Sicker at all ages? Two decades of changing disease histories and mortality in the United States.

BACKGROUND

Stagnating cardiovascular disease mortality and rising drug- and suicide-related deaths are key drivers of stalled US mortality improvements. However, prior research has largely overlooked disease histories preceding death, despite increasing multimorbidity. We examined changes in disease-history profiles in the US population and assessed how shifts in profile composition and within-profile mortality shaped mortality trends.

METHODS

We analysed adults aged ≥30 years (N = 510 372) from the IPUMS-NHIS (National Health Interview Survey) 1997-2018 linked mortality files (follow-up through 2019). Latent class analysis on the pooled sample across all ages and years identified disease-history profiles based on self-reported cancer, diabetes, cardiovascular disease, mental health, internal diseases, obesity, and physical limitations. Logistic regression was used to estimate profile-specific 5-year mortality over time, and counterfactual decomposition was used to quantify the contributions of compositional change and within-profile mortality improvement to overall mortality trends.

RESULTS

Five disease-history profiles were identified. Prevalence and 5-year death probability ranged from disease-free (62.4%; men 3.7% [95% confidence interval (CI) 3.5-3.8], women 2.4% [95% CI 2.3-2.5]) to high-burden multimorbidity (5.9%; men 17.8% [95% CI 16.8-18.8], women 12.6% [95% CI 11.9-13.3]), with substantial gradients across intermediate profiles. Profile composition varied substantially by gender, education, and race/ethnicity. The disease-free population declined by nearly 1% annually, replaced primarily by cardiometabolic and high-burden multimorbidity profiles. Most profiles experienced stagnating or increasing mortality after the mid-2000s. The decomposition shows that within-profile mortality improvement was the dominant driver of declining mortality; compositional shifts, mostly driven by the declining disease-free share, partially offset these gains.

CONCLUSION

Stagnating US mortality reflects both the growing prevalence of high-morbidity profiles and worsening or stagnating survival within them, highlighting the limited reach of prevention and disease-management efforts.