OBJECTIVE. The purpose of this work was to examine the relation of scores on tests of mental ability in childhood with food consumption and physical activity in adulthood.
METHODS. Based on a cohort of >17000 individuals born in Great Britain in 1970, 8282 had complete data for mental ability scores at 10 years of age and reported their food intake and physical activity patterns at 30 years of age.
RESULTS. Children with higher mental ability scores reported significantly more frequent consumption of fruit, vegetables (cooked and raw), wholemeal bread, poultry, fish, and foods fried in vegetable oil in adulthood. They were also more likely to have a lower intake of chips (French fries), nonwholemeal bread, and cakes and biscuits. There was some attenuation in these associations after adjustment for markers of socioeconomic position across the life course, which included educational attainment, with statistical significance lost in some analyses. Higher mental ability was positively associated with exercise habit, in particular, intense activity (defined by being out of breath/sweaty). The associations between mental ability and these behaviors were similar in both men and women, and they were somewhat stronger for verbal than nonverbal ability.
CONCLUSIONS. It is plausible that the skills captured by IQ tests, such as the ability to comprehend and reason, may be important in the successful management of a person's health behaviors.
Developed a century ago, written tests of intelligence (mental ability) have been used regularly in educational and workplace settings. However, in the last decade, a number of investigators have examined the predictive value of mental ability for health behaviors, such as alcohol consumption and smoking.1,2 In recent reports, high childhood mental ability scores were associated with a reduced risk of smoking initiation3,4 and alcohol-induced hangovers (a proxy for binge drinking)5 and increased smoking cessation rates6 in some, if not all,7 studies.
One mechanism advanced for these findings is the differential interpretation of, and responses to, health advice by persons with different mental ability test scores.6 This explanation is consistent with the established definition of intelligence as the ability to learn, reason, and solve problems.8 On this basis, similar associations between childhood mental ability and other health behaviors, such as physical activity and dietary choice, would be hypothesized. The reports that link high childhood mental ability scores with a reduced risk of adult obesity/overweight provide some indirect support for a relation of early life mental ability with physical activity and diet.9–14 However, we are unaware of any studies that directly report on associations between ability and these behaviors.
Examining the links between mental ability and these health behaviors might also assist in the identification of some of the mechanisms responsible for the recent consistent finding that high mental ability tests scores in childhood seem to confer protection against the later risk of total mortality.15 It has been suggested2,16 that this relation may be at least partially mediated by established adult risk factors for premature mortality, including physical inactivity17–19 and some dietary characteristics (such as low fruit and vegetable consumption and high red meat intake).20,21 We investigated the relation between childhood mental ability and adult patterns of food intake and physical exertion using data from the 1970 British Cohort Study.
The 1970 British Cohort Study is an ongoing longitudinal study that takes as its subjects all of the 17198 live births occurring to parents residing in Great Britain between April 5 and 11, 1970.22 After the initial 1970 survey, there have been 5 major contacts with cohort members to monitor their physical, educational, and social development; transitions to adult life23; and, in the latest study, adult identity. These studies were conducted in 1975 (when the cohort members were aged 5 years), 1980–1981 (aged 10 years), 1986 (aged 16 years), 1996 (aged 26 years), and 1999–2000 (aged 30 years). For the surveys in 1975, 1980–1981, and 1986, the cohort was augmented by the inclusion of immigrants to Britain who had also been born in the target week in 1970. The present analyses use data from 1980 to 1981, when study participants completed cognitive ability tests at 10 years old,24 and from 1999 to 2000 when, at aged 30 years, they responded to enquiries about their diet and physical activity levels,25 among other health behaviors.
