Comparative Methodologic and Practical Considerations for Life Expectancy as a Public Health Mortality Measure

Abstract

Introduction Life expectancy is a public health metric used to assess mortality. We describe life expectancy calculations for US counties and present methodologic considerations compared with years of potential life lost before age 75 (YPLL-75) and premature age-adjusted mortality (PAAM), 2 commonly used length-of-life metrics. Methods We used death data from the National Center for Health Statistics for 2015-2017 and other health measures from the 2019 County Health Rankings & Roadmaps. We calculated life expectancy from birth at the county level using an abridged life table and the Chiang method of variance. Studentized residuals identified counties with discordant life expectancy and YPLL-75 or PAAM values. Correlations tested associations of life expectancy with key health measures (eg, smoking, child poverty, uninsured). Results Among 3073 US counties, life expectancy ranged from 62.4 to 98.0 years, with a mean of 77.4 years. Life expectancy was strongly and negatively correlated with YPLL-75 ( r = −0.91) and PAAM ( r = −0.95) at the county level. Life expectancy was also associated with other key health metrics, such as smoking, employment, and education rates, where an improvement in the health factor indicated improvement in the respective length-of-life measure. Counties with discordant life expectancy and YPLL-75 or PAAM values had differing age structures. Practice Implications Commonly used length-of-life metrics in population health settings are differentiated by methodological matters, such as computation complexity, data availability, and differential risk among age groups, especially among the very old or very young. The choice of metric should consider these factors, in addition to practical concerns, such as the communication needs of the audience.

Social Determinants of Remaining Life Expectancy at Age 60: A District-Level Analysis in Germany

“…In Germany, like in many other high-income countries (e.g., [31][32][33][34][35][36]), life expectancy at birth differs markedly among different parts of the country, i.e., among federal states, regions and districts [1,2,8,37,38]. A recent study on district-level life expectancy in Germany [2] demonstrated that average life expectancy at birth varied among Germany's 401 districts by 5.3 years for men and 3.9 years for women, based on the period life tables of 2015/17.…”

Study the forest, not only the trees: Environmental exposures, not genomes, generate most health disparities

“…Global-scale evidence indicates that environmental, not genetic, explanations are strongly implicated in populationlevel health disparities. For example, life expectancy is a common public health measure (Roubal et al, 2021), and despite a fairly constant genetic background, it shifts dramatically based on environment (The World Bank, 2019a; The World Bank, 2019b). For example, immigrants to the United States from nations with lower life expectancies than the United States experience increased life expectancy beyond the average US-born individual, though their US-born descendants do not (Argeseanu Cunningham et al, 2008;Mehta et al, 2016;Bastian et al, 2020).…”