A cohort approach to tobacco use and mortality: the case of Quebec

Prev Med. 1996 Nov-Dec;25(6):730-40. doi: 10.1006/pmed.1996.0113.


Background: Cigarette smoking is the No. 1 cause of premature death among Canadians. However, tobacco consumption dropped by 35% in Canada during the 1980s.

Methods: Using data collected during a Quebec health survey, we evaluated the prevalences of tobacco use by birth cohort. Since mortality has been decreasing for both sexes, we computed gender-specific cohort mortality indices for tobacco-related causes of death. In the analysis, we took into account the inevitable time lag between adopting the behavior and becoming the victim of a smoking-attributable disease.

Results: Results indicate a systematic decrease in tobacco use from older male cohorts to younger ones across all ages; however, for females an increase in tobacco use has been observed from one cohort to the next but there seems to be hope for a future trend toward breaking the habit. While a decline in tobacco-related mortality has been observed among men (e.g., death from ischemic heart disease), female lung cancer mortality increased considerably. Finally, we present an indicator measuring the years of potential life lost for some tobacco-related causes of death.

Conclusions: The cohort approach allows us to observe birth-cohort-specific trends, thereby more easily relating past behavior to current mortality levels. In terms of prevention, increasing female mortality from lung cancer should become a major concern and a motivator in the fight against tobacco use.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Age Distribution
  • Aged
  • Cause of Death* / trends
  • Cohort Effect
  • Cohort Studies
  • Female
  • Humans
  • Longevity
  • Male
  • Middle Aged
  • Myocardial Ischemia / epidemiology
  • Prevalence
  • Quebec / epidemiology
  • Respiratory Tract Neoplasms / epidemiology
  • Sex Distribution
  • Smoking / adverse effects*
  • Smoking / epidemiology*
  • Smoking Cessation / statistics & numerical data