The relationship between intelligence and health in Luxembourg

Even though the general health and life expectancies of Western societies have been consistently rising throughout the 20th century, socioeconomic health inequalities continue to persist. Individuals from lower socioeconomic groups have substantially worse health and an increased mortality risk compared to individuals from higher socioeconomic groups. As external factors such as material resources cannot fully account for these health inequalities, personal factors such as intellectual abilities have been suggested as additional important explanatory factors. The research field concerned with the effects of intelligence on different health outcomes is called cognitive epidemiology. Results from this field of research have now established that childhood intelligence is an important predictor of different health outcomes in adulthood. Specifically, children with higher intelligence exhibit a lower mortality risk and enjoy better health in adulthood compared to children with lower childhood intelligence.
Despite these findings, several open research questions remain: (1) Almost all previous studies on the relation between childhood intelligence and adult health have been conducted in English-speaking or Scandinavian countries. Can these findings be generalized to countries with different cultural backgrounds, health-care systems, or levels of social mobility? Specifically, Luxembourg offers universal access to quality health care, which may compensate for some of the effects of individual differences in intelligence on health, and as a result, intelligence may lose its impact. (2) Physical health is a multidimensional concept with three distinct subdimensions: a physical subdimension (e.g., presence of diagnosed diseases, number of doctor visits in a certain time period), a subjective subdimension (e.g., satisfaction with one’s own health), and a (social-)functional subdimension (e.g., unimpaired participation in social and occupational activities or performing household tasks). Most previous studies on the relation between intelligence and health outcomes have focused on the physical health subdimension. Hence, considerably less is known about the effects of intelligence on the functional and subjective subdimensions. This issue is of particular importance as childhood intelligence may be differentially related to different aspects of adult health. (3) It remains unclear whether different facets of childhood intelligence (e.g., general, fluid, or crystallized intelligence) predict adult health equally well, as most studies on the topic have used only global measures of childhood intelligence to predict later health. However, investigating different facets of childhood intelligence as predictors of adult health would provide insights into which facets of intelligence are important in personal health management and could be targeted by interventions. (4) Despite considerable evidence that has emphasized the relevance of education and further indicators of subsequent socioeconomic status (SES) as potential mediators between childhood intelligence and later health outcomes, previous research has yielded inconsistent results regarding the extent to which these relations are mediated. Some studies have reported pronounced mediation via education and subsequent SES, yet others have reported little or no mediation. However, knowing and understanding which mediational mechanisms underlie the intelligence-health relation and the extent to which they mediate this relation are crucial for applying findings from cognitive epidemiology to public health. (5) As is the case for studies on intelligence and health outcomes, most studies on the relation between intelligence and mortality risk have been conducted in English-speaking or Scandinavian countries. Thus, it remains unclear whether the results of these studies can be generalized to Luxembourg. Further, there is controversy about whether the effect of intelligence on mortality exists across the entire range of intelligence scores or whether individuals at the lower end of the intelligence distribution constitute a risk group with a particularly high mortality risk. Identifying potential risk groups is crucial for determining which groups should be targeted by interventions to reduce inequalities in health and mortality risk.
The present Ph.D. thesis addressed these five research questions with three distinctive studies. Study I investigated whether childhood intelligence would predict adult physical, functional, and subjective health 40 years later even when controlling for the effects of childhood SES. Study I also investigated whether a global measure of childhood general intelligence or whether more specific facets such as fluid and crystallized intelligence would better predict adult health. Study II investigated whether and the extent to which educational attainment and SES in adulthood would mediate the effects of childhood intelligence on the three adult health dimensions. Study III investigated whether childhood intelligence would predict adult mortality risk when controlling for childhood SES and whether individuals at the lower end of the intelligence distribution would constitute a risk group with a particularly high mortality risk.
All three studies were embedded in the general framework of the Luxembourgish MAGRIP project. This large-scale longitudinal study comprised two waves of measurement over a 40-year period. In the first wave of measurement in 1968, detailed intelligence and socioeconomic data were collected on a randomly selected nationally representative sample comprising 2,824 students at the end of their primary education (M = 11.9 years; SD = 0.6 years; 50.1% male). In the second wave conducted 40 years later, 745 participants (M = 51.7 years, SD = 0.6 years; 46.7% male) provided data on their educational careers, adult SES, and functional, subjective, and physical health. In addition, the mortality rate was established for the participants in the first wave of MAGRIP: 166 participants (69.9% male) had died.
The results of the three studies demonstrated that childhood intelligence, particularly childhood fluid intelligence, showed a significant association with adult health: Lower childhood intelligence scores were associated with worse health outcomes on all three dimensions of physical, functional, and subjective health in adulthood, even when controlling for childhood SES. These effects were entirely mediated via educational attainment and adult SES, with educational attainment playing a crucial role in these mediational processes. Further, childhood intelligence showed a significant association with adult mortality such that lower childhood intelligence scores were associated with an increased mortality risk. This effect was particularly strong among men at the lower end of the intelligence distribution. These results suggest that even high-quality public health care cannot fully offset the cumulative effects of childhood intelligence on adult health. Intelligence may thus be an important explanatory factor for socioeconomic inequalities in health. Promising means for reducing these socioeconomic health inequalities consist of interventions that are designed to improve childhood intelligence, to improve environments for childhood physical and intellectual development, and to make public health care and preventive measures or treatments accessible to adults with lower intellectual abilities.