Data Collected at the 10-Year Follow-up
Written informed consent was given by parents of study participants. Testing of the children's mental ability took place in schools. Mental ability at the age of 10 years was assessed using a modified version of the British Ability Scales,26 which had been adapted to facilitate administration by teachers. Verbal ability was assessed using 2 subscales: word definitions and word similarities. The word definitions subscale consisted of a list of 37 words. The teacher articulated each word in turn and quizzed the child about its meaning. The word similarities subscale consisted of 42 items composed of 3 words (eg, orange, banana, and strawberry or sad, worried, and happy). For each item, the teacher enunciated the 3 words and asked the child to name another word consistent with the theme. Nonverbal ability was also assessed using 2 subscales: recall of digits and matrices. The recall of digits subscale consisted of 34 items. For each item, the teacher read out digits at half-second intervals and asked the child to repeat them. The matrices subscale consisted of 28 incomplete patterns arrayed as a grid. For each item, the teacher asked the child to draw in the missing part of the pattern. Test results were scored by trained coders. Reliability of coding was monitored throughout the survey and results fed back to the coders. Regular checks conducted on a 5% random sample of tests showed that the percentage of tests where the original code was not confirmed on recoding was low: 0.8% for recall of digits, 4.7% for word definitions, 1.9% for word similarities, and 2.3% for matrices.27
Information on father's occupation was collected during the interview with the child's parents. Social class was derived from the father's then-current occupation using 6 categories according to the 1980 Registrar General's Classification of Occupations (professional, managerial, skilled nonmanual, skilled manual, semiskilled, and unskilled).28 Such classifications are broadly generalizable: occupational categories used in European countries other than the United Kingdom and in the United States are similarly based on the skills and status of different jobs.29
Data Collected at the 30-Year Follow-up
After written informed consent, information was collected by interview in the participant's home. As part of a series of questions on health, cohort members were asked to respond to items regarding how often they ate a range of foods: fresh fruit, cooked vegetables, raw vegetables or salads, wholemeal bread, other bread, red meat, poultry, fish, eggs, pulses, cakes and biscuits, sweets and chocolates, chips (French fries), food fried in hard fat, and food fried in vegetable oil. Responses were recorded using 7 response categories: more than once a day, once a day, 3 to 6 days a week, 1 to 2 days a week, <1 day a week, occasionally, or never. Participants were also asked 3 questions regarding their participation in regular physical activity. If, in the first, they answered positively to taking part in any regular exercise they were asked to indicate how often they participated by checking 1 of 6 categories (every day, 4–5 days a week, 2–3 days a week, once a week, 2–3 times a month, or less often). They were also asked to indicate how often they got out of breath or sweaty during exercise using 1 of 4 responses (most of the time, sometimes, rarely, or never).
Participants were asked about their highest academic or vocational qualification. Academic qualifications, in order of increasing attainment, were: Certificate of Secondary Education (CSE) grades 2 to 5 (normally taken at minimum school leaving age); ordinary level General Certificate of Education ([O level] normally taken at minimum school leaving age); advanced level General Certificate of Education ([A level] normally taken at 18 years old); degree or diploma (bachelor's degree or higher education diploma); or higher degree (masters or doctorate). Vocational qualifications consisted of National Vocational Qualifications (NVQs) and other vocationally based credentials of an equivalent standard. NVQs are based on national occupational standards and are awarded for evidence of competency in work-based situations at 5 levels, reflecting increasing job complexity and personal responsibility. These academic and vocational qualifications were subsequently collapsed into 6 categories, reflecting increasing attainment: no qualifications, CSE grades 2 to 5/NVQ level 1 and equivalent, O levels/NVQ level 2 and equivalent, A levels/NVQ level 3 and equivalent, degree or diploma/NVQ level 4 and equivalent, or higher degree/NVQ level 5. The subject's current social class (professional, managerial, skilled nonmanual, skilled manual, semiskilled, or unskilled) was derived from his or her own occupation.30 Participants were asked to provide information on their gross and net earnings from employment (United Kingdom pounds sterling).
In all, 14875 children took part in the 10-year follow-up, representing 93% of those eligible to participate (alive and living in Great Britain). Cognitive testing was completed on 11563 (78%). Parental refusal for this part of the study and refusal of teachers to administer the tests were the main reasons why some children did not sit the cognitive tests. By the time of the 30-year follow-up, the cohort consisted of 16695 individuals. Of these study participants, 14087 were traced, 13394 were eligible to take part, and 11261 were interviewed for the 30-year follow-up, representing 84% of those eligible to participate. In total, 8282 (74% of those interviewed) had data on cognitive function at the age of 10 years and were, therefore, included in our analyses. Compared with these 8282 men and women, cohort members who did not participate in the 30-year follow-up had slightly lower verbal and nonverbal mental ability scores at age 10 years: their mean (SD) verbal ability score was 20.9 (7.1) compared with 22.6 (6.9), and nonverbal ability score was 36.2 (8.2) compared with 38.1 (7.9; P < .001 for both).
