Jha · The New England journal of medicine 2013 · prospective cohort study · n=202248

21st-century hazards of smoking and benefits of cessation in the United States.

Cited 1805 times in the scientific literature.

Level 3 - non-randomized controlled study

Nationally representative prospective cohort study with mortality linkage

PubMed 23343063 · doi:10.1056/NEJMsa1211128 · record verified 2026-08-30

What was done

Smoking and cessation histories were obtained from 113,752 women and 88,496 men aged 25 years or older interviewed between 1997 and 2004 in the U.S. National Health Interview Survey. These data were linked to mortality records through December 31, 2006, encompassing 8,236 deaths in women and 7,479 in men. Hazard ratios (HR) for all-cause mortality in participants aged 25 to 79 were calculated comparing current smokers to never-smokers, adjusting for age, educational level, adiposity, and alcohol consumption.

What was found

All-cause mortality was roughly tripled among current smokers compared to never-smokers (HR 3.0; 99% CI, 2.7 to 3.3 in women; HR 2.8; 99% CI, 2.4 to 3.1 in men). Survival probability from age 25 to 79 was 70% vs. 38% in women and 61% vs. 26% in men for never-smokers versus current smokers. Current smokers lost more than 10 years of life expectancy. Cessation at ages 25–34, 35–44, or 45–54 gained approximately 10, 9, and 6 years of life, respectively, compared with continued smoking. Quitting before age 40 reduced excess mortality associated with continuing to smoke by about 90%.

Why it matters

This study provides modern, nationally representative estimates showing that smoking hazards are now equally extreme in women and men, while confirming that cessation—especially before age 40—averts nearly all excess mortality.

Limits

Smoking and cessation status were self-reported at baseline survey interview, risking misclassification and unmeasured status changes during follow-up. Observational data cannot fully rule out residual confounding despite adjustment for education, body weight, and alcohol intake.

Cited by