Abstract

Background Birthweight is associated with cognition and educational attainment across the full birthweight range in the normal population, independently of social background. However, the extent to which birthweight reflects fetal growth, or is a marker of subsequent size, with respect to this association, is not clear. We therefore investigated the independent effects of birthweight and postnatal height adjusted for postnatal weight on cognitive function and educational attainment while controlling for family background.

Methods Using the British 1946 birth cohort we investigated the association between cognitive function at various ages and birthweight, height adjusted for weight in childhood and adulthood, and educational attainment, controlling for sex, father's social class, maternal education, birth order, and maternal age.

Results Birthweight was positively associated with cognition up to age 26, and with the likelihood of obtaining advanced educational qualifications. Height was positively associated with cognition at all ages, and also with educational attainment. Weight was not associated with cognition at ages 8 and 15, but was negatively associated with verbal ability at age 26, with verbal memory at age 43, and with educational attainment. These effects were independent of each other, and of family background. Conditional analyses suggested the positive effect of height growth on cognition at two intervals, one in early childhood, and the other in late adolescence. In addition, weight gain after age 15 was negatively associated with cognition at 26.

Conclusion Birthweight and postnatal growth are independently associated with cognition.

While the adverse effects of clinically low birthweight on cognitive development are well known,1 birthweight is also associated with cognition across the full birthweight range in the normal population,2 independently of social background.3–8 One possible explanation is that insulin-like growth factors (IGF) target brain areas responsible for cognition while playing a critical role in determining body growth.9 An important question, with implications for intervention,10 is whether the critical period for such development occurs only before birth, as implied by the fetal origins hypothesis,11 or continues during postnatal growth. The extent to which birthweight reflects fetal growth, or is a marker of subsequent size, is clearly of relevance to this question. If later size attenuates the effect of birthweight on a particular outcome, then later size may be more relevant to that outcome than birthweight.10

Using the British 1946 birth cohort, we previously reported a positive association between birthweight and cognition in childhood, adolescence and early adulthood, and between birthweight and educational attainment, adjusting for sex, father's social class, maternal education, birth order, and maternal age.6 We therefore extended this study to include measures of later size by adding height measurements at or near to the time of these cognitive assessments, adjusted for concurrent weight. Thus we were able to investigate the independent effects of birthweight, postnatal height and postnatal weight on cognitive function and educational attainment while controlling for family background.

Methods

Participants

Participants comprised the study population of the MRC National Survey of Health and Development (NSHD), also known as the British 1946 birth cohort, and initially consisting of 5362 children of non-manual and agricultural workers and a random sample of one in four of manual workers selected from all single and legitimate births that occurred in England, Scotland and Wales during one week in March 1946.12 The cohort has been studied on 21 occasions between birth and age 53 years, when information about sociodemographic factors and medical, cognitive and psychological function was obtained by interview and examination. This occurred most recently in 1999 at age 53 when sample size was 3035. In 1989 the cohort was shown to be a representative sample, in most respects, of the UK population legitimately and singly born in the immediate post-war era.13 Exceptions were an over-representation among non-responders of the never married, the least literate, those always in manual social class circumstances, and the mentally ill.13 By age 43 years permanent losses comprised 365 (6.8%) deaths, and 540 (10.1%) refusals, and temporary losses were 607 (11.3%) emigrations or residence overseas and 370 (6.9%) failures to contact. Losses through death have been greater among those in manual rather than non-manual social classes.14

Cognitive measures

The following cognitive measures were used in this study:

Age 815

Reading comprehension; word pronunciation; vocabulary; non-verbal reasoning.

Age 1516

Verbal and non-verbal intelligence (AH4); reading comprehension (Watts-Vernon); mathematics.

Age 26

Reading comprehension (Watts-Vernon, with an additional 10 items of increased difficulty).

Age 43

Verbal memory (word list learning), assessed by a 15-item word learning task devised by the NSHD. Total number of words correctly recalled over three trials was summed to provide an overall score for this test.

All test scores were standardized to give a mean of 0 and a standard deviation of 1, using the total population with a valid score for each test. Global scores representing overall cognitive function at ages 8, 11, 15 and 26 were obtained by setting the standard deviation of the distribution of the sum of these scores to one.

Educational attainment

The highest educational or training qualification achieved by age 26 years was dichotomized into ordinary, i.e. qualifications typically obtained by the end of the compulsory schooling years at age 16 (up to ‘O’ Levels and their training equivalents), versus advanced, i.e. ‘A’ Levels or entry to university examinations, and their equivalents, or degree level or equivalent.