Further comparison of cohort members with and without complete data for potential confounding or mediating variables showed that the relations between mental ability scores and the adult dietary and exercise outcomes were similar in the 2 groups. On the basis of the results of these preliminary analyses, therefore, the present analytical sample seems to be representative of the full cohort. However, in a further effort to avoid any suggestion of selection bias, we retained cohort members with incomplete data in the analysis by creating an extra category for missing data within the father's and current social class variables.
We used analysis of variance, correlation coefficients, and the χ2 test to examine the relation between test scores and characteristics of the participants. We used ordinal logistic regression to assess the association of mental ability at 10 years old with increased frequency of eating various foods at 30 years old, taking exercise more frequently, and exercising with greater intensity (as measured by frequency of getting out of breath or sweaty). Ordinal logistic regression is a method of analyzing categorical data with >2 categories where the categories are ordered in a natural way,31 such as “most of the time,” “sometimes,” “rarely,” or “never.” It makes use of information about the ordering, producing a cumulative odds and assuming that the odds for each “cutoff” category are proportional. Testing the proportional odds assumption in preliminary analyses using binary logistic regression modeling of each dichotomized response suggested that this was a valid approach. We therefore used binary logistic regression to examine the relation between mental ability at 10 years old and the likelihood of engaging in any regular exercise at 30 years old (“yes” versus “no”). Odds ratios (ORs) with accompanying 95% confidence intervals (CIs) were expressed per 1 SD increase in mental ability score and are shown adjusted for gender, father's social class at 10 years old, own occupational social class, highest academic/vocational qualification, and annual net earnings at 30 years old.
Table 1 shows the characteristics of the study participants in relation to verbal and nonverbal mental ability scores at 10 years old. As expected, test performance was strongly associated with socioeconomic circumstances across the life course. Thus, those persons who were from, or themselves in, a nonmanual occupation, had markedly higher childhood test scores than their counterparts in manual occupations. There was also a strong relation with academic/vocational qualifications achieved by 30 years old, with greater mental ability test scores at 10 years of age apparent in persons with higher attainment. Data on current annual earnings were available for 6045 (73%) of the participants. Annual earnings at 30 years old were positively related to both verbal (rs = 0.30; P < .001) and nonverbal mental ability (rs = 0.22; P < .001) in childhood. Men had a slightly higher average nonverbal mental ability score than women.
Table 2 shows the ORs for greater frequency of consumption of various foods at 30 years old in relation to verbal ability at 10 years old. In general, persons with higher childhood mental ability scores reported what could be considered to be a healthier consumption of food than those in the lower-scoring groups. Thus, in gender-adjusted analyses, study participants with higher verbal ability in childhood were more likely than those of lower verbal ability to have a greater frequency of consumption of fresh fruit (OR per 1 SD increase in mental ability; 95% CI: 1.30; 1.25–1.35), cooked (1.26; 1.23–1.34) and raw vegetables (1.27; 1.22–1.32), wholemeal bread (1.23; 1.18–1.28), poultry (1.10; 1.05–1.14), fish (1.27; 1.21–1.31), and foods fried in vegetable oil (1.19; 1.15–1.24). Persons with higher mental ability scores were more likely to have a lower frequency of consumption of chips (0.74; 0.71–0.77) and cakes and biscuits (0.95; 0.92–0.99). These relationships were generally attenuated toward unity after separate adjustment for childhood and current socioeconomic position and particularly by educational attainment (correlation with verbal ability: rs = 0.40; P < .001). Statistical significance at conventional levels was retained after multiple adjustment for mental ability in relation to the consumption of fruit, cooked or raw vegetables, fish, chips, and foods fried in vegetable oil but lost for other food items. Frequency of consumption of red meat and of bread other than wholemeal was inversely related to mental ability, but these associations were lost after control for adult indicators of socioeconomic circumstances.
