Losing life and livelihood: a systematic review and meta-analysis of unemployment and all-cause mortality

Authors:
ROELFS David J., et al
Journal article citation:
Social Science and Medicine, 72(6), March 2011, pp.840-854.
Publisher:
Elsevier

Meta-analyses of data from 42 studies from developed nations revealed that unemployment is associated with a significant increase in overall mortality, especially in males and those at the beginning, or middle stage, of their careers. Future studies should continue to investigate mediating, moderate and confounding factors, especially in terms of those that may be modifiable risk factors.

Extended abstract:
Author

ROELFS David J

Losing life and livelihood: a systematic review and meta-analysis of unemployment and all-cause mortality.

Journal citation/publication details

Social Science and Medicine, 72(6), March 2011, pp.840-854.

Summary

Meta-analyses of data from 42 studies from developed nations revealed that unemployment is associated with a significant increase in overall mortality, especially in males and those at the beginning, or middle stage, of their careers. Future studies should continue to investigate mediating, moderate and confounding factors, especially in terms of those that may be modifiable risk factors.

Context

The number of unemployed people worldwide has risen steadily over the last decade, prompting an increase in research on the health effects of unemployment. Most studies have found that unemployment is associated with decreased longevity, but there is no consensus on the degree to which longevity is reduced in population sub-groups, or on the most important mediating, moderating, and confounding factors involved. The aim of this study was to focus on these factors by: evaluating the impact of pre-existing health status and health behaviours on all-cause mortality; comparing the potential moderating effect in countries with and without national health care systems, and; assessing the potential moderating roles of gender, age, time, duration of follow-up, and case-control group composition on the association between unemployment and mortality.

Methods

What sources were searched?
The electronic databases Medline, EMBASE, CINAHL, and Web of Science were searched in June 2005 and again in July 2008 and January 2009. Hand searches were carried out on the bibliographies of eligible publications and related articles.

What search terms/strategies were used?
Searches were performed using terms for psychological stress, stress disorders, mortality, unemployment, and a wide range of social factors. Full details of the search algorithm for Medline are presented in an appendix and details of the other search strategies are available from the authors on request. The searches were performed by a research librarian.

What criteria were used to decide on which studies to include?
Studies were included if all-cause mortality was the outcome variable, unemployment was measured at the individual level, and the results were compared between a study population that experienced unemployment and  one that did not experience unemployment at all or experienced it to a lesser extent. Searches were carried out in English but publications found to be published in other languages were included if relevant.

Who decided on their relevance and quality?
Two named authors were responsible for study selection and coding, and a third author was consulted as necessary. Study quality was assessed, using the Newcastle-Ottawa scale for non-randomised trials, by the same two authors working independently; the average rating for each study was used in the analysis. The study selection process is outlined in Figure 1 and includes the number of studies included at each stage.

How many studies were included and where were they from?
A total of 1,570 publications was identified from the database searches; 48 articles met the study inclusion criteria and 30 were included in the review. In all, 232 articles identified from hand search were also included. The pool of 262 publications reported on a range of psychosocial stressors; the current review used a subset of 42 articles that focused on the association between unemployment and all-cause mortality. There were eight studies from the USA, seven each from the UK and Sweden, six from Finland, two each from Denmark, Israel and Japan, and single studies from eight other countries. 

How were the study findings combined?
Mortality risk estimates were extracted from the 42 included studies. Odds ratios and relative risks were converted to hazard ratios. Meta-analyses and meta-regression analyses were conducted using a random effects model. Further details of statistical methods are included in the study text and appendices.


 

Findings of the review

In all, 235 mortality risk estimates from 42 studies, and representing more than 20 million people, were analysed. The majority were from men, and almost all were individuals of working-age at baseline.

Unemployed persons were significantly more likely to die than those in a comparator group; the hazard ratio adjusted for age and other covariates was 1.63, showing that unemployment is associated with a 63% higher risk of mortality. The average effect was higher for men than women with an increased risk of 78% compared to 37%, respectively. Unemployed people in their early or mid careers faced an increased risk of 73% and 77%, compared to 25% for those in their late careers. The risk of death was over 70% in the first ten years of follow-up but fell to 42% after that, although the trend was not significant in the final meta-regression model.

Studies that controlled for any measure of health showed no significant difference in the magnitude of risk compared to remaining studies but the hazard ratio was reduced by 24% for studies that controlled for one or more health behaviours, compared to the other studies. This suggests that health behaviour may confound the association between unemployment and mortality and also that pre-existing health behaviour and health conditions do not account for all of the relationship.

No significant difference in mortality was found between unemployed people in the USA, which has no universal health coverage, and the Scandinavian nations combined, where public health care coverage is most comprehensive, or the remaining nations. This suggests that national-level differences in policy may not affect the rate of mortality after unemployment in developed countries.


 

Authors' conclusions

‘Unemployment was associated with a substantially increased risk of death among broad segments of the population. Future research should continue to focus on possible mediating, moderating, and confounding factors and on whether this risk is modifiable, either at the health system level or the individual level.’

Implications for policy or practice

None are discussed.

Subject terms:
mortality, risk, unemployment;
Content type:
systematic review
Link:
Journal home page
ISSN print:
0277-9536

Key to icons

  • Free resource Free resource
  • Journal article Journal article
  • Book Book
  • Digital media Digital media
  • Journal Journal

Give us your feedback

Social Care Online continues to be developed in response to user feedback.

Contact us with your comments and for any problems using the website.

Sign up/login for more

Register/login to access resource links, advanced search and email alerts