Article Text

Download PDFPDF
OP85 Psychiatric symptoms and premature mortality in the 1946 British birth cohort
  1. G Archer1,
  2. M Hotopf2,
  3. M Stafford1,
  4. D Kuh1,
  5. M Richards1
  1. 1MRC Unit for Lifelong Health and Ageing, University College London, London, UK
  2. 2Institute of Psychiatry, Kings College London, London, UK


Background Despite accounting for nearly half of all ill-health in those aged under 65, and 23% of the total UK burden of disease, mental health receives only 13% of NHS expenditure. Moreover, these figures do not take into account the potential effect of mental health on physical morbidity. Several previous studies have shown an association between mental disorders and earlier mortality; however the relationship remains largely unexplained. This study will establish whether there is an association between psychiatric symptoms and premature mortality, and if so, whether this relationship can be explained by a range of potential mediators and confounders.

Methods In the nationally representative prospective 1946 British birth cohort study, psychiatric symptoms were assessed at age 36 (1982) using the Present-State-Examination, a clinically validated interview; symptoms were coded according to an index-of-definition into ‘none’, ‘mild’ or ‘severe’. Mortality was obtained from the National Health Service Central Register until end of May 2012, an average follow-up of 29.4 years. The study sample consisted of 3305 participants, of which there were 343 deaths. Covariates included measures of adult and childhood socio-economic position, health behaviours, physical health function, and social networks. Stata 12 was used to perform all analyses. Multiple-imputation was used for missing covariate data, and Cox regression was used to estimate hazard ratios.

Results After adjustment for sex, ‘severe’ psychiatric symptoms appeared to increase the risk of all-cause mortality by 50% compared to those who had no symptoms (HR=1.50, 95% CI 1.14–1.98). After full adjustment for covariates, this effect was no longer statistically significant (HR=1.20, 0.89–1.61). The association was partially explained by smoking, pulse rate, marital status, and frequency of contact with friends. Excluding violent deaths and suicides led to similar results. An almost identical trend was observed with respect to cancer mortality.

Conclusion Severe psychiatric symptoms in adulthood appear to increase the risk of premature mortality. This effect may be partially explained by smoking, social networks, and physical health function. Quantifying the effect of mental health on mortality could help influence policy regarding NHS expenditure, whilst establishing potential pathways between psychiatric symptoms and mortality may highlight effective points of intervention.

  • mental health
  • premature mortality
  • birth cohort

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.