The question the researchers set out to answer may sound paradoxical at first. If individuals with lower education experience higher mortality from nearly every cause of death, and if everyone ultimately must die from some condition, then what actually kills people across the social spectrum? One might expect that groups facing higher risks of, say, lung cancer or heart disease would show a heavier concentration of deaths from those specific conditions. Yet when Bergeron-Boucher and Strozza examined the actual distribution of causes of death, the answer was surprising. The relative proportions of deaths attributed to different diseases turned out to be remarkably similar across education levels in both countries.

To reach this conclusion, the researchers drew on comprehensive national data. For Denmark, they used the Causes of Death and Population registers, which cover the entire population. For the United States, they combined the Multiple Causes of Death dataset from the National Center for Health Statistics with population counts from the American Community Survey. Education was classified according to internationally standardized ISCED codes into three categories: low, corresponding to less than a high school education; medium, corresponding to high school completion; and high, corresponding to a college degree. Deaths were grouped into 15 broad categories reflecting the leading chapters of the International Classification of Diseases, tenth revision, plus one residual group for less common causes.

The analysis centered on women aged 40 and above in 2022 and relied on a demographic tool called the multiple-decrement life table, calculated separately for each education level. Life tables are prized in mortality research because they are age-standardized, removing distortions that arise when populations have different age structures. Crucially, the study went beyond the single underlying cause of death typically listed first on a death certificate. Because dying often involves a cascade of interacting conditions, death certificates also record contributing causes, meaning conditions that played a role in the death without initiating the fatal process. The researchers incorporated these contributing causes into their life tables, deduplicating any cause mentioned more than once on a single certificate to avoid double-counting. Confidence intervals were generated by bootstrapping the individual cause-of-death records within each education group, using 1,000 resampling simulations.

The gaps in longevity between education groups were substantial. Life expectancy at age 40 for American women was 39.4 years for the low-education group, 40.3 years for the medium group, and 44.6 years for the high-education group. In Denmark, the corresponding figures were 41.0, 44.6, and 46.9 years. In other words, highly educated women in both countries could expect to live roughly four to six years longer after age 40 than their less educated counterparts. Yet when the researchers compared the distribution of underlying causes of death across these groups, the differences were minimal. Formal chi-square homogeneity tests flagged statistically significant differences, but the researchers caution that with population-level data and enormous sample sizes, such tests can signal significance even for trivially small effects. They therefore turned to Cramér’s V, a measure of association strength ranging from zero to one, and found effect sizes below 0.06 in both countries, a level generally considered weak to negligible.

Where differences did appear, they were modest. Between high- and low-education groups, the share of deaths attributed to any given cause typically differed by less than three percentage points. Only a handful of exceptions crossed that threshold: diseases of the nervous system in the United States showed a gap of 4.4 percentage points, while in Denmark neoplasms differed by 3.5 points and respiratory diseases by 3.1 points. The researchers also computed the Gini-Simpson index, a diversity measure indicating how evenly deaths are spread across causes, and found nearly identical levels of cause diversity in every education group. The same held true for contributing causes: their distributions, their average number per death, and their diversity varied only minimally with education. Taken together, the pathways leading to death appear strikingly consistent across social strata, even as the timing of death shifts dramatically.

Behind this empirical pattern lies an elegant mathematical principle. If mortality risks rise proportionally across all diseases in a disadvantaged group, the ratios between diseases remain constant, and so does the cause-of-death distribution. The authors illustrate this with a simple example: imagine a population facing two diseases with annual death probabilities of 5 percent and 10 percent. If both probabilities double in a second population, death arrives earlier, but the split between the two diseases, one-third versus two-thirds, stays exactly the same. Reality complicates this tidy logic in two ways. First, mortality risks vary with age, and educational mortality differences are often larger at younger ages, which could shift all-cause distributions. Second, risks are not elevated proportionally across diseases. Indeed, when the researchers calculated age-standardized death rate ratios comparing low- and high-education groups, they found elevated risks for nearly every cause in the low-education group in both countries, the lone exception being diseases of the nervous system in the United States.

The key insight is that even these non-proportional elevations are not disproportionate enough to reshape the overall distribution. For some diseases, the low-education disadvantage is stronger; for others, weaker. But because everyone must eventually die from something, an elevated mortality rate for particular causes does not necessarily translate into a higher proportion of deaths from those causes. When the researchers compared the ratios of cause-specific shares of death between education groups, these ratios clustered tightly around one, much more so than the mortality rate ratios themselves. Causes where the educated advantage is less pronounced simply occupy a somewhat larger share of deaths among the highly educated, and vice versa. The statistical significance of these ratios was reached more often in the United States than in Denmark, likely reflecting larger sample sizes and death counts.

The findings align with what sociologists call the structural theory of health inequality. Within this framework, the socioeconomic conditions of social groups, including access to care, opportunities, and lifelong exposures, produce differences in health outcomes across virtually all diseases, even those with strong genetic components, while behaviors such as smoking or drug use act as mechanisms linking structural position to specific illnesses. The new study supports this interpretation but extends it: consistently higher mortality among lower-education groups across a wide range of conditions ultimately produces a similar cause-of-death profile. Social disadvantage, in other words, appears to create broad vulnerability to disease rather than redirecting people toward particular fatal pathways.

