↵Hack M, Breslau N. Very low birth weight infants: effects of brain growth during infancy on intelligence quotient at 3 years of age. Pediatrics.1986;77 :196– 202
- ↵Wechsler D. Manual for the Wechsler Preschool and Primary Scale of Intelligence-Revised. New York, NY: Psychological Corp; 1989
- ↵Wechsler D. Wechsler Intelligence Scale for Children. San Antonio, TX: Psychological Corp; 1991
- ↵Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression: development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry.1987;150 :782– 786
- ↵Bray OF, Shields WD, Wolcott GJ, Madsen JA. Occipitofrontal head circumference, an accurate measure of intracranial volume. J Pediatr.1969;78 :904– 908
- ↵Gale CR, Walton S, Martyn CN. Foetal and postnatal head growth and risk of cognitive decline in old age. Brain.2003;126 :2273– 2278
- ↵Tisserand DJ, Bosma H, Van Boxtel MPJ, Jolles J. Head size and cognitive ability in nondemented older adults are related. Neurology.2001;56 :969– 971
- Copyright © 2006 by the American Academy of Pediatrics