Birthweight, postnatal height and postnatal weight

Information on birthweight was obtained from clinical records. As in our previous analysis,6 this measure was divided into five categories (<2.51 kg, 2.51–3.00 kg, 3.01–3.50 kg, 3.51–4.00 kg and 4.01–5.00 kg). Height and weight measurements taken at the same ages as the cognitive tests (at age 7 corresponding to the cognitive assessment at age 8) were also divided into five categories, the divisions of which allowed sufficient numbers within each category at each age. In addition, height at ages 2 and 4 years were obtained. All heights and weights were measured by doctors or nurses, with the exception of those at age 26, which were self-reported. Measured height at age 36 was therefore used to represent adult height, corresponding to the cognitive tests obtained at ages 26 and 43 years.

Statistical methods

Associations between birthweight, height adjusted for weight and cognitive function at ages 8, 15, 26 and 43 years were investigated using regression models. Distributions of all outcome measures, except the score at 26 years, were symmetric. For the score at age 26 years, a model using a square transformation of the outcome to normalize the distribution reached the same conclusions as using the raw score. Thus for ease of interpretation the model using the raw score is presented here. Birthweight was fitted as a categorical variable. Since associations between several of the height, weight and cognition variables were not linear, heights and weights were fitted as categorical variables. Unadjusted estimates of differences in means with the middle group as the baseline group were initially obtained. Models were then fitted including all three body size variables, and identical confounders to those used in our previous study.6 These confounders were sex, father's social class (categorized into six groups and fitted in the model using five dummy variables), maternal education (categorized into those with primary or secondary education only with no formal qualifications, and those with formal qualifications or any further education), birth order (included as a continuous variable) and maternal age (included as a continuous variable). To test whether change in height or weight was associated with cognitive score, each of the above models were further adjusted for height and weight at the previous age. The model for cognitive function at age 8 years was adjusted for height and weight at age 4 years. The fit of each model was investigated by calculating the standardized residuals and using standard model diagnostic plots.

Interactions between birthweight and postnatal height and birthweight and postnatal weight were assessed for each age. A significant interaction would suggest that the benefit of postnatal growth on cognition is dependent on birthweight, i.e. early body size is acting to modify the cognitive effect of later body size,10 or possibly that the benefit of large size at birth depends on postnatal size or growth.

The effect of adult height and weight at 26 years on education was investigated using logistic regression. Adjustment for birthweight and the confounding variables was carried out with variables defined as for the cognitive score models. The Hosmer-Lemeshow goodness-of-fit test was carried out.

Results

Missing data

As previously reported,6 762 cohort members had not undergone any cognitive testing (ages 8–43). Of the 4600 cohort members with at least one recorded cognitive test score, 3900 (85%) had complete birthweight and confounder information. Those with missing information had lower mean cognitive scores at ages 8 (P = 0.09), 11 (P = 0.02), 15 (P = 0.05) and 26 (P = 0.1) years compared with the others.

Cognitive test scores

Tables 1–3 show mean differences (95% CI) in standardized cognitive scores for each birthweight, current height and current weight category at ages 8, 15 and 26. These are shown unadjusted, then adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age.

Birthweight was significantly associated with cognitive test scores at ages 8 to 43, and with likelihood of obtaining advanced educational qualifications, before and after adjustment, although the effect became more linear after adjustment for the background variables, as previously reported.6 However, although there remained a trend across the birthweight categories at age 8, the effect at ages 15 and 26 was due largely to the lower mean cognitive score in the smallest birthweight group compared with those of higher birthweight.

Height was strongly and positively associated with cognition at three ages before adjustment. These scores were attenuated after adjustment at ages 8 and 15, the former falling just below the 5% significance level. Nevertheless, at this age the difference in mean score between shortest and tallest groups was 0.24 standard deviations (95% CI : 0.06–0.42), with the difference at age 15 being very similar. At age 26 years, the effect of height was larger, remaining stable after adjustment, with the difference in mean score between shortest and tallest groups being 0.46 standard deviations (95% CI : 0.27–0.64).

Weight was also positively associated with cognition at ages 8 and 15 years before adjustment, but was strongly attenuated, and no longer significant at the 5% level, at both ages after adjustment. However, the association between weight at age 26 and cognition at this age was strengthened by adjustment, becoming significant at the 5% level. The difference in mean score between the lightest and heaviest groups was 0.27 standard deviations (95% CI : 0.08–0.46).

Table 4 shows similar analyses for verbal memory at age 43 years. Consistent with our previous analyses,6 birthweight was not associated with this test, before or after adjustment. However, adult height was positively associated after adjustment. Current weight was strongly negatively associated before adjustment, and while scores were attenuated after adjustment, the association remained significant at the 5% level.

There was no obvious lack of fit of any model apart from the slight skewness exhibited by the cognitive test score at 26 years (see Method: statistics).

