Smoking as a risk factor for injury death: a meta-analysis of cohort studies

Prev Med. 1998 Nov-Dec;27(6):871-8. doi: 10.1006/pmed.1998.0374.


Background: Injury and tobacco effects represent one-quarter of the global burden of disease. Understanding the causes of injury and the effects of smoking may help reduce those burdens. Some smokers have high risks of injury. We provide an initial meta-analysis of cohort associations between smoking and fatal injury.

Methods: Three authors independently searched MEDLINE, and bibliographies of the pertinent studies found, for cigarette smoker-specific injury death data which allowed estimation of an appropriate relative risk (RR) and 95% confidence interval (CI). Relative risks and dose response were summarized by fixed effects and Poisson modeling, respectively.

Results: Six studies covering 10 pertinent cohorts were located. Associations between smoking and injury death have been significant after adjustment or, in effect, stratification for age, race, sex, country, and, respectively, alcohol, marriage, education, and body mass; job and time period; job, alcohol, and exercise; etc. Summary dose-response trends were significantly positive (P < 0.00005). Cigarette smoking predicted summary injury death crude RRs of 1.61 (CI 1.44-1.81) vs never smokers and 1.39 (CI 1.25-1.55) vs ex-smokers.

Conclusions: Smoking has significant, consistent, dose-response, often strong and independent, prospective associations with injury death, internationally.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Cohort Studies
  • Effect Modifier, Epidemiologic
  • Female
  • Humans
  • Male
  • Middle Aged
  • Poisson Distribution
  • Predictive Value of Tests
  • Research Design
  • Risk
  • Risk Factors
  • Smoking / adverse effects*
  • Wounds and Injuries / etiology*
  • Wounds and Injuries / mortality*