Dementia developed in about 17% of patients aged 90 years or older during a mean follow-up of 2 years in the prospective LifeAfter90 cohort study.
Investigators recruited Kaiser Permanente Northern California members aged 90 years or older who had been health system members at any time from 1964 to 1992 and were proficient in English or Spanish. Clinical evaluations occurred every 6 months from July 2018 through November 2024, with approximately one-third conducted remotely. The primary outcome was incident all-cause dementia based on physician diagnosis, the Clinical Dementia Rating, or the Functional Activities Questionnaire that determined cognitive impairment and clinically significant functional impairment.
Among the 1,120 participants initially available, the investigators excluded 96 with prevalent dementia and 219 who completed just one clinical evaluation. The median age was 92 years; 61% were female; 26% were Asian, 24% were African American or Black, 20% were Hispanic or Latinx, 28% were White, and 2% were from other racial or ethnic groups.
The investigators standardized dementia incidence to the 2020 US Census population aged 90 years or older. Cox proportional hazards models used age as the time scale, and Fine-Gray models accounted for mortality as a competing risk. Cox proportional hazards models were adjusted for age.
The age-standardized dementia incidence rate was 117 cases per 1,000 person-years, and 37% (n = 295) of the participants died over the study period. Female participants had an age-standardized incidence rate of 145 cases per 1,000 person-years compared with 80 among males. After accounting for age and competing mortality, female patients had an 89% higher dementia risk. In the Fine-Gray competing-risk model, Black participants had a 75% higher dementia risk compared with Asian participants. Hispanic or Latinx participants had the highest age-standardized incidence rate at 155 cases per 1,000 person-years. The investigators found no statistically significant differences in dementia risk by educational attainment.
APOE genotype data were available for 413 participants. APOE epsilon 4 carriers had an age-standardized incidence rate of 249 cases per 1,000 person-years compared with 112 among noncarriers, although the association with dementia risk was not statistically significant. APOE epsilon 2 carrier status was associated with lower dementia risk in the Fine-Gray model accounting for competing mortality. The participants with available APOE data were more likely to have attended graduate school and included a higher proportion of White participants compared with those without genetic data.
In sensitivity analyses, age-specific dementia incidence increased from 59 cases per 1,000 person-years among participants aged 90 to 94 years to 86 among those aged 95 years or older. Cumulative dementia incidence among those who survived without dementia to age 90 years was approximately 10% at 3 years and 18% at 5 years after accounting for competing mortality.
The findings could not be generalized to uninsured populations because the participants were long-term Kaiser Permanente Northern California members. The requirement for English or Spanish proficiency might also have limited generalizability to some Asian populations and other groups with low English proficiency. The investigators noted that volunteer study bias may have underestimated dementia incidence risk and contributed to selection bias. Because the outcome was all-cause clinical dementia and evaluations did not include biomarkers or neuropathologic data, they could not distinguish specific underlying dementia etiologies.
“Although much of the past research was conducted in young participants, our findings extend these observed disparities into the tenth decade of life,” wrote lead study author Hilary L. Colbeth, PhD, of the Department of Public Health Sciences at the University of California, Davis, and colleagues.
The study was supported by the National Institute on Aging of the National Institutes of Health and residual class settlement funds from April Krueger vs Wyeth, Inc. Full disclosures can be found in the study.
Source: The Lancet Healthy Longevity