Dementia in men: +24% risk of death and more hospitalizations

A new study in JAMA Neurology (online August 11, 2025) analyzed data from 5,721,711 patients aged 65+ with newly diagnosed dementia in the United States (Medicare, 2014-2021). After adjusting for age, comorbidities, and social factors, men had a 24% higher risk of death and an 8% higher risk of any hospitalization than women. They also had higher risks of hospitalization for behavioral/neurodegenerative diagnoses and neuroimaging; men spent fewer days in hospice.

Background

• At the population level, dementia is more common in women. In 2025, approximately 7.2 million Americans aged 65+ will be living with Alzheimer's disease (the most common cause of dementia), and nearly two-thirds will be women. This is largely due to women's longer life expectancy and the accumulation of risk factors.
• However, the picture may differ “after diagnosis.” Across cohorts, men with dementia have been found to use acute hospital care more often, while women have higher rates of outpatient visits and drug therapy; the results vary across countries and health systems. This leaves open the question of who dies more often and who is hospitalized more often after diagnosis—and why.
• Demographic differences and risk factors by sex are known, but do not explain everything. Reviews show that women have a higher lifetime risk of dementia; the profile of modifiable factors (vascular, metabolic) and their contribution may differ between sexes. Large cohorts with consistent data are needed to separate the influence of sex from age, comorbidity, and socioeconomic factors.
• Why this work is important. We used the 2014–2021 National Medicare Cohort with up to 8 years of follow-up and included 5,721,711 people aged 65+ with newly diagnosed dementia and ≥1 year of prior fee-for-service Medicare. This scale and single source of billing/claims allow us to reliably compare mortality and service use in men and women after diagnosis, adjusting for age, race/ethnicity, poverty (dual eligibility), comorbidity burden, and access to resources.
• What was assessed. The primary outcome was all-cause mortality (Cox model). Secondary outcomes were all hospitalizations, nursing home stays, neuroimaging, physical/occupational therapy, and hospice care—that is, the full “trace” of service use after dementia diagnosis.
• Reader context: Medicare is the largest payer for older Americans; its fee-for-service segment pays for services at approved rates and generates a detailed billing database useful for the epidemiology of health care utilization. This makes the study's findings relevant for resource planning and policy for care of people with dementia.
• The gap this paper fills: While women are more likely to develop dementia, it was unclear whether the higher population contribution to mortality in women was due to greater incidence or greater mortality after diagnosis. This new paper specifically addresses this question in the context of the United States and current practice (2014-2021).

What exactly did they do?

• Design and scope: National Medicare cohort, up to 8 years of follow-up (2014-2021). Included people aged 65+ with a primary diagnosis of dementia (ICD-10) and ≥1 year of previous service in the traditional Medicare system. The primary outcome was the risk of all-cause mortality (Cox model); secondary outcomes were hospitalizations, stays in nursing facilities, neuroimaging, physical/occupational therapy, and hospice.
• Comparison subjects: Women (3,302,579) and men (2,419,132) with incident dementia. Gender was taken from the US Social Security registry.

Key results (with figures)

• Mortality. Unadjusted annual risk of death: 27.2% in men versus 21.8% in women. After adjustments, HR = 1.24 (95% CI 1.23–1.26) in favor of a higher risk in men.
• All hospitalizations. Unadjusted HR 1.13; adjusted HR = 1.08 (95% CI 1.08-1.09) for men.
• Specific use of care. The risk of hospitalization due to a neurodegenerative diagnosis/behavioral disorders is higher (HR ≈ 1.46), the probability of neuroimaging (≈ +4%) and hospice hospitalizations (≈ +8%) is slightly higher. Men spent fewer days in hospice (−8%) and nursing facilities (−3%).

What does this mean and why is it important?

• Gender “double gap.” At the population level, women bear a greater overall burden of dementia deaths (they are more common among those diagnosed), but once diagnosed, men die faster and are hospitalized more often. This is a signal to health systems that programs to reduce hospitalizations and deaths in men with dementia will have a disproportionately large effect.
• Resource planning: The higher rate of hospitalization in men places a burden on hospitals; however, shorter stays in hospices and nursing facilities may indicate access barriers, cultural attitudes, or delayed referrals—areas for organizational intervention.

Possible explanations (authors' hypotheses and context)

• Differences in comorbidity profiles, delays in seeking help by men, and behavioral and social factors may contribute to excess mortality and hospitalizations. The study does not draw causal conclusions but provides directions for targeted programs.
• Given that women are more likely to develop dementia, the authors draw the general conclusion that population-level mortality inequality is primarily due to higher incidence in women rather than higher mortality after diagnosis. This differentiates the goals of primary prevention (for women) and secondary/tertiary prevention (for men).

Restrictions

• US Medicare data: Transferability of results to other countries/systems requires caution.
• Retrospective analysis of administrative data: coding errors, underreporting of cognitive deficit severity and social support.
• Gender from the registers is a binary variable; gender aspects (care role, help-seeking behavior) were not measured directly.

Practical implications for families and physicians

• Men with newly diagnosed dementia are a high-risk group: early care plans, proactive work to prevent hospitalizations (control of behavioral symptoms, management of comorbidities, support for cuidadores) are useful.
• Women as a group with higher morbidity are the target audience of primary prevention programs for cognitive decline (control of vascular factors, active lifestyle, cognitive and social activity). These emphases are consistent with the authors' conclusions.

Source: Lusk JB et al. Sex Differences in Mortality and Health Care Utilization After Dementia Diagnosis. JAMA Neurology (online August 11, 2025), analysis of the 2014–2021 Medicare cohort, n = 5.72 million; doi: 10.1001/jamaneurol.2025.2236.