Is intelligence associated with health and longevity and if so, why? Discuss with reference to relevant empirical studies.
Associations between measures of socioeconomic status (SES), health and survival rates have been firmly established by social epidemiologists. Nevertheless, in spite of improving western living conditions and wide access to health services and information, social inequalities in health and mortality still prevail. In some instances, wide accessibility to health care appears only to deepen socioeconomic divisions of health and mortality rates (Gottfredson, 2004). Why this should be the case has led researchers to seek out further predictors of morbidity and mortality across the social spectrum. From these efforts a new field of study, cognitive epidemiology (Deary & Der 2005) is seeking to establish general intelligence scores as risk factors for mortality and certain health outcomes. This is a comparatively new field of research, and findings with regard to the extent and exact causal mechanisms of an association between intelligence, health and longevity remain far from conclusive. For instance, some researchers claim that intelligence is the ‘elusive fundamental cause’ of health inequalities (Gottfredson, 2004); whilst others negate the impact of intelligence once SES factors and education are controlled for (Singh-Manoux & Kivimaki, 2009; Hauser & Palloni, 2008). Intelligence tests are long established as valid predictors of occupational success and educational attainment, both of which have a direct bearing on health outcomes. The question is to what extent, if at all, SES and other risk factors, are proxies of intelligence, and do levels of mental ability have a direct impact on health outcomes, or is its effect indirect and mediated by other factors?
Given the heterogeneity of its study, it is important to note that intelligence, as referred to in this essay, is based on individual psychometric test scores of g (Spearman, 1904), or general intelligence, which concerns differing abilities to reason, comprehend, solve problems, and learn from experience. A unitary general intelligence is embraced by cognitive epidemiology as it provides a reliable and viable measurement which remains largely stable across the lifespan. Furthermore, general intelligence appears to adhere to a normal distribution in a population (Neisser, 1996). The large genetic component of general intelligence, and an increase in size of genetic effect across the lifespan, can lend intelligence an element of autonomy against SES, group membership and environmental factors (Plomin & Bergeman, 1991).
Over the past decade, a number of studies have confirmed an inverse association between general intelligence and mortality from late childhood to mid-adulthood. The Scottish Mental Health Surveys of 1932 and 1947 provided an opportune data set to empirically test for an association between childhood psychometric intelligence scores and later all-cause mortality risk. Through retrospective cross-reference between pre-morbid intelligence scores, subsequent death registers and medical records, researchers were able to report that a one S.D disadvantage (15 points) in intelligence at the age of 11 was associated with a 21% increased risk of dying by age 76 (Whalley & Deary, 2001). National conscription records also provided ready-made data sets of mental ability scores for follow-up study. The Swedish Conscripts Study (Batty & Wennerstad, 2009) reported that a 1 S.D advantage in intelligence at the age of enlistment (17-18) related to a 24% decrease in mortality risk. The Whitehall II Study (Marmot & Brunner, 2006) took their baseline at mid-adulthood and found that a 1 S.D disadvantage in general intelligence test scores was associated with a 16% risk of all-cause mortality. Moreover, these findings have been replicated by further studies which provide ‘strength in numbers’ (Deary, 2010).

Despite these compelling and seemingly consistent results, childhood SES factors are potential confounders. In order to clarify the situation, a systematic review of 9 longitudinal cohort studies was conducted. In the three cases where relevant data was available, researchers made statistical adjustments for early life SES factors and, albeit with modest attenuation, found the association between pre-morbid intelligence and mortality risk to remain significant. So maybe early individual differences in cognitive ability are independent of early SES inequalities in predicting all-cause mortality (Batty, et.al. 2007). However, 3 out of 9 studies hardly provide a firm basis upon which to delineate between early SES and cognitive function, and doesn’t preclude the impact of other confounding variables such as parenting style, cognitive stimulation, sibling order and diet.