For most of the associations described above, the direction and strength of the relation with childhood verbal mental ability was similar in both men and women. In univariate analysis, the frequency of consumption of fruit, raw vegetables, and food fried in vegetable oil was more strongly associated, albeit in the same direction, with verbal mental ability in women than it was in men. However, after multivariate adjustment only the relation of verbal ability with consumption of food fried in vegetable oil remained statistically significant (P value for interaction =.02). Thus, the multivariate-adjusted ORs (95% CI) for greater frequency of consumption of food cooked in vegetable oil were 1.09 (1.01–1.16) in men and 1.15 (1.06–1.24) in women in relation to verbal mental ability. When we related nonverbal childhood mental ability with food consumption, the same pattern of results as those seen for verbal ability was found with the magnitude of the relationships somewhat weaker (data not shown).
In Table 3, we show how verbal mental ability at 10 years of age relates to exercise habits at 30 years old after adjusting for potential confounding or mediating variables. In total, 79% of the participants reported that they took regular exercise. In gender-adjusted analyses, higher verbal mental ability was associated with taking regular exercise (OR: 1.20; 95% CI: 1.14–1.27), and getting out of breath/sweaty more frequently when exercising (OR: 1.30; 95% CI: 1.24–1.36). Both of these relationships were weakened somewhat when childhood and current social class or earnings were added separately to the multivariable model, but effect estimates did not cross unity. After adjustment for education, the relationship between verbal mental ability and likelihood of taking regular exercise was of borderline statistical significance. However, higher verbal ability in childhood remained a significant predictor of exercise intensity, even in the fully adjusted model: the OR for getting out of breath or sweaty more frequently during exercise was 1.15 (95% CI: 1.08–1.22). The relations between exercise habits and childhood verbal mental ability were similar in both men and women. Nonverbal ability was also associated with a higher likelihood of taking exercise and of exercising more intensively in gender-adjusted analyses, although less strongly than verbal ability (data not shown). These relations were also attenuated by adjustment for socioeconomic position.
In the present analyses we examined the hypothesis that high childhood mental ability test scores were related to favorable levels of food intake and physical activity in adult life. In the only study of which we are aware to examine these associations, we found that higher mental ability scores were positively associated with an increased consumption of foods that could be described as healthy and with higher levels of some indices of physical activity. These relations were attenuated by adjustment for indices of socioeconomic position across the life course, remaining statistically significant at conventional levels in some, if not all, analyses.
Although paternal social class may be considered as a confounding variable in the present analyses, education, income, and the subject's own social class may be conceptualized as mediators. Some attenuation of the relation of mental ability with diet and physical activity was seen when adult markers of socioeconomic disadvantage were taken into account. Although no formal mediation analyses were conducted (eg, using Structural Equation Modeling), these results implicate a pathway of events that lead to a more favorable pattern of food consumption and physical activity. Given the relatively close correlation between childhood mental ability and subsequent educational achievement, the inclusion of the latter in our multivariable models is a point of some debate.1 That is, the correlation between education and mental ability might be explained partly by the fact that variance in educational outcomes is, to a substantial extent, a reflection of differences in mental ability.32
Comparison With Other Studies
Because this is the first examination of the link between preadult intelligence quotient and later food intake and physical activity, direct comparison with other studies is not possible. However, several investigators have examined the link between education (a correlate of mental ability) and food intake and physical activity. Based on a review of 33 studies of adults representing 13 European countries, Roos et al33 reported that individuals with higher educational qualifications, particularly those from northern and western countries, tended to report higher consumption of fruit, vegetables,34 and low-fat milk products and cheese35 and lower quantities of meat than persons with more basic educational credentials. Differences in fat consumption, both total and saturated, across educational groups were less clear.36 Similar socioeconomic-food intake gradients have been observed in analyses of the present data set.37 Fewer studies have examined the link between educational attainment and physical activity.38 However, in adults from both the United Kingdom39 and Australia,40 there is evidence of a positive gradient.
