Background: We investigate the contribution of socio-behavioral factors to changes in US adult mortality over the period 1997–2019 using National Health Interview Surveys for the years 1997–2018 linked to death records through 2019. The variables studied include alcohol consumption, cigarette smoking, health insurance coverage, educational attainment, mental distress, obesity, and race/ethnicity. We evaluate the contribution of each socio-behavioral variable to mortality change by estimating the mortality risks associated with each variable in a hazards model and applying the risks to changes in the variable’s distribution. Results: When all variables are included in the model, we find that reductions in cigarette smoking and increases in educational attainment are the largest contributors to recent mortality improvements, accounting between them for 66% of mortality improvements. The contribution of educational attainment nearly doubles when variables that may be considered downstream to it are omitted from the analysis. In a secondary analysis, we compare two subperiods to investigate whether the variables can account for a widely observed slowdown in the rate of mortality reduction that occurred within the period of study. Rising levels of psychological distress, combined with very high risks associated with distress, contributed substantially to the slowdown. However, most of the slowdown remains unaccounted for. Contribution: We develop and apply a rigorous method to identify the role of many socio-behavioral factors in recent mortality change in the United States. Results highlight the role of a behavioral change (declines in smoking) and a broad social advance (educational expansion).

中文翻译：

---

导致美国近期死亡率趋势的社会行为因素（作者：Samuel Preston、Yana Vierboom、Mikko Myrskylä）


背景：我们利用 1997-2018 年全国健康访谈调查，调查了社会行为因素对 1997-2019 年美国成人死亡率变化的贡献，并与 2019 年的死亡记录相关联。研究的变量包括饮酒、吸烟、健康保险覆盖范围、教育程度、精神困扰、肥胖和种族/民族。我们通过估计与危害模型中每个变量相关的死亡风险并将风险应用于变量分布的变化来评估每个社会行为变量对死亡率变化的贡献。结果：当模型中包含所有变量时，我们发现吸烟的减少和教育程度的提高是近期死亡率改善的最大贡献者，占死亡率改善的 66%。当分析中忽略可能被视为下游的变量时，教育程度的贡献几乎翻倍。在二次分析中，我们比较了两个子时期，以调查这些变量是否可以解释研究期间广泛观察到的死亡率下降速度的放缓。心理困扰程度的上升，加上与困扰相关的极高风险，在很大程度上导致了经济放缓。然而，大部分放缓的原因仍然不明。贡献：我们开发并应用严格的方法来确定许多社会行为因素在美国近期死亡率变化中的作用。结果强调了行为改变（吸烟减少）和广泛的社会进步（教育扩展）的作用。