Interactions between birthweight and current height, and birthweight and current weight were tested at ages 8, 15 and 26 years. No interactions were significant at the 5% level. There was little evidence, therefore, that the effects of height and weight on cognition at these ages were dependent on birthweight. Nor were any height × sex or weight × sex interactions significant at the 5% level at these ages.

To test whether change in height was associated with cognitive score, each of the above models up to age 26 were further adjusted for height at the previous age. The association between height at age 7 and cognitive test score at age 8 was reduced by simultaneously adjusting for height at ages 2 and 4 (P = 0.21). There was no association at the 5% level between height at age 2 and cognition at age 8 (P = 0.89), and therefore no evidence that birthweight was a marker for height growth in the first 2 years of life in its effect on cognition. On the other hand, height at age 4 was significantly associated with cognition at age 8 (P = 0.03). Overall this suggests that height growth after age 2 is associated with cognitive development, with the suggestion of an important period between ages 2 and 4. It should be noted that the significant effect of height growth between ages 2 and 4 on cognition was due to the effect of the shortest group (≤0.96 m) in particular.

The association between height and the cognitive test score at age 15 was strongly reduced after adjusting for height at age 8, and was no longer significant at the 5% level (P = 0.32), suggesting no effect of height growth between ages 8 and 15 on cognition. However, the association between height and cognition at age 26 remained significant at the 5% level after adjusting for height at age 15 (P = 0.001), suggesting a second important period of height growth for cognition, between ages 15 and 26. Furthermore, the negative association between weight and cognition at age 26 remained significant when weight at age 15 was added, and indeed was strengthened in this way (P = 0.001), suggesting that weight gain during this time accounts for the negative association with adult cognition. As before, all these associations were independent of birthweight and the family background variables.

Educational attainment

After adjustment, birthweight, adult height and current weight all showed independent associations with educational attain-ment (Table 5). Those in the bottom two birthweight groups had a lower chance of advanced educational attainment than those in the heavier three groups. The effects of current weight and height acted in opposite directions, with increasing odds of high educational attainment with increasing height (odds ratio [OR] = 2.57, 95% CI : 1.55–4.26 for the difference between shortest and tallest category) but decreasing odds for increasing weight (OR = 0.32, 95% CI : 0.19–0.54 for the difference between lightest and heaviest category).

The Hosmer-Lemeshow test for the full logistic regression model for education indicated that the model was a reasonable fit to the data (χ2 = 8.11, d.f. = 8, P = 0.4).

Discussion

In a longitudinal population-based study we confirmed our previous findings6 that birthweight was positively associated with cognition up to early adulthood, and with likelihood of obtaining advanced educational qualifications, after controlling for social background. However, we have also shown that this association is independent of postnatal size. In addition, height was positively and independently associated with cognitive ability through midlife, and also with educational attainment. Weight, on the other hand, although not independently associated with cognition in childhood and adolescence, was negatively associated with cognition in early adulthood, and with educational attainment. Conditional analyses also suggested two important periods for the effect of height growth on cognition, between ages 2 and 4, and after age 15. Weight gain after age 15 was also negatively associated with cognition in early adulthood, although there was no independent effect of weight gain on cognition in childhood or adolescence. Overall, these results suggest two things. First, birthweight is specifically associated with cognitive and educational outcomes, and is not a marker for postnatal body size in this respect. Rather, this association reflects factors occurring during fetal development, as implied by the fetal origins hypothesis.11 Second, postnatal size is independently associated with cognition, with the suggestion of positive effects of height growth in early childhood and late adolescence, and a negative effect of weight gain in late adolescence.

We should highlight several potential limitations of the present study. First, cohort members with missing birthweight and confounding information had lower cognitive scores than those with complete information. However, in a previous study the unadjusted associations between birthweight and cognitive score in analyses of cohort members with birthweight and cognitive data but no confounder information were similar to those with complete confounder information.6

Second, although our analyses adjusted for important features of the early family background, these are crude measures, leaving the possibility of residual effects of social background operating after statistical adjustment. It is worth noting in this context that a recent study found an association between birthweight and cognitive performance using a within-sibship design,7 although this study did not investigate postnatal height and weight.

Third, we should note potential limitations of our data. For example, gestational age was unknown, which limits the interpretation of birthweight as an index of fetal growth in this cohort. Weight at age 26 years was self-reported rather than measured, where there is a tendency to under-report this aspect of body size.17 However, all other body size measures in this study were taken directly by doctors or nurses. Also, for practical reasons heights and weights were measured over a period of months around each survey interval. However, results were unchanged when the actual measurement dates were taken into account.

Finally, we should point out that the effects of body size on cognition, although statistically significant, were relatively small, at least in comparison to the effects of the social background variables, particularly birth order (not shown).