The researchers observed the same pattern in pre-pandemic years and among men, and the conclusion held when causes were regrouped in different ways, for example by separating heart disease and lung cancers into their own categories. This robustness suggests the result is not an artifact of classification choices. If cause-of-death distributions prove similarly consistent across other social determinants such as income or marital status, the authors note, it may point to a fundamental structure of mortality within each country, one that is shared across subpopulations and only modestly modified by behavioral differences, though cross-country comparisons would still reflect differences in death registration systems.

The policy implications are significant. Expanding educational attainment is known to raise life expectancy, and this benefit is likely to continue as education levels rise globally. But if education mainly postpones death rather than changing its cause, then the diseases populations face will remain largely the same regardless of educational investment. Health systems should therefore anticipate treating the same conditions in future decades, although the patients themselves will differ in age and educational profile. As populations become more educated, the emphasis may need to shift toward improving the treatment and management of familiar diseases at older ages. The work, supported by the SCOR Foundation for Science through the SCOR Chair in Mortality Research 2023–2026, points toward a future research agenda: identifying the mechanisms through which education and other social determinants shape when we die, knowledge the authors argue will be essential for designing interventions that not only extend life but also address the fundamental disease processes that determine how it ends.

This persistent mortality disparity invites critical scrutiny of the underlying factors beyond mere healthcare access. The study suggests that health outcomes are not solely dictated by technology availability but are deeply influenced by broader social determinants and public policy frameworks. According to the researchers, many of the avoidable excess deaths in the U.S. could be mitigated through the implementation of health and social policies that have been demonstrably successful in other high-income contexts. The implication is profound: mortality is a multifaceted issue, requiring interventions that transcend clinical care, encompassing a holistic approach to social welfare, preventive health, and economic equity.

An important consideration highlighted by the study involves the limitations and challenges inherent in multinational data comparability. Cross-country analyses often contend with variable death coding standards, data completeness issues, and reporting inconsistencies, all of which can introduce uncertainties in interpreting mortality figures. The researchers caution that these methodological nuances must be carefully weighed when drawing conclusions. However, even with these caveats, the overarching trend of excess mortality in the U.S. remains robust and calls for urgent attention.

Embedded within the mortality figure disparities are possible contributions from social inequalities pervasive in American society. Income inequality, in particular, emerges as a critical social determinant potentially exacerbating health outcomes. Whereas many peer countries have instituted social safety nets and health policies that mitigate the impact of socio-economic disparities on mortality, the U.S. appears less effective at bridging these gaps. The interplay between socio-economic status and access to preventive health services likely exacerbates the observed mortality differentials.

Furthermore, legislative frameworks and government policies are instrumental in shaping health trajectories on a population level. The investigation points to the necessity of evaluating the efficacy of existing U.S. legislation related to public health, social welfare, and economic support. Comparative policy analyses reveal that other high-income nations’ regulatory environments may afford protective factors that reduce mortality by addressing root causes such as poverty, education inequalities, and healthcare access barriers.

From a methodological standpoint, the study’s use of repeated cross-sectional data allows for temporal tracking of mortality trends, thus enabling identification of persistent patterns as well as emergent issues. This approach provides a dynamic perspective on mortality, contrasting with static snapshots often seen in previous research. The longitudinal design strengthens the validity of findings by demonstrating consistent excess mortality over more than two decades.

Technological advancements in information processing and data analysis have empowered this kind of rigorous cross-national mortality research. Utilizing sophisticated statistical tools and harmonizing international datasets requires meticulous attention to detail and an interdisciplinary approach. The study exemplifies the integration of epidemiology, demography, social sciences, and health informatics, setting a benchmark for future comparative mortality research.

The study also emphasizes human health implications, illustrating that mortality metrics are not just abstract statistical outcomes but reflections of real-world lived experiences. High death rates in the U.S. translate into millions of premature deaths, diminished quality of life, and substantial socio-economic costs. Public health stakeholders, policymakers, and society at large must reckon with these findings as a clarion call for systemic change.

Moreover, this research aligns with growing recognition within the scientific community that tackling mortality disparities requires inclusive and equitable public policy initiatives. Health interventions cannot be siloed within clinical settings but must integrate social, economic, and environmental factors. Initiatives such as expanding access to affordable healthcare, addressing food insecurity, improving education, and reducing income inequality could collectively shift the trajectory of U.S. mortality rates.

In conclusion, the comprehensive analysis conducted by Jacob Bor, PhD, and colleagues provides an urgent narrative about mortality in the United States compared to its high-income peers. Despite technological prowess, social and policy inadequacies perpetuate unacceptable mortality differentials. The study paves the way for developing evidence-based policies that, if adopted, could save hundreds of thousands of lives annually. It is a critical scientific contribution demanding both national introspection and international collaboration to close the mortality gap.

Subject of Research: Cross-national mortality comparison and health policy implications
Article Title: Not specified in the provided content
News Publication Date: Not specified in the provided content
Web References: Not provided
References: DOI: 10.1001/jamanetworkopen.2026.6147
Image Credits: Not provided
Keywords: Mortality rates, United States population, Income inequality, Legislation, Public policy, Data analysis, Human health