The situation becomes more confounded by adult SES and education and the possible interaction with cognitive ability. Certainly, there appears to be inconsistent results in this regard with some studies; Wisconsin Longitudinal Study (Hauser and Palloni, 2008), US National Longitudinal Survey of Youth (Jokela, Elovaio, Singh-Manoux & Kivimaki, 2009, and the Vietnam Experience Study (Weiss et. al. 2009) reporting that any association between early life IQ and mortality risk was significantly attenuated when adjusted for SES and education. Other studies, notably Aberdeen Children of 1950’s (Batty et. al. 2004) were more modestly attenuated. A meta-analysis of 16 longitudinal cohort studies supported the intelligence-mortality hypothesis that early life intelligence was associated with a 24% lower risk of all-cause mortality, and that this estimate retained its significance after adjustment for covariates of SES and education (Calvin, 2010). However, it is generally agreed that the exact nature of the interaction between intelligence, health and mortality, SES and educational attainment remains unclear (Calvin, 2010). In this light, Gottfredson’s claim that SES and Education are surrogates of intelligence appears somewhat ‘gauntlet throwing’ (Deary, 2005). Certain studies, notably Danish Metropolit and Whitehall II Studies do not support her hypothesis. However, The West of Scotland Twenty-07 Study, and other similar studies, found that intelligence ranked as highly as other risk factors, including SES. In this light, it would be unwise to overlook the potential impact of intelligence in any meaningful exploration of prevailing health inequalities (Lubinski & Humphreys, 2004). Bearing, this cautious summation in mind, I shall now examine a number of possible explanations, childhood and adult SES aside, for an association between intelligence, morbidity and mortality risk.

One explanation suggests that the levels of intelligence involved in personal health management are often underestimated. Increased levels of health literacy may relate to higher intelligence due to an increased capacity for undirected learning, assimilation of ever complex medical advice and compliance in healthcare settings. Certainly, claims that health literacy can predict health outcomes even after SES factors are controlled for are gaining ground (Gottfredson 2004). Greater intelligence may also be associated with the deployment of preventive health behaviors, such as increased exercise, reduced smoking, moderate drinking and adherence to dietary advice. Accidents and unintentional injury are also a major cause of early life morbidity and mortality. It is well established that some lower-status occupations involve risk-laden environments, though it would appear that some people are just more accident-prone than others. Could this be due to individual intelligence differences? Cognitive epidemiologists suggest that accident prevention is a highly cognitive process, (Batty, Deary & Gottfredson, 2007). An early cohort study, Australian Veterans Health Study (O’Toole, 1990) reported that IQ was the most efficient predictor of death from motor accidents amongst veterans under 40. It is important to note, that the above explanations do encroach on theories of practical and emotional intelligence (Sternberg, 2000), and personality traits such as conscientiousness and extraversion, all of which bear a predictive load in terms of health management, personal safety and risk management.

Cohort studies have also highlighted the association between cognitive ability and specific health conditions, and further explanations focus on the physiological correlates of intelligence scores. Childhood or early adult scores may reflect possible developmental problems, injuries, low birth weight and generally poor nurture as indications of later poor health and increased mortality risk. Also a number of cognitive epidemiological studies (Deary & Der, 2005) have investigated reaction times as indicators of levels of intelligence, and thus a reflection of brain function and system integrity. In turn, neural efficiency (Eysenck, 1986; Vernon, 1987) may indicate wider somatic efficiency of organs, nervous system and metabolic function. It has been suggested that there may be an association between metabolic syndrome, (hypertension, obesity, and poor glucose metabolism), and intelligence. The Vietnam Experience Study (Weiss et. al. 2009) found that a one S.D advantage in intelligence at enlistment was related to a 14% reduced risk of metabolic syndrome in mid life. The intelligence-metabolic syndrome association was replicated in the UK’s National Survey of Health & Development (Richards, et. al. 2009).