Previous authors have reported that associations between body size and cognitive performance are dependent on socioeconomic circumstances,18 and are more apparent in children of developing countries,19,20 or in economically disadvantaged children of developed countries, than in advantaged children.20 Our finding that the association between growth and cognitive ability was independent of social background is therefore important, particularly in regard to childhood. The potential role of adult social factors in the association between growth between 15 and 26 and cognition requires further investigation.

If the association between growth and cognition is independent of social background, what are the most likely mechanisms? Further studies will be required to investigate possible environmental factors, including diet, childhood illness and psychosocial stress. In terms of biological factors, one explanation is that IGF target brain areas responsible for cognition while playing a critical role in determining body growth.9 If so, then these results suggest that both fetal and postnatal programming associated with the hypothalamic-pituitary axis have significant consequences for cognitive development. The lack of an effect of height growth from childhood to adolescence on cognition seems at odds with the finding that IGF-I concentration accelerates most sharply during the adolescent growth spurt.21 It is possible, however, that high levels of IGF at this stage have a latent effect on central nervous system function, reflected in the apparent cognitive benefit of height growth after adolescence.

Another biological candidate is that of thyroid hormone, which also has dual somatic and central effects, stimulating linear growth, development and maturation of bone and also influencing neural development and learning capacity. This may also be relevant to the negative association between weight and cognition in early adulthood, since among many things hypothyroidism is associated with cognitive impairment and weight gain. Thyroid hormone may also interact with IGF.22 An alternative biological explanation for the negative association between weight in adulthood and cognition may be found with cortisol, a glucocorticoid hormone of the adrenal cortex. Cortisol increases appetite and stimulates lipogenesis in certain adipose tissues, and is negatively associated with cognitive performance, although evidence for the latter largely derives from studies of cognitive ageing.23 Glucocorticoid regulation is also partly under the control of thyroid hormone.24 Finally, cortisol also diminishes IGF production, so the IGF may provide a common pathway for the effects of several hormones involved in growth. However, such biological pathways should be considered alongside environmental factors that mediate the association between weight gain and cognitive disadvantage, such as diet and physical lifestyle. Further investigation is clearly warranted.

Table 1

Mean differences (95% CI) in standardized cognitive score in tests at age 8 years (n = 2758) for each birthweight, current height and current weight category (compared with middle category, 1. = lowest, 5. = highest). Differences are shown unadjusted, then adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age

Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.27 (–0.46 to –0.09)  –0.22 (–0.39 to –0.06)  
    2. 2.51–3.0 kg –0.06 (–0.17–0.04) <0.001 –0.02 (–0.12–0.08) <0.001 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg 0.15 (0.06–0.24)  0.17 (0.09–0.25)  
    5. 4.01–5.0 kg 0.04 (–0.08–0.17)  0.17 (0.06–0.29)  
Current height Current height  Current height  
    1. ≤1.12 m –0.43 (–0.56 to –0.30)  –0.14 (–0.28 to –0.01)  
    2. 1.12–1.14 m –0.10 (–0.22–0.03) <0.001 0.05 (–0.07–0.17) 0.06 
    3. 1.14–1.19 m   
    4. 1.20–1.24 m 0.15 (0.06–0.24)  0.03 (–0.05–0.12)  
    5. 1.25–1.42 m 0.32 (0.20–0.43)  0.10 (–0.02–0.22)  
Current weight Current weight  Current weight  
    1. ≤19.1 kg –0.41 (–0.53 to –0.28)  –0.14 (–0.28–0.00)  
    2. 19.2–21.0 kg –0.12 (–0.23 to –0.03) <0.001 –0.02 (–0.11–0.08) 0.32 
    3. 21.1–24.0 kg   
    4. 24.1–25.4 kg 0.09 (–0.02–0.21)  0.01 (–0.10–0.12)  
    5. 25.5–42.2 kg 0.19 (0.08–0.30)  –0.04 (–0.16–0.08)  
Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.27 (–0.46 to –0.09)  –0.22 (–0.39 to –0.06)  
    2. 2.51–3.0 kg –0.06 (–0.17–0.04) <0.001 –0.02 (–0.12–0.08) <0.001 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg 0.15 (0.06–0.24)  0.17 (0.09–0.25)  
    5. 4.01–5.0 kg 0.04 (–0.08–0.17)  0.17 (0.06–0.29)  
Current height Current height  Current height  
    1. ≤1.12 m –0.43 (–0.56 to –0.30)  –0.14 (–0.28 to –0.01)  
    2. 1.12–1.14 m –0.10 (–0.22–0.03) <0.001 0.05 (–0.07–0.17) 0.06 
    3. 1.14–1.19 m   
    4. 1.20–1.24 m 0.15 (0.06–0.24)  0.03 (–0.05–0.12)  
    5. 1.25–1.42 m 0.32 (0.20–0.43)  0.10 (–0.02–0.22)  
Current weight Current weight  Current weight  
    1. ≤19.1 kg –0.41 (–0.53 to –0.28)  –0.14 (–0.28–0.00)  
    2. 19.2–21.0 kg –0.12 (–0.23 to –0.03) <0.001 –0.02 (–0.11–0.08) 0.32 
    3. 21.1–24.0 kg   
    4. 24.1–25.4 kg 0.09 (–0.02–0.21)  0.01 (–0.10–0.12)  
    5. 25.5–42.2 kg 0.19 (0.08–0.30)  –0.04 (–0.16–0.08)  
Table 2

