Satizabal · The New England journal of medicine 2016 · prospective cohort study · n=5205

Incidence of Dementia over Three Decades in the Framingham Heart Study.

Cited 1055 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective longitudinal cohort study assessing temporal incidence trends over multiple epochs.

PubMed 26863354 · doi:10.1056/NEJMoa1504327 · record verified 2026-08-28

What was done

Participants aged 60 years or older (n = 5,205) from the Framingham Heart Study under surveillance for incident dementia since 1975 were evaluated across four successive epochs spanning three decades (late 1970s/early 1980s, late 1980s/early 1990s, late 1990s/early 2000s, and late 2000s/early 2010s). Cox proportional-hazards models adjusted for age and sex determined 5-year cumulative incidence rates for each epoch. Investigators analyzed interactions with epoch for age, sex, apolipoprotein E ε4 status, and education level, alongside the effects of vascular risk factors and cardiovascular disease.

What was found

The 5-year age- and sex-adjusted cumulative hazard rates for dementia per 100 persons were 3.6 in the first epoch, 2.8 in the second, 2.2 in the third, and 2.0 in the fourth. Compared with the first epoch, dementia incidence declined by 22% in the second epoch, 38% in the third epoch, and 44% in the fourth epoch. The risk reduction was observed only among participants with at least a high school education (hazard ratio, 0.77; 95% confidence interval, 0.67 to 0.88). While most vascular risk factors (excluding obesity and diabetes) and the dementia risk linked to stroke, atrial fibrillation, or heart failure declined, these factors did not fully explain the reduction.

Why it matters

This study demonstrates a 44% decline in age-specific dementia incidence over 30 years in a community cohort, showing that dementia risk is not fixed and is modifiable over time.

Limits

The Framingham Heart Study population is predominantly white and of European descent, limiting generalizability to more diverse populations. The observational design cannot definitively establish the causal mechanisms underlying the decline.

Cited by