Strengths and Limitations
The strengths of this study are its size (resulting in high statistical power), the representativeness of the sample (resulting in a high degree of generalizability), and the breadth of data on socioeconomic position (allowing an examination of the role of potential confounding and mediating variables). Inevitably, there are also some limitations. First, although this is the first study to examine the links between mental ability and food choice/diet, the construction of the enquiries pertaining to food consumption did not permit computations of dietary characteristics, such as saturated fat, fiber, or micronutrient intake. Second, the longitudinal nature of the present study has inevitably led to some attrition: only 46% of the participants at the 30-year follow-up had taken part in all of the earlier surveys of the cohort, although 80% of the participants in the 30-year follow-up had missed none or only 1 of these earlier sweeps, and 74% of them had taken the British Ability Scale tests as part of the 10-year follow-up.41 The participants at the 30-year follow-up gained significantly higher scores at age 10 for both verbal and nonverbal ability than those who did not take part, but the size of these differences was modest (0.2 SD). Unless the relation of childhood mental ability with adult dietary and exercise habits is in the opposite direction in nonresponders or those who have died to that found in the present analyses (a highly unlikely scenario), little bias will have been introduced. Finally, it is plausible that persons with low mental ability scores in childhood, a characteristic largely carried forward to adult life,42 might be less precise in their reporting of food intake and exercise. However, we are unaware of any examination of this hypothesis, and we were unable to test it using the present data set. If some participants were misclassified because of inaccuracies in recording diet or exercise this would tend to weaken any association, biasing estimates of risk toward unity.
Public Health Implications
At least 2 potential interventions could be used to address the apparent cognitive variation in diet and physical activity identified in the present study. First, the skills captured by mental ability tests, such as verbal comprehension and reasoning, may be important in the successful management of a these behaviors. Exercise and dietary modification advice could, therefore, be simplified to make it more appropriate for persons with lower cognitive ability. Potentially, this would be a rapidly introduced, adult-targeted intervention to address cognitive variation in these behaviors. Second, in a more far-reaching approach, early life educational programs could be implemented to increase childhood mental ability. Two recent reviews of early learning and school readiness interventions,43,44 1 of which focused on randomized trials only,43 concluded that these programs led to important improvements on tests of reading, arithmetic ability, and general intelligence that extended to secondary school ages. However, with a modest duration of follow-up in these trials, crucially, the extent to which these improvements are actually maintained across the life course is not clear.
In this study, we found that, in comparison with their lower scoring counterparts, persons with higher childhood mental ability scores were more likely to engage in physical activity and have a food intake that would generally be considered congruent with current information about a healthy lifestyle. These associations were strongest for verbal mental ability and were not always fully explained by differences in life course socioeconomic disadvantage.
The 10-year follow-up of the 1970 British Cohort Study was conducted by the Department of Child Health, Bristol University. The 30-year follow-up was conducted under the auspices of the Joint Centre for Longitudinal Research (comprising the Centre for Longitudinal Studies, Institute of Education; the International Centre for Health and Society, University College Medical School [both University of London]; and the National Centre for Social Research, London). We thank the United Kingdom Data Archive, University of Essex, for providing the data. David Batty is a Wellcome Fellow. Ian Deary is the recipient of a Royal Society-Wolfson Research Merit Award.
- Accepted August 24, 2006.
- Address correspondence to David Batty, PhD, Medical Research Council Social and Public Health Sciences Unit, University of Glasgow, 4 Lilybank Gardens, Glasgow, United Kingdom G12 8RZ. E-mail:
Dr Batty generated the idea for the analyses, which was developed by Drs Gale and Deary; Dr Gale conducted all of the data analyses; and Drs Gale and Batty wrote the first draft of the article to which all of the coauthors made substantial subsequent contributions.
The original data creators, depositors, or copyright holders; the funding agencies; and the United Kingdom Data Archive bear no responsibility for the analyses and interpretation presented here.
