Childhood cognitive ability and deaths up until middle age: a post-war birth cohort study

Int J Epidemiol. 2004 Apr;33(2):408-13. doi: 10.1093/ije/dyh043.


Background: Childhood IQ has been related to mortality in later life in four studies. Cognitive ability may be a mediator between early disadvantage and mortality, a marker of the efficiency of information processing in the central nervous system, or predict entry to safe adult environments or healthy behaviours. We examined mortality in relation to cognitive ability at age 8 years in a birth cohort and investigated these possible reasons.

Methods: Cox's proportional hazards models were used to investigate the effect of early cognitive ability on all-cause mortality in 2057 women and 2192 men born in England, Scotland, and Wales in March 1946 and followed until age 54 years. We tested whether the relationship was accounted for by childhood socioeconomic conditions or serious illness, education, adult socioeconomic conditions, or smoking.

Results: Cognitive ability was related to mortality in men but not women. The excess mortality rate in men was concentrated in the bottom quarter of the cognitive score (hazard ratio [HR] for bottom versus top quarter 1.8, 95% CI: 1.0, 3.0) and there was no gradient across the range of ability. Adjustment for childhood socioeconomic conditions and serious illness had a small effect on the HR for deaths between 9 and 54 years while adjustment for education or early adult socioeconomic conditions halved the HR for deaths from age 26 years. Smoking was not a mediator of the effect of early ability on adult mortality.

Conclusions: Greater cumulative exposure to poor lifetime socioeconomic conditions is the most likely explanation for the observed relationship between low cognitive ability in childhood and mortality. This relationship may therefore be elucidated further by studying the causes of lifelong socioeconomic inequalities in health.

MeSH terms

  • Adult
  • Child
  • Educational Status
  • England / epidemiology
  • Epidemiologic Methods
  • Female
  • Hospitalization / statistics & numerical data
  • Humans
  • Intelligence*
  • Male
  • Middle Aged
  • Mortality*
  • Scotland / epidemiology
  • Smoking / adverse effects
  • Smoking / epidemiology
  • Social Class
  • Wales / epidemiology