Why Dementia Risk Persists After Age Ninety, Globally

It is easy to assume that reaching your 90th birthday without cognitive decline means the worst is behind you. The data from a new long-term study upend that assumption and show a more complicated reality: risk remains, and it is unevenly distributed.

Dementia risk can remain substantial even after age 90, with new research revealing persistent differences linked to sex, race, ethnicity, and genetics. 

Tracking the tenth decade: the LifeAfter90 effort

The LifeAfter90 study was launched precisely to fill a gap many researchers had ignored. Most dementia research focuses on people in their late 60s or 70s. Very few studies follow those who enroll after 90, and even fewer include racially and ethnically diverse cohorts. LifeAfter90, run by UC Davis Health in partnership with Kaiser Permanente, follows members who joined the study at age 90 or older while showing no signs of dementia. Participants undergo cognitive evaluations every six months so researchers can catch new cases early and map how risk changes in extreme old age.

The recent analysis, published in The Lancet Healthy Longevity, used medical records from more than 800 participants whose median age was 92. Some of those participants had been Kaiser Permanente members for decades, giving researchers access to health histories that sometimes stretch back to the 1960s. That long window matters because dementia often reflects exposures and health states from many years earlier.

Rachel Whitmer, professor, Departments of Neurology and Public Health Sciences, UC Davis Health. 

Clear patterns: sex, race and ethnicity still matter

One of the study’s most striking findings is that women who entered the study dementia-free were about twice as likely as men to develop dementia during follow-up. That mirrors patterns seen in younger elderly groups and suggests the sex gap does not simply vanish for survivors in their nineties.

Race and ethnicity also shaped outcomes. Black participants faced substantially higher incidence compared with Asian participants, with both Black and Hispanic groups showing higher dementia rates than white and Asian participants. These disparities echo earlier research on younger cohorts, but the persistence of the pattern past age 90 is notable. It implies that social, medical, and environmental forces that drive disparities across midlife and early old age still echo into the tenth decade.

Why would disparities persist so late in life? The dataset allows researchers to look back into midlife exposures and medical histories. Hypertension, diabetes, access to consistent health care, lifelong socioeconomic differences, and even regional patterns of care all leave fingerprints. Those factors influence brain health over decades. The LifeAfter90 analysis suggests that surviving to 90 without dementia does not erase the cumulative effects of lifetime risk and resilience.

Genetics still plays a role, but not uniformly

APOE, a gene associated with Alzheimer’s disease, remains an important piece of the puzzle. APOE2 is known to be associated with lower Alzheimer’s risk, and that protective effect showed up clearly in the LifeAfter90 cohort. Participants carrying APOE2 had roughly 60 percent lower risk of developing dementia even after age 90.

APOE4, the variant most commonly linked to increased late-onset Alzheimer’s risk, produced a more nuanced picture. When the cohort was analyzed in aggregate, APOE4 did not significantly raise dementia incidence. But when researchers examined subgroups defined by sex and ethnicity, patterns emerged: APOE4 appeared to pose greater risk for men in the study and was associated with nearly double the risk for Black participants. In short, genetic risk is not the same across every group, and population-specific effects can change how we interpret risk estimates.

Those findings help explain a recurring problem in genetic research: risk models developed in largely white study populations often fail to generalize. If genetic influence interacts with sex or ancestral background, then diverse and inclusive sampling is not optional. It is essential.

The mystery of resilient brains

Perhaps the most intriguing question the study raises is why some people remain cognitively healthy despite carrying known risk markers. Several LifeAfter90 participants had conventional midlife risk factors such as high blood pressure or high cholesterol yet they reached their nineties without cognitive impairment. A number of APOE4 carriers remained free of dementia as well.

Resilience might come from multiple sources. Better access to continuous medical care and treatment during midlife could blunt the long-term impact of vascular risk factors. Social and behavioral elements matter too. Lifelong education, engaging social networks, regular physical activity, and mentally stimulating occupations are associated with preserved cognition. Biology likely contributes as well: some people may carry protective genetic variants or favorable inflammatory and metabolic profiles that delay neurodegeneration.

• Medical care across decades
• Lifestyle and cognitive engagement
• Biological protective factors

Figuring out which of these elements matter most could open routes to interventions that help more people preserve cognitive function into very late life.

Expert Insight

Dr. Elena Morales, a fictional neurologist and aging researcher with a long career studying late-life cognition, reflects: "This study forces us to rethink survivorship. Reaching 90 without symptoms is a remarkable achievement, but it does not grant immunity. We need targeted risk-reduction strategies that account for sex, ancestry, and genetic background. At the same time, studying people who remain well may reveal protective mechanisms we can amplify."

Her comment underscores a pragmatic implication: prevention messages and clinical guidance should not assume safety after a fixed age. Clinical counseling may need to remain personalized and evidence-based even for those in their tenth decade.

Implications for research and policy

LifeAfter90 demonstrates the value of long-term medical records and diverse cohorts. The study’s design, with frequent cognitive testing and decades of prior health data for many participants, allows investigators to connect midlife exposures with outcomes that appear very late. For policymakers and funders, the takeaways are clear. Investment in inclusive longitudinal studies is critical. So is attention to social determinants of health that generate baseline disparities.

Clinicians should also take note. The message is not alarmist. It is precise. Certain groups remain at higher relative risk well into advanced age, and genetic risk can vary by sex and ancestry. That knowledge should inform conversations about prevention, monitoring, and care planning.

Conclusion

Reaching 90 without dementia is an extraordinary milestone. It does not, however, erase the uneven distribution of risk shaped by biology, lifetime health, and social factors. The LifeAfter90 study offers a richer, more inclusive picture of brain aging and points toward next steps: discover what protects resilient brains and translate those lessons into equitable strategies to preserve cognition for more people.