Mean differences (95% CI) in standardized cognitive score in tests at age 15 years (n = 2900) for each birthweight, current height and current weight category (compared with middle category, 1. = lowest, 5. = highest). Differences are shown unadjusted, then adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age

Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.22 (–0.40 to –0.04)  –0.17 (–0.33 to –0.02)  
    2. 2.51–3.0 kg –0.07 (–0.18–0.03) 0.02 –0.05 (–0.14–0.04) 0.02 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg 0.05 (–0.03–0.14)  0.06 (–0.03–0.12)  
    5. 4.01–5.0 kg –0.02 (–0.15–0.10)  0.08 (–0.03–0.19)  
Current height Current height  Current height  
    1. ≤1.50 m –0.37 (–0.49 to –0.24)  –0.17 (–0.30 to –0.04)  
    2. 1.50–1.55 m –0.12 (–0.22 to –0.01) <0.001 –0.01 (–0.10–0.09) 0.02 
    3. 1.55–1.60 m   
    4. 1.60–1.65 m 0.11 (0.07–0.21)  0.02 (–0.07–0.11)  
    5. 1.65–1.88 m 0.27 (0.17–0.37)  0.08 (–0.02–0.18)  
Current weight Current weight  Current weight  
    1. ≤42.2 kg –0.26 (–0.38 to –0.14)  –0.08 (–0.19–0.04)  
    2. 42.3–47.0 kg –0.12 (–0.22 to –0.01) <0.001 –0.02 (–0.11–0.08) 0.74 
    3. 47.1–54.0 kg   
    4. 54.1–60.0 kg 0.10 (–0.01–0.20)  0.01 (–0.08–0.10)  
    5. 60.1–108.4 kg 0.11 (0.01–0.22)  0.02 (–0.08–0.12)  
Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.22 (–0.40 to –0.04)  –0.17 (–0.33 to –0.02)  
    2. 2.51–3.0 kg –0.07 (–0.18–0.03) 0.02 –0.05 (–0.14–0.04) 0.02 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg 0.05 (–0.03–0.14)  0.06 (–0.03–0.12)  
    5. 4.01–5.0 kg –0.02 (–0.15–0.10)  0.08 (–0.03–0.19)  
Current height Current height  Current height  
    1. ≤1.50 m –0.37 (–0.49 to –0.24)  –0.17 (–0.30 to –0.04)  
    2. 1.50–1.55 m –0.12 (–0.22 to –0.01) <0.001 –0.01 (–0.10–0.09) 0.02 
    3. 1.55–1.60 m   
    4. 1.60–1.65 m 0.11 (0.07–0.21)  0.02 (–0.07–0.11)  
    5. 1.65–1.88 m 0.27 (0.17–0.37)  0.08 (–0.02–0.18)  
Current weight Current weight  Current weight  
    1. ≤42.2 kg –0.26 (–0.38 to –0.14)  –0.08 (–0.19–0.04)  
    2. 42.3–47.0 kg –0.12 (–0.22 to –0.01) <0.001 –0.02 (–0.11–0.08) 0.74 
    3. 47.1–54.0 kg   
    4. 54.1–60.0 kg 0.10 (–0.01–0.20)  0.01 (–0.08–0.10)  
    5. 60.1–108.4 kg 0.11 (0.01–0.22)  0.02 (–0.08–0.12)  
Table 3

Mean differences (95% CI) in standardized cognitive score in tests at age 26 years (n = 2098) for each birthweight, adult height and current weight category (compared with middle category, 1. = lowest, 5. = highest). Differences are shown unadjusted, than adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age

Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.37 (–0.58 to –0.16)  –0.35 (–0.54 to –0.17)  
    2. 2.51–3.0 kg –0.08 (–0.21–0.04) 0.008 –0.09 (–0.20–0.02) 0.003 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg –0.01 (–0.11–0.09)  0.00 (–0.08–0.09)  
    5. 4.01–5.0 kg –0.08 (–0.23–0.07)  0.02 (–0.12–0.15)  
Adult height Adult height  Adult height  
    1. ≤1.60 m –0.28 (–0.42 to –0.15)  –0.29 (–0.43 to –0.15)  
    2. 1.60–1.65 m –0.03 (–0.17–0.10) <0.001 –0.10 (–0.22–0.03) <0.001 
    3. 1.65–1.70 m   
    4. 1.70–1.78 m 0.13 (0.00–0.26)  0.14 (0.02–0.27)  
    5. 1.78–1.97 m 0.20 (0.06–0.33)  0.17 (0.02–0.32)  
Current weight Current weight  Current weight  
    1. ≤52.0 kg –0.06 (–0.21–0.08)  0.21 (0.06–0.36)  
    2. 52.1–60.0 kg 0.04 (–0.07–0.15) 0.55 0.15 (0.04–0.26) 0.01 
    3. 60.1–70.0 kg   
    4. 70.1–80.0 kg 0.05 (–0.07–0.16)  0.00 (–0.11–0.10)  
    5. 80.1–146.1 kg –0.04 (–0.17–0.10)  –0.06 (–0.19–0.07)  
Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.37 (–0.58 to –0.16)  –0.35 (–0.54 to –0.17)  
    2. 2.51–3.0 kg –0.08 (–0.21–0.04) 0.008 –0.09 (–0.20–0.02) 0.003 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg –0.01 (–0.11–0.09)  0.00 (–0.08–0.09)  
    5. 4.01–5.0 kg –0.08 (–0.23–0.07)  0.02 (–0.12–0.15)  
Adult height Adult height  Adult height  
    1. ≤1.60 m –0.28 (–0.42 to –0.15)  –0.29 (–0.43 to –0.15)  
    2. 1.60–1.65 m –0.03 (–0.17–0.10) <0.001 –0.10 (–0.22–0.03) <0.001 
    3. 1.65–1.70 m   
    4. 1.70–1.78 m 0.13 (0.00–0.26)  0.14 (0.02–0.27)  
    5. 1.78–1.97 m 0.20 (0.06–0.33)  0.17 (0.02–0.32)  
Current weight Current weight  Current weight  
    1. ≤52.0 kg –0.06 (–0.21–0.08)  0.21 (0.06–0.36)  
    2. 52.1–60.0 kg 0.04 (–0.07–0.15) 0.55 0.15 (0.04–0.26) 0.01 
    3. 60.1–70.0 kg   
    4. 70.1–80.0 kg 0.05 (–0.07–0.16)  0.00 (–0.11–0.10)  
    5. 80.1–146.1 kg –0.04 (–0.17–0.10)  –0.06 (–0.19–0.07)  
Table 4

Mean differences (95% CI) in standardized verbal memory score at age 43 years (n = 2136) for each birthweight, adult height and current weight category (compared with middle category, 1. = lowest, 5. = highest). Differences are shown unadjusted, then adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age

Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.18 (–0.39–0.04)  –0.15 (–0.36–0.05)  
    2. 2.51–3.0 kg –0.08 (–0.21–0.05) 0.23 –0.10 (–0.22–0.02) 0.36 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg –0.07 (–0.17–0.03)  –0.03 (–0.13–0.07)  
    5. 4.01–5.0 kg –0.14 (–0.29–0.01)  –0.02 (–0.17–0.12)  
Adult height Adult height  Adult height  
    1. ≤1.60 m –0.05 (–0.19–0.09)  –0.17 (–0.32 to –0.02)  
    2. 1.60–1.65 m 0.11 (–0.03–0.25) 0.10 0.01 (–0.13–0.14) 0.01 
    3. 1.65–1.70 m   
    4. 1.70–1.78 m –0.05 (–0.18–0.08)  0.10 (–0.04–0.23)  
    5. 1.78–1.97 m –0.03 (–0.17–0.12)  0.16 (0.00–0.32)  
Current weight Current weight  Current weight  
    1. ≤60.0 kg 0.21 (0.09–0.34)  0.15 (0.01–0.29)  
    2. 60.1–70.0 kg 0.09 (–0.03–0.20) <0.001 0.03 (–0.09–0.15) 0.03 
    3. 70.1–80.0 kg   
    4. 80.1–90.0 kg –0.13 (–0.26–0.01)  –0.09 (–0.22–0.04)  
    5. 90.1–138.0 kg –0.11 (–0.26–0.04)  –0.08 (–0.23–0.06)  
Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg –0.18 (–0.39–0.04)  –0.15 (–0.36–0.05)  
    2. 2.51–3.0 kg –0.08 (–0.21–0.05) 0.23 –0.10 (–0.22–0.02) 0.36 
    3. 3.01–3.5 kg   
    4. 3.51–4.0 kg –0.07 (–0.17–0.03)  –0.03 (–0.13–0.07)  
    5. 4.01–5.0 kg –0.14 (–0.29–0.01)  –0.02 (–0.17–0.12)  
Adult height Adult height  Adult height  
    1. ≤1.60 m –0.05 (–0.19–0.09)  –0.17 (–0.32 to –0.02)  
    2. 1.60–1.65 m 0.11 (–0.03–0.25) 0.10 0.01 (–0.13–0.14) 0.01 
    3. 1.65–1.70 m   
    4. 1.70–1.78 m –0.05 (–0.18–0.08)  0.10 (–0.04–0.23)  
    5. 1.78–1.97 m –0.03 (–0.17–0.12)  0.16 (0.00–0.32)  
Current weight Current weight  Current weight  
    1. ≤60.0 kg 0.21 (0.09–0.34)  0.15 (0.01–0.29)  
    2. 60.1–70.0 kg 0.09 (–0.03–0.20) <0.001 0.03 (–0.09–0.15) 0.03 
    3. 70.1–80.0 kg   
    4. 80.1–90.0 kg –0.13 (–0.26–0.01)  –0.09 (–0.22–0.04)  
    5. 90.1–138.0 kg –0.11 (–0.26–0.04)  –0.08 (–0.23–0.06)  
Table 5

