Health & function · Research Analysis
FINGER slightly improved cognitive-test scores. It did not prove dementia prevention
A two-year package of diet, exercise, cognitive activity and vascular-risk monitoring produced a small advantage on a cognitive-test composite. The trial cannot identify a winning component—and it did not show that dementia, disability or death was prevented.
Discovery, counterevidence, claim mapping, identifiers, funding, correction state and prose were reviewed before release.
Evidence box
Small cognitive-test difference for one combined program
- Study type
- Randomized controlled trial
- Studied in
- Finnish adults age 60–77 selected for elevated CAIDE dementia risk and cognitive performance at or slightly below age expectations
- Participants / sample
- 1,260 randomized; post-baseline cognition available for 591 intervention and 599 control participants
- Publication status
- Peer-reviewed primary reports and current registry records; correction status checked August 2, 2026
- Outcome type
- Two-year neuropsychological-test-battery composite; adverse events
- Development stage
- Completed human trial of a combined research program; not one standardized treatment
- Conflicts / funding
- Public and nonprofit research funding; investigator disclosures are retained in the primary report
- Our assessment
- A small package-level cognitive-test result—not dementia prevention, disability prevention, rejuvenation, or longer life
- Reporting confidence
- High for the bounded randomized result; low for component-specific or clinical-aging translation
Start here
Key takeaways
- The estimated difference in composite change per year was 0.022 standard units (95% CI 0.002–0.042).
- Diet, exercise, cognitive and social activity, and vascular-risk monitoring were delivered together; the trial cannot credit one component.
- The primary report did not test whether the program prevented dementia, disability, death, or shortened lifespan.
Keep the outcomes separate
What we know / what we do not know
What we know
- The combined program modestly outperformed general health advice on the prespecified cognitive-test composite over two years.
- Reported adverse events were more frequent in the intervention arm, mainly musculoskeletal pain.
What we do not know
- Which part of the package caused the difference—or whether all parts were necessary.
- Whether this test-score difference changes dementia, independence, mortality, healthspan, or lifespan.
Evidence at a glance
The answer changes with the outcome.
Rows preserve outcome boundaries; they are not a score, ranking, or recommendation.
| Outcome | Result | Boundary |
|---|---|---|
| Cognitive-test composite | Difference in change per year 0.022 (95% CI 0.002–0.042). | Small standardized package-level result. |
| Dementia and disability | Not established by the primary report. | A cognitive-test endpoint cannot be relabeled as prevention. |
| Safety | Adverse events 7% versus 1%, mainly musculoskeletal pain. | A lifestyle package should not be described as risk-free. |
The bottom line
FINGER randomized 1,260 at-risk Finnish adults to a two-year multidomain program or general health advice. The intervention group improved slightly more on a composite of cognitive tests: the estimated difference in change per year was 0.022 standard units (95% CI 0.002–0.042; p=.030). That is a statistically detectable, small test-score difference—not proof that dementia was prevented.
Why this matters
FINGER is often summarized as evidence that lifestyle prevents cognitive decline. The more useful reading is narrower: a resource-intensive combined program changed a test composite slightly in a selected population. Keeping that scope visible prevents a promising result from becoming a universal prescription.
What researchers did
Participants received diet guidance, exercise, cognitive and social activity, and vascular-risk monitoring as one package. Controls received general health advice. Because the elements were bundled, the randomized comparison cannot tell us whether one element drove the result.
What they found
Composite z-score change was 0.20 in the intervention group and 0.16 in control. Post-baseline cognitive data were available for 94% and 95% of participants. Twelve percent discontinued. Adverse events were reported for 7% versus 1%, most often musculoskeletal pain.
How strong is the evidence?
Randomization and high cognitive-assessment retention support confidence in the measured two-year test result. The small standardized difference, selected Finnish population, combined intervention, and lack of clinical endpoints sharply limit broader conclusions.
What this study does not show
FINGER does not show which component works best, that the program prevents dementia or disability, or that it reduces death or extends healthspan or lifespan. It also does not establish that a simplified commercial program would reproduce the protocol.
Safety, interpretation, and funding
This was a supervised research program, not a risk-free checklist. Musculoskeletal pain was the most common reported adverse event. Funding and investigator disclosures are available in the primary report; neither a research result nor this article is individualized medical advice.
What happens next
The most useful follow-up is longer outcome research that measures dementia, disability, independence and safety, plus trials that test which components are necessary and whether results transfer to other populations.
Primary sources
Sources, roles and limits
- Ngandu et al., two-year FINGER randomized trial
- Identifier
- PMID:25771249 · DOI:10.1016/S0140-6736(15)60461-5 · NCT01041989
- Role
- Primary randomized support
- Limitation
- Combined program and cognitive-test endpoint; no component-specific or dementia-prevention result.
- Kivipelto et al., FINGER protocol
- Identifier
- PMID:23332672
- Role
- Protocol and intervention definition
- Limitation
- A protocol defines planned methods; it is not a result.
- ClinicalTrials.gov record
- Identifier
- NCT01041989
- Role
- Registry identity and study history
- Limitation
- Registration does not validate efficacy or guarantee that every later claim was prespecified.
Connected topics
Continue through the knowledge system
Continue with real coverage
Keep reading
These are learning paths, not rankings, recommendations, or evidence that the stories support one another.
Another cognitive-aging trial with an important overall-versus-subgroup distinction
Hearing intervention did not slow cognitive decline across the full ACHIEVE trial
The three-year trial was null across its full cohort. Signals in higher-risk groups need confirmation and do not establish dementia prevention.
Evidence scopeHuman outcomes · randomized trial with subgroup signals
A vascular-risk intervention with hard clinical outcomes
SPRINT reduced cardiovascular events—not aging itself
Intensive treatment lowered cardiovascular events and mortality in eligible high-risk adults, with more hypotension, electrolyte problems and kidney injury.
Evidence scopeHuman clinical outcomes · large randomized trial
How mature exercise evidence changes by outcome
Exercise has strong evidence for function—not proof of slower biological aging
Human trials support mobility and physical function in defined populations. Lifespan and biological-age claims require separate evidence.
Evidence scopeHuman function · mature randomized evidence
Disclosures and history
- August 2, 2026 — source discovery, verification and counterevidence completed before prose drafting.
- August 2, 2026 — atomic claim map, article draft, identifiers, funding, correction state and inference boundaries reviewed.
- Corrections: none identified for the FINGER primary report or protocol as of publication.