The stakes of that question are considerable. Lung cancer kills roughly 2,200 Norwegians every year, accounting for 19.3 percent of all cancer deaths in 2022 and ranking as the fifth leading cause of death overall in the country. More than 90 percent of lung cancers occur in current or former smokers, and the heaviest smoking burdens are concentrated in the older cohorts now reaching the peak ages for tobacco-related disease. Because prevalence figures alone cannot capture the underlying dynamics of when people start smoking, when they stop, and how many cigarettes they consume daily, the research team set out to deconstruct cross-sectional smoking data into its fundamental behavioral components.

To accomplish this, the researchers, led by Jacopo Di Silvestre and colleagues, drew on nationally representative surveys conducted by Statistics Norway from 1978 to 2024, covering people aged 16 to 79 years. After rigorous data cleaning that removed occasional smokers, inconsistent records, and surveys lacking cessation information, the final analytical sample comprised 78,282 respondents. From these anonymous self-reports, the team reconstructed individual smoking histories, including ages at initiation and cessation and the usual number of cigarettes smoked per day, grouped into ten consumption categories.

The analytical engine behind the study was an age-period-cohort, or APC, model, a statistical framework previously refined by the U.S. Cancer Intervention and Surveillance Modeling Network Lung Working Group and applied to survey data from the United States, Canada, Brazil, and Argentina. APC models disentangle three overlapping influences on smoking behavior: age effects that trace how smoking changes across the life course, period effects that capture changes common to all cohorts such as policy shifts, and cohort effects that distinguish one birth generation from another. The model treated smoking as a set of transitions between three mutually exclusive states, never smoker, current smoker, and former smoker, with initiation possible from age 8 onward and cessation modeled from age 15 onward, conditional on a hypothetical scenario with no mortality-related transitions.

The results paint a striking portrait of generational change. Daily ever-smoking prevalence, the proportion of people who had ever smoked daily, reached its zenith in the 1950 birth cohort, at 66 percent among women and 70 percent among men. From that peak, prevalence declined steadily across successive cohorts. Yet the decline was not uniform between the sexes. Women exhibited what epidemiologists call a catch-up pattern: their ever-smoking prevalence was lower than men’s in the 1950 and 1955 cohorts, exceeded male prevalence in the 1960 through 1975 cohorts, converged with male levels in the 1980 cohort, and then fell back below male prevalence in all cohorts born after 1985. This trajectory mirrors the classic model of the tobacco epidemic, in which smoking uptake spread later among women than men before the two sexes converged as the epidemic matured.

Smoking initiation probabilities followed a bell-shaped pattern across age in every cohort, rising through adolescence, peaking around age 19, and then falling steeply. The highest initiation probabilities occurred in the 1950 cohort, at 12 percent for women and 13 percent for men, and declined steadily in later generations. Notably, initiation probabilities were higher among women in the 1960 to 1975 cohorts, while men showed higher initiation in cohorts born after 1980. Cessation probabilities told an equally revealing story. In the earliest cohorts, the probability of quitting daily smoking increased only modestly with age, climbing from roughly 0 to 5 percent over four decades. But beginning with the 1965 cohort, both sexes showed markedly sharper increases in cessation, with probabilities rising from about 5 to 15 percent over just 20 years. Men initially quit at higher rates than women, but women surpassed men in the more recent cohorts.

Smoking intensity, measured as cigarettes per day among daily smokers, added a third dimension to the picture. Consumption remained relatively stable across the early cohorts, peaking at 12 cigarettes per day for women and 15 for men, before declining in more recent generations. Between the 1980 and 2000 birth cohorts, peak modeled consumption fell from approximately 13.5 to 10 cigarettes per day among men and from 11.5 to 7.5 among women. Men consistently smoked more intensively than women, and the age-related curve in intensity was broader for women, reflecting more gradual changes across the life course, and narrower for men, indicating faster swings in consumption.

The researchers attribute these declines to a combination of behavioral shifts and Norway’s aggressive tobacco control trajectory, which includes the 1975 Tobacco Act, the 1988 Clean Air Act, the 2004 nationwide indoor smoking ban, and the 2017 introduction of standardized packaging alongside smoke-free legislation extended to e-cigarettes. The authors also flag a potentially important confounder for the youngest cohorts: the rise of snus, the smokeless tobacco product that has gained substantial popularity in Norway since around 2000. Snus substitution may have contributed to declining cigarette initiation and higher cessation, particularly among cohorts born after 1980, though the authors caution that its causal impact remains uncertain.

The study’s methodological strengths lie in its use of nationally representative data spanning multiple decades and its application of an APC framework validated in other countries, supporting comparability across contexts. The team quantified uncertainty through a non-parametric bootstrap analysis with 1,000 resampled datasets, finding that uncertainty intervals were generally wider for earlier cohorts and older ages but did not materially alter the qualitative patterns. Still, the authors acknowledge limitations: occasional smokers, who comprised 8.6 percent of respondents, were excluded, potentially underestimating initiation and prevalence; cessation probabilities conflate complete abstinence with a transition to non-daily smoking; the model does not allow relapse from former to current smoking, which may overestimate sustained cessation in younger groups; and residual survivorship bias may persist at older ages because smokers experience higher mortality.

What elevates this study beyond descriptive epidemiology is its intended downstream use. The cohort-specific estimates of initiation, cessation, and intensity are designed to serve as key inputs for disease simulation models and health economic evaluations, the same modeling infrastructure that informed the 2021 U.S. Preventive Services Task Force lung cancer screening guidelines. As Norway evaluates a national screening program coupled with smoking cessation interventions, these granular behavioral parameters will allow modelers to project future lung cancer burden, estimate the cost-effectiveness of screening and cessation strategies, and quantify the long-term health and economic consequences of tobacco control policies. In a country where daily smoking among the youngest generations has become vanishingly rare, the smoking crisis of the mid-twentieth century is now a historical artifact, but its consequences will continue to unfold in the aging lungs of the 1950 and 1960 cohorts for decades to come.

Subject of Research: Birth cohort patterns of smoking initiation, cessation, and intensity in Norway from 1978 to 2024

Article Title: Age- and sex-specific smoking trends by birth cohort in Norway, 1978–2024: an analysis of cross-sectional survey data

Article References: Di Silvestre, J., Salgado, M. V., Jeon, J., Mok, Y., Holford, T. R., Vedøy, T. F., Pedersen, K., Burger, E. A., & Meza, R. (2026). Age- and sex-specific smoking trends by birth cohort in Norway, 1978–2024: an analysis of cross-sectional survey data. The Lancet Regional Health – Europe, 71, Article 101886. https://doi.org/10.1016/j.lanepe.2026.101886

Image Credits: AI Generated

DOI: Not provided

Keywords: Norway, smoking, birth cohort, lung cancer, tobacco control, smoking cessation, age-period-cohort model, cigarettes per day, public health, cancer screening, snus, epidemiology