Sex and All-Cause Mortality in the US, 1999 to 2019

While life expectancy in the US has increased among both sexes, a significant sex gap in mortality has been consistently observed during the past 2 decades.

To examine differences in mortality rates between males and females for overall mortality and the leading 9 causes of mortality (coded by the National Death Index as diseases of the heart, malignant neoplasms, chronic lower respiratory diseases, unintentional injuries, cerebrovascular diseases, Alzheimer disease, diabetes, influenza and pneumonia, and nephritis, nephrotic syndrome, and nephrosis).

This prospective cohort study included adults 20 years or older participating in the National Health and Nutrition Examination Survey (NHANES) cycles between 1999 and 2016. Data were linked to the National Death Index from the date of survey participation through December 31, 2019. Data were analyzed from July 16, 2024, to August 14, 2025.

Participation in the 1999-2016 NHANES cycles.

Cox proportional hazards regression was used to estimate male-to-female hazard ratios (MF HR) for overall and cause-specific mortality, adjusting for sociodemographic characteristics (eg, age, race and ethnicity), behavioral factors (eg, smoking, alcohol use), and chronic conditions (eg, diabetes, hypertension). The adjusted MF HRs were stratified by self-rated health, race and ethnicity, income, and educational level.

Among 47 056 participants (52.0% female and 48.0% male [weighted]), 12.9% (12.2% females and 13.6% males) were deceased at the end of follow-up. Males had a 63% higher risk of all-cause mortality compared with females, with the largest difference in mortality among individuals who died from heart disease (MF HR, 1.96; 95% CI, 1.72-2.24). The MF HR differed significantly across race and ethnicity for heart disease (range, 0.92 [95% CI, 0.36-2.35] for individuals of other race or ethnicity to 2.11 [95% CI, 1.80-2.48] for White individuals; P = .02 for interaction) and across income quartiles for mortality due to cerebrovascular diseases (range, 0.59 [95% CI, 0.25-1.38] for highest to 2.25 [95% CI, 1.31-3.86] for lowest income; P = .03 for interaction) and accidents (range, 0.57 [95% CI, 0.16-1.99] for highest income to 2.40 [95% CI, 1.03-5.58] for third highest income; P = .02 for interaction). No difference in the MF HR was found across self-rated health categories or education attainment levels. Information on covariates was largely collected by self-report, resulting in underreporting or overreporting due to social desirability bias. How these factors may have changed leading up to the time of mortality could not be accounted for.

In this cohort study of adult NHANES participants, after accounting for a wide range of risk factors, the sex gap in mortality remained for most causes of mortality, suggesting there may be intrinsic biological factors (eg, sex hormones, chromosomes, immune response) associated with sex differences in mortality. Further research should investigate the effects of sex-linked biological factors on mortality.