Why Smoking Deaths Look Like They’re Rising When They Aren’t

Smoking Isn’t Getting More Dangerous, Aging Smokers Are Skewing the Numbers

Mortality risk tied to smoking hasn’t climbed much since the late 1980s, once researchers account for age, even though the gap in death rates between smokers and nonsmokers appeared to widen for decades. A new analysis in PNAS Nexus traces that illusion to a simple shift: the people who still smoke have gotten older.

Researchers tracked nearly 900,000 adults for almost thirty years, linking a major health survey to national death records through 2018. Comparing smokers and nonsmokers of the same age, the mortality gap tied to smoking barely budged between the late 1980s and 2013, even as standard comparisons made it look like smoking had grown far more dangerous.

Smokers as a group simply got older, the study found. As younger generations quit or never started, the remaining pool of smokers skewed older, and older people die at higher rates regardless of smoking status. By 2013, 59% of people who had ever smoked were over 50, compared with 42% in 1987, and once researchers accounted for that shift, most or all of the apparent rise disappeared.

Is Smoking Getting Deadlier? Three Decades of Data Say No

Data came from the National Health Interview Survey, a long-running federal survey covering 1987 through 2013, linked to death records through 2018. The sample included 892,908 adults ages 25 to 84, grouped by smoking history and tracked for deaths within five years of being surveyed.

First, the team repeated the kind of analysis used in earlier, well-known studies. That confirmed the alarming pattern others had found: the estimated five-year mortality gap between current and never smokers climbed from 3.4 to 4.9 percentage points between the earliest and latest survey periods, a 44% increase, with a similar pattern for ever smokers.

Next, researchers split the data into three periods, roughly 1987-1995, 1996-2004, and 2005-2013, comparing the mortality gap for each age group separately, asking a sharper question: had the gap changed for a 55-year-old smoker, compared with a 55-year-old smoker three decades earlier? It had not. Death rates for smokers were still much higher than for nonsmokers within any bracket, sometimes double, but that gap wasn’t growing once age was properly accounted for.

Aging Smokers, Not Rising Danger, Drive the Numbers

Standard analysis showed a rising gap because of who was smoking, not because smoking changed. Death rates climb steeply with age regardless, so an aging smoking population will naturally show a bigger gap in conventional comparisons, even if the underlying risk hasn’t changed. Standard age adjustments weren’t fully catching this shift, according to the study.

To test that theory, researchers ran several approaches separating aging from any true change in the smoking-mortality relationship. Holding the age makeup of smokers constant at 1987-1990 levels cut the apparent rise for current smokers by 63%, and the gap for ever smokers actually shrank slightly. Two more checks agreed, with changing age patterns explaining roughly 78% to 82% of the rise in one method, and 60% to 90% in a third decomposition similar to techniques economists use to break down pay gaps between groups.

Smokers Became More Disadvantaged Over Time, but That Doesn’t Explain the Trend

Researchers also asked whether smokers in later survey years were simply a different, higher-risk group than smokers in earlier years. The data told a partial story. By the later years, smokers were more likely to have less education, report worse health, and describe more emotional distress than nonsmokers, and those gaps had widened, while nonsmokers stayed steady on the same measures.

That might seem like an easy explanation, but it wasn’t. Models adjusting for education and self-reported health barely moved the age-specific mortality gap tied to smoking. For one age group, changing education levels added only about 0.3 percentage points to the gap, a small slice next to the roughly 6.6 percentage point gap already tied to smoking at that age. Smokers surveyed in later years may have carried more disadvantages than in the past, but those disadvantages aren’t what drove the gap’s rise.

Aging Populations Distort Risk Estimates Beyond Smoking

This lesson reaches past cigarettes. The paper’s authors argue the approach matters for anyone studying how a health risk changes over time, whether that risk is obesity, poor diet, or pollution exposure, since simple age adjustments aren’t enough when a risk factor’s effects differ sharply by age and the people carrying that risk are aging as a group.

There’s a practical angle too. Agencies like the Centers for Disease Control and Prevention use risk estimates from smokers and nonsmokers to forecast future smoking-related deaths, and projections could end up skewed if those estimates don’t separate age effects from real changes in the smoking-mortality relationship.

None of this means smoking is safe. It remains one of the biggest preventable causes of death in the country, with death rates for smokers in most age groups still more than double those of people who never smoked. The smoking-mortality gap hasn’t grown with each passing decade, despite appearances. What changed is simply who’s still lighting up, and how old they’ve gotten while doing it.

---

Disclaimer: This article summarizes findings from a peer-reviewed study and is intended for general informational purposes only. It is not medical advice. Anyone with questions about smoking, quitting, or personal health risk should talk to a licensed healthcare provider.

---