Smoking kills. But why does smoking seem more likely to kill now than a few decades ago? In 1987, a smoker had a 3.4% chance of dying within five years. By 2013, a smoker had a 4.9% chance of drying within five years. Ralph Lawton, Ellen Meara, and colleagues discovered an answer to this puzzle by analyzing representative data from the 1987–2013 waves of the US National Health Interview Survey (NHIS). By breaking the years of the NHIS into three periods of time and estimating the age-specific mortality in each, the authors demonstrate that the difference in risk of death between smokers and non-smokers at each age is about the same in each eight-year period. For example, a 65-year-old smoker had a 10 percent higher chance of death than a 65-year-old non-smoker in the late 1980s, and that risk was the same in the 2010s. The authors conclude that changing age distributions explain most of the increase in smoking-related mortality risk. The US population is aging and, in addition, fewer young people are picking up the habit as smoking declines in social acceptance. Smokers are, on average, older than they were decades ago, and mortality risks associated with smoking rise with age. According to the authors, similar methods could be used to examine the effects of other risk factors that vary with age, potentially including obesity, poor diet, or lack of education.
PNAS Nexus
Why the mortality penalty of smoking is rising over time
22-Sep-2026