Odds ratios (95% CI) for obtaining advanced educational qualifications for each birthweight, adult height and age 26 weight category (compared with middle category, 1. = lowest, 5. = highest). Differences are shown unadjusted, then adjusted for each other, and for sex, birth order, father's occupational social class, mother's education, and mother's age (n = 2405)

Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg 0.46 (0.28–0.76)  0.46 (0.26–0.81)  
    2. 2.51–3.0 kg 0.74 (0.57–0.95) 0.009 0.69 (0.51–0.93) 0.02 
    3. 3.01–3.5 kg 1.00  1.00  
    4. 3.51–4.0 kg 0.94 (0.78–1.15)  0.93 (0.73–1.17)  
    5. 4.01–5.0 kg 0.85 (0.63–1.14)  0.94 (0.66–1.35)  
Adult height Adult height  Adult height  
    1. ≤1.60 m 0.53 (0.39–0.72)  0.55 (0.36–0.82)  
    2. 1.60–1.65 m 0.82 (0.61–1.09) <0.001 0.83 (0.59–1.17) 0.006 
    3. 1.65–1.70 m 1.00  1.00  
    4. 1.70–1.78 m 1.50 (1.15–1.94)  1.20 (0.87–1.66)  
    5. 1.78–1.97 m 2.00 (1.51–2.64)  1.41 (0.95–2.08)  
Weight at age 26 Weight at age 26  Weight at age 26  
    1. ≤52.0 kg 0.96 (0.71–1.28)  2.57 (1.68–3.90)  
    2. 52.1–60.0 kg 0.85 (0.68–1.07) 0.18 1.42 (1.05–1.92) <0.001 
    3. 60.1–70.0 kg 1.00  1.00  
    4. 70.1–80.0 kg 1.17 (0.92–1.46)  0.96 (0.71–1.28)  
    5. 80.1–146.1 kg 1.09 (0.83–1.43)  0.81 (0.57–1.16)  
Category boundary Unadjusted P-value Adjusted P-value 
Birthweight Birthweight  Birthweight  
    1. ≤2.5 kg 0.46 (0.28–0.76)  0.46 (0.26–0.81)  
    2. 2.51–3.0 kg 0.74 (0.57–0.95) 0.009 0.69 (0.51–0.93) 0.02 
    3. 3.01–3.5 kg 1.00  1.00  
    4. 3.51–4.0 kg 0.94 (0.78–1.15)  0.93 (0.73–1.17)  
    5. 4.01–5.0 kg 0.85 (0.63–1.14)  0.94 (0.66–1.35)  
Adult height Adult height  Adult height  
    1. ≤1.60 m 0.53 (0.39–0.72)  0.55 (0.36–0.82)  
    2. 1.60–1.65 m 0.82 (0.61–1.09) <0.001 0.83 (0.59–1.17) 0.006 
    3. 1.65–1.70 m 1.00  1.00  
    4. 1.70–1.78 m 1.50 (1.15–1.94)  1.20 (0.87–1.66)  
    5. 1.78–1.97 m 2.00 (1.51–2.64)  1.41 (0.95–2.08)  
Weight at age 26 Weight at age 26  Weight at age 26  
    1. ≤52.0 kg 0.96 (0.71–1.28)  2.57 (1.68–3.90)  
    2. 52.1–60.0 kg 0.85 (0.68–1.07) 0.18 1.42 (1.05–1.92) <0.001 
    3. 60.1–70.0 kg 1.00  1.00  
    4. 70.1–80.0 kg 1.17 (0.92–1.46)  0.96 (0.71–1.28)  
    5. 80.1–146.1 kg 1.09 (0.83–1.43)  0.81 (0.57–1.16)  

Funding for this study was provided by the Medical Research Council.