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Deary IJ, Batty GD. Cognitive epidemiology: a glossary of terms. J Epidemiol Community Health.2006; In press
- ↵Deary IJ, Batty GD. Commentary: pre-morbid IQ and later health: the rapidly evolving field of cognitive epidemiology. Int J Epidemiol.2006;35 :670– 672
- ↵Martin LT, Fitzmaurice GM, Kindlon DJ, Buka SL. Cognitive performance in childhood and early adult illness: a prospective cohort study. J Epidemiol Community Health.2004;58 :674– 679
- ↵Batty GD, Deary IJ, MacIntyre S. Childhood IQ and life course socio-economic position in relation to alcohol-induced hangovers in adulthood: the Aberdeen Children of the 1950s Study. J Epidemiol Community Health.2006;60 :872– 874
- ↵Taylor MD, Hart CL, Davey Smith G, et al. Childhood mental ability and smoking cessation in adulthood: prospective observational study linking the Scottish Mental Survey 1932 and the Midspan studies. J Epidemiol Community Health.2003;57 :464– 465
- Hart CL, Taylor MD, Davey Smith G, et al. Childhood IQ, social class, deprivation, and their relationships with mortality and morbidity risk in later life: prospective observational study linking the Scottish Mental Survey 1932 and the Midspan studies. Psychosom Med.2003;65 :877– 883
- ↵Batty GD, Deary IJ, Gottfredson LS. Pre-morbid (early life) IQ and later mortality risk: systematic review. Ann Epidemiol.2006; In press
- ↵Batty GD, Deary IJ. Early life intelligence and adult health. BMJ.2004;329 :585– 586
- ↵Batty GD, Lee IM. Physical activity for preventing strokes. BMJ.2002;325 :350– 351
- Batty GD, Lee IM. Physical activity and coronary heart disease. BMJ.2004;328 :1089– 1090
- ↵Batty GD, Thune I. Does physical activity prevent cancer? Evidence suggests protection against colon cancer and probably breast cancer. BMJ.2000;321 :1424– 1425
- ↵Ness AR, Powles JW. Fruit and vegetables, and cardiovascular disease: a review. Int J Epidemiol.1997;26 :1– 13
- ↵Butler N, Bonham D. Perinatal Mortality. Edinburgh, United Kingdom: Livingstone; 1963
- ↵Ferri E, Bynner J, Wadsworth M. Changing Britain, Changing Lives. London, United Kingdom: Institute of Education; 2003
- ↵Butler N, Bynner J. 1970 British Cohort Study: Ten-YearFollow-up. Colchester, Essex, United Kingdom: UK Data Archive; 1997
- ↵Joint Centre for Longitudinal Research. National Child Development Study and 1970 British Cohort Study (BSC70) Follow-ups, 1999–2000. Colchester, Essex, United Kingdom: UK Data Archive; 2003
- ↵Elliott C, Murray D, Pearson L. British Ability Scales. Windsor, United Kingdom: National Foundation for Educational Research; 1978
- ↵Butler N, Haslum M, Barker W, Morris A. Child Health and Education Study. First Report to the Department of Education and Science on the 10 Year Follow Up. Bristol, United Kingdom: University of Bristol; 1982
- ↵OPCS. Classification of Occupations 1980. London, United Kingdom: HMSO; 1980
- ↵Jones S. Identifying deprived areas using indices from the 1991 census and information about the recipients of community charge and council tax benefit. J Epidemiol Community Health1995;49(suppl 2) :S65– S71
- ↵Deary I, Strand S, Smith P, Fernandes C. Intelligence and educational achievement. Intelligence.2006; In press
- ↵Roos G, Prättälä R. FAIR-97-3096 Disparities Group (tasks 4 and 5). Disparities in Food Habits: Review of Research in 15 European Countries. Helsinki, Finland: National Public Health Institute; 1999
- ↵Schoon I, Parsons S. Life styles and health behaviour in three birth cohorts. In Ferri E, Bynner J, Wadsworth M, eds. Changing Britain, Changing Lives. Three Generations at the Turn of the Century. London, United Kingdom: Institute of Education Press; 2003
- ↵Activity and Health Research. Allied Dunbar National Fitness Survey. London, United Kingdom: Sports Council; 1992
- ↵Cerin E, Leslie E, Bauman A, Owen N. Levels of physical activity for colon cancer prevention compared with generic public health recommendations: population prevalence and sociodemographic correlates. Cancer Epidemiol Biomark Prev.2005;14 :1000– 1002
- ↵Plewis I, Calderwood L, Hawkes D, Nathan G. National Child Development Study and 1970 British Cohort Study Technical Report: Changes in the NCDS and BCS70 Populations and Samples Over Time. London, United Kingdom: Centre for Longitudinal Studies, Institute of Education; 2004
- ↵Ramey CT, Ramey SL. Early learning and school readiness: can early intervention make a difference? Merrill-Palmer Q Dev Psychol.2004;50 :471– 491
- ↵Barnett WS, Camilli G. Compensatory preschool education, cognitive development, and “race”. In: Fish JM, ed. Race and Intelligence. Separating Science From Myth. London, United Kingdom: Lawrence Erlbaum Associates; 2002:369– 406
- Copyright © 2007 by the American Academy of Pediatrics