In conclusion, associations have been established between early life intelligence, certain illnesses and all-cause mortality, but Gottfredson’s claim that intelligence is the ‘fundamental elusive cause’ of health inequalities remains largely unsupported until further light is shed on just exactly how and why intelligence, given its strong genotype-environment correlation (Plomin, 1999), is mediated by other predictors such as SES, education, and other individual characteristics such as personality traits and physiological differences. Furthermore, for reasons of space, evolutionary explanations (Kamazawa, 2006) of intelligence-mortality association have not been touched on. Current empirical evidence is inconsistent, and methodologically flawed in so much that it is reliant on westernized pre-existing data sets, e.g. male-dominant conscription records, rather than sample populations truly representative of gender differences and socio-cultural diversity. The way-forward points to collaborative research between cognitive epidemiology, intelligence theorists, medical sociology, geneticists and individual difference psychologists in order to uncouple SES, education and intelligence, thoroughly test the system integrity hypothesis, and evaluate the impact of other individual differences such as personality traits on health outcomes. It would be fair to state that intelligence has claimed ‘a seat at the table of epidemiology’ (Deary & Batty, 2011:702), but it is perhaps premature to state that the strength of intelligence-mortality association is comparable with traditional early life physiological risk factors such as birth weight, cardiovascular function and blood pressure (Deary, 2008). However, there are certainly indications that this may actually be the case, and thus it may not be wholly inconceivable to envisage a future whereby psychometric intelligence testing may be a paid-up member epidemiological study, and even general medical practice, alongside SES, education and familial health history.
Word count 1,644

Dear reader, once again I am struggling with the required word count. Please bear in mind that I would have wished to discuss evolution and intelligence, and also explore associations between mental heath and cognitive ability

References

Batty, G, Deary I. J. & Gottredson, L. 2007. Premorbid (early life) IQ and Later Mortality Risk: Systematic Review. Annals of Epidemiology; 17

Calvin, C., Deary, I. J., Fenton, C., Roberts, B., Der, G., Leckenby, N., & Batty, G. D. 2010. Intelligence in youth and all-cause mortality: systematic review with meta-analysis. International Journal of Epidemiology, October

Deary, I. J. 2010. Cognitive epidemiology: Its rise, its current issues, and its challenges. Personality and Individual Differences, (49).

Deary, I. J. 2008. Why do intelligent people live longer? Nature. Vol 456. 13.

Deary I. J. & Batty, G. D. 2011 Intelligence as a Predictor of Health, Illness and Death, Chapter 20. The Cambridge Handbook of Intelligence, 2011. Sternberg & Kaufman (eds). Cambridge University Press.

Deary, I. J. Whiteman, M., Starr, J. ,Whalley, L. J., & Fox, H. C. 2004 The Impact of Childhood Intelligence on Later Life: Following up the Scottish Mental Surveys of 1932 and 1947. Journal of Personality and Social Psychology, Vol. 86, No. 1

Gottfredson, L. 2004. Intelligence: Is it the Epidemiologists ‘Elusive Fundamental Cause’ of Social Inequalities in Health. Journal of Personality & Social Psychology, Vol. 86, No. 1

Gottfredson, L.& Deary J., 2004. Intelligence Predicts Health and Longevity, but Why? 2004. Current Directions in Psychological Science, Vol.13. No.1

Kanazawa S., 2006. Mind the gap ... in intelligence: Re-examining the relationship between inequality and health. British Journal of Health Psychology (11)

Link, B., Phelan, J., Miech, R., & Leckman Westin, E. 2008. The Resources That Matter: Fundamental Social Causes of Healh Disparities and the Challenge of Intelligence. Journal of Health and Social Behaviour, Vol. 49 (March)

Lubinski, D. 2009. Cogntive Epidemiology. With emphasis on untangling cognitive ability and socioeconomic status. Intelligence., (37).

Plomin R. 1999. Genetics and general cognitive ability. Nature. Vol 402. Supp. 2 December
Neisser, U., 1996. Intelligence: Knowns and Unknowns. American Psychologist, Vol. 51. No.2

Singh-Manoux A., Ferrie J., Lynch J. & Marmot, M. 2005. The Role of Cognitive Ability (Intelligence) in Explaining the Association between Socioeconomic Position and Health: Evidence from the Whitehall II Prospective Cohort Study, American Journal of Epidemiology, Vol. 161, No.9