References

1
Breslau N. Psychiatric sequelae of low birthweight.
Epidemiol Rev
 
1995
;
17
:
96
–106.
2
Record RG, McKeown T, Edwards JH. The relation of measured intelligence to birth weight and duration of gestation.
Ann Hum Genet
 
1969
;
33
:
71
–79.
3
Seidman DS, Laor A, Gale R, Stevenson DK, Mashaich S, Danon YL. Birth weight and intellectual performance in late adolescence.
Obstet Gynecol
 
1992
;
79
:
543
–46.
4
Paz I, Laor A, Gale R, Harlap S, Stevenson DK, Seidman DS. Term infants with fetal growth restrictions are not at increased risk for low intelligence scores at 17 years.
J Pediatr
 
2001
;
138
:
87
–91.
5
Sorensen HT, Sabroe S, Olsen J, Rothman KJ, Gilman MW, Fischer P. Birth weight and cognitive function in young adult life: historical cohort study.
BMJ
 
1997
;
315
:
401
–03.
6
Richards M, Hardy R, Kuh D, Wadsworth MEJ. Birth weight and cognitive function in the British 1946 birth cohort: longitudinal population based study.
BMJ
 
2001
;
322
:
199
–203.
7
Matte TD, Breshnahan M, Begg M, Susser E. Influence of variation in birth weight within normal range and within sibships on IQ at age 7 years: cohort study.
BMJ
 
2001
;
323
:
310
–14.
8
Shenkin SD, Starr JM, Pattie A, Rush MA, Whalley LJ, Deary IJ. Birth weight and cognitive function at age 11 years: the Scottish Mental Survey 1932.
Arch Dis Child
 
2001
;
85
:
189
–96.
9
Berger A. Insulin like growth factor and cognitive function.
BMJ
 
2001
;
322
:
203
.
10
Lucas A, Fewtrell MS, Cole TJ. Fetal origins of adult disease—the hypothesis revisited.
BMJ
 
1999
;
319
:
245
–49.
11
Barker DJP. Mothers, Babies and Health in Later Life. 2nd Edn. Edinburgh: Churchill Livingstone, 1998.
12
Wadsworth MEJ. The Imprint of Time: Childhood, History and Adult Life. Oxford: Clarendon Press, 1991.
13
Wadsworth MEJ, Mann SL, Rodgers B, Kuh DL, Hilder WS, Yusuf EJ. Loss and representativeness in a 43 year follow-up of a national birth cohort.
J Epidemiol Community Health
 
1992
;
46
:
300
–04.
14
Pless IB, Crips HA, Davies JMC, Wadsworth MEJ. Chronic physical illness in childhood and psychological and social circumstances in adolescence and early adult life.
Dev Med Child Neurol
 
1989
;
31
:
746
–55.
15
Pigeon DA. Tests used in the 1954 and 1957 surveys. In: Douglas JWB (ed.). The Home and the School. London: Macgibbon & Kee, 1964, Appendix 1.
16
Pigeon DA. Details of the fifteen years tests. In: Douglas JWB, Ross JM, Simpson HR (eds). All Our Future. London: Davies, 1968, Appendix 1.
17
Casey VA, Dwyer JT, Coleman KA, Krall EA, Gardner J, Valadian I. Accuracy of recall by middle-aged participants in a longitudinal study of their body size and indices of maturation earlier in life.
Ann Human Biol
 
1991
;
18
:
155
–66.
18
Tanner JM. Galton, eugenics and the study of growth: the relation of body size, intelligence test score, and social circumstances in children and adults.
Eugenics Rev
 
1966
;
58
:
122
–35.
19
Cheung YB, Yip PSF, Karlberg JPE. Fetal growth, early postnatal growth and motor development in Pakistani infants.
Int J Epidemiol
 
2001
;
30
:
66
–72.
20
Karp R, Martin R, Sewell T, Manni J, Heller A. Growth and academic achievement in inner-city kindergarten children. The relationship of height, weight, cognitive ability, and neurodevelopmental level.
Clin Pediatr
 
1992
;
31
:
336
–40.
21
Juul A, Bang P, Hertel NT et al. Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index.
J Clin Endocrinol Metab
 
1994
;
78
:
744
–52.
22
Sinclair D, Dangerfield P. Human Growth After Birth. Oxford: Oxford University Press, 1998.
23
Kalmijn S, Launer LJ, Stolk RP et al. A prospective study on cortisol, dehydroepiandrosterone sulfate, and cognitive function in the elderly.
J Clin Endocrinol Metab
 
1998
;
83
:
3487
–92.
24
Sapolsky RM. The importance of the well-groomed child.
Science
 
1997
;
277
:
1620
–21.