Using 29 years of nationally representative US data, researchers mapped the exact ages at which physical, cognitive, and psychological decline accelerate, and found that the longer you live, the later your decline begins — but the steeper and more brutal it is once it starts. Centenarians hold function nearly a decade longer than people who die in their 80s and 90s, yet end life as the most disabled group of all.
For decades, the centenarian has been treated as biology’s lottery winner — someone who somehow sidesteps the diseases and frailty that claim everyone else. A large new analysis from the University of Southern California complicates that fairy tale. Centenarians, it turns out, are not exempt from aging at all. They simply delay its onset, and then decline faster than anyone.
The team, led by Erfei Zhao, drew on the Health and Retirement Study, following 6,851 Americans born in 1922 or earlier for up to 29 years, until almost all had died. By sorting people according to the age they reached — 73-79, 80-89, 90-99, and 100-plus — and using a technique called joinpoint regression to detect the precise age at which decline speeds up, the researchers could ask a question that disease-counting studies miss: not whether the oldest-old decline, but whenthe floor gives way.
The pattern is strikingly consistent across every functional domain. People destined to die in their seventies were already accumulating disability and disease on a steady upward slope from age 70. Centenarians, by contrast, stayed remarkably flat. Their first serious difficulty with independent living (managing money, medications, shopping) typically did not appear until age 94 — seventeen years later than the shortest-lived group. Their disability didn’t accelerate until age 93.7, and cognitive impairment held off until 91.
But the delay comes with a sting in the tail. Once a centenarian’s trajectory bends, it bends violently. Disability accumulation jumps roughly eight-fold, from 0.04 to 0.34 new impairments per year. The yearly rate of new dementia roughly sextuples, from about 1% to 6%. Centenarians end up the most functionally disabled group at death — the price of a long, healthy plateau is a short, steep fall.
The disease picture is more nuanced. Centenarians genuinely carry less illness, and once they have dodged cancer, diabetes, or lung disease in midlife, those conditions rarely appear later. But cardiovascular disease and stroke keep climbing into extreme age — vascular aging, it seems, can be delayed but never fully escaped. Crucially, in the oldest groups, disease counts plateau even as physical and cognitive function keeps collapsing — implying that what kills the oldest-old is the exhaustion of physiological reserve, not the arrival of new diagnoses.
The “Big Idea”: successful aging is not the absence of decline but its postponement. If interventions can push the tipping point even a few years later, the public-health payoff in disability-free years could be enormous.
Actionable Insights
This is an epidemiological study, not a trial — so it identifies correlates of a delayed tipping point, not proven interventions. Treat everything below as **association, not causation. The magnitudes, however, are large enough to be motivating.
The size of the prize. The gap between the shortest- and longest-lived groups is not subtle. First IADL (independent-living) limitation arrives at age 77 in the 73-79 group versus age 94 in centenarians — a 17-year postponement of functional dependence. The onset of accelerated disability shifts from ~84.8 (nonagenarians) to ~93.7 (centenarians), and accelerated cognitive impairment from ~84.3 to ~91.2 — roughly 7-9 extra years of preserved brain and body.
Rate differences while healthy. Before any acceleration, centenarians accrue disability at 0.04 ADL/year versus 0.15/year in the 73-79 group — a ~73% lower rate (rate ratio ≈ 0.27). Depressive-symptom prevalence rises 0.5%/year in centenarians versus 1.9%/year in the shortest-lived (rate ratio ≈ 0.26; ~74% slower). Pre-acceleration cognitive impairment grows ~1%/year versus 3%/year in octogenarians (RR ≈ 0.33).
What the authors point to as levers. The paper cites (from prior literature, not its own data) that genes explain ~50% of exceptional-longevity variance, leaving the rest to diet and lifestyle — specifically reduced caloric intake, high dietary quality, and sustained physical activity that preserves muscle strength and executive function. The honest take-home: the realistic, modifiable goal is compressing morbidity by pushing your personal tipping point later, chiefly by defending muscle, metabolic health, and vascular health — because cardiovascular disease is the one major condition that keeps rising even in the oldest-old.
Source:
- Open Access Paper: Finding the turning point in aging: a comparison of physical, cognitive, and psychological trajectories among centenarians and non-centenarians
- Institution: Leonard Davis School of Gerontology, University of Southern California (lead author now at the Jean Mayer USDA Human Nutrition Research Center on Aging, Tufts University).
- Country: United States.
- Journal: The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences (Oxford University Press, on behalf of the Gerontological Society of America).
- Data source: Health and Retirement Study (HRS), AHEAD cohort, 1993-2022.
Impact Evaluation
Most recent Clarivate Journal Impact Factor (2024, released June 2025): 3.8 (the journal has historically ranged ~4-6.6; it was 5.1 in 2022 and 4.3 in 2023).
The impact score of this journal is 3.8, evaluated against a typical high-end range of 0-60+ for top general science, therefore this is a Low-to-Medium impact journal on that absolute scale. Important context: within its own field, J Gerontol A is a leading, top-quartile (Q1) specialist journal in geriatrics/gerontology — a 3.8 here is far more prestigious than the raw number suggests against a general-science yardstick, because subfield citation norms are much lower than for general-science or oncology journals.