In the almost 60 years since the publication of the landmark US Surgeon General’s report on the health consequences of smoking in 1964,1 even stronger evidence on the health risks of smoking has accrued from a wealth of research.2 However, there is a dominance of studies from high-income countries and hence uncertainty on whether risk estimates are generalisable to countries at different economic levels. In The Lancet Global Health, Thirunavukkarasu Sathish and colleagues3 have, for the first time, to our knowledge, studied systematically whether risks associated with smoking differ across country income categories.3

By analysing data from three large epidemiological multi-country studies and following a standardised approach, the authors found that risk estimates of adverse outcomes for smokers in high-income countries (HICs) were higher than for smokers in middle-income countries (MICs) and low-income countries (LICs). For example, the hazard ratio for current versus never smoking was 2·58 (95% CI 1·97–3·37) in HICs, but only 1·51 (1·40–1·62) in MICs and 1·22 (1·11–1·34) in LICs. Notably, never smokers were defined as those who had never used any tobacco products, and these risk estimates were adjusted for several potential confounders, including smoking history. Similar patterns were found for several disease specific outcomes (cardiovascular events, cancers, and respiratory diseases), and these were remarkably consistent in the three studies across diverse study designs, decades, and sets of countries, which substantiates the evidence.

Notably, Sathish and colleagues also further explored potential explanations for the lower risk estimates in MICs and LICs versus HICs. One potential explanation is the higher exposure of toxicants in smokers from HICs, leading to a greater excess risk associated with smoking. Another possible reason is a higher exposure to second-hand smoke (SHS) among never smokers in MICs and LICs, as a higher background risk in the reference group would result in a lower excess risk associated with smoking. Indeed, the analysis of mean urinary total nicotine equivalent (TNE) concentrations, which is a robust measure of nicotine intake and was available in a substudy, yielded clear gradients of TNE concentrations in current and never smokers by country income category. Although TNE concentrations in current smokers were highest in HICs and lowest in LICs, the opposite was the case for TNE concentrations among never smokers. Furthermore, when comparing current and never smokers within country income categories, large differences in TNE concentrations were observed in HICs, but only small differences in LICs. Although reasons for these patterns remain somewhat speculative, greater toxicant exposure from smoking in HICs as well as higher SHS exposure in MICs and LICs are indeed plausible explanations.

The study’s findings imply that region-specific assessments of risks associated with tobacco are required to obtain valid regional or global quantifications of the tobacco-related disease burden. However, beyond these methodological implications noted by Sathish and colleagues, the study also has important global health implications. First, although it is possible that smokers in MICs and LICs are less likely to fall seriously fall and die due to smoking than smokers in HICs, the fact that never smokers are so much more affected by SHS in these countries is reason for concern, also given that children are often particularly exposed.4 Second, it is crucial to note that the public health impact of smoking in a given region is not only determined by the size of relative risks, but also by the overall number of people exposed. With high numbers of smokers, the associated disease burden can still be tremendous even if the relative risks are small. In this context, a recently published Global Burden of Disease study5 observed an alarming trend. Although smoking prevalence declined significantly over the past 3 decades in the vast majority of the included countries, these decreases in smoking prevalence were offset by the “demographic force of population growth” in most of the countries, resulting in constant or increasing absolute numbers of smokers in these countries over time.5

The largest relative increases in absolute numbers of smokers were observed in North Africa and the Middle East and in sub-Saharan Africa. Since the full health consequences of smoking are usually delayed by decades,6 it is likely that the tobacco-associated disease burden will increase in the years to come in such countries, including many MICs and LICs whose health systems might not cope well with this challenge.7^, 8

Hence, there is an urgent need to strengthen tobacco control globally. The WHO global report⁹ on trends in prevalence of tobacco use in 2000–25 describes the remarkable achievements, but also the shortcomings, in curbing the smoking epidemic following the adoption of the WHO Framework Convention on Tobacco Control in 2003, hence further strengthening the implementation of the convention must remain a global health priority. Although it is indisputable that tobacco control efforts can achieve substantial inroads to reducing smoking, it will take years to decades until today‘s efforts will be reflected in decreases in the disease burden. Therefore, action should be taken sooner rather than later. Any inaction will have devastating consequences in the years to come, particularly in MICs and LICs.