Nearby greenery tracks lower dementia risk
A 16-year Stockholm cohort linked more vegetation within 100 metres of home to lower dementia incidence, but the observational result does not show that adding trees prevents dementia.
The 60-second version
In a Stockholm cohort, more satellite-measured vegetation within 100 metres of home was associated with lower dementia incidence, without proving causation.
Key points
- Researchers followed 2,993 dementia-free older adults for up to 16 years, and 477 developed clinically diagnosed dementia.
- Each interquartile-range increase in 100-metre NDVI was associated with HR 0.79, while 250- and 500-metre estimates were statistically inconclusive.
- NDVI measures foliage around an address, not whether greenery is accessible, attractive or used.
- The stronger estimate in women relied on an interaction p-value of 0.082 under a relaxed 0.10 threshold and requires replication.
- Residual confounding, reverse causation, exposure timing and the affluent Stockholm sample limit causal and global claims.
Verdict. The study makes nearby greenery a credible research and planning signal, not a proven personal prescription for preventing dementia.
A peer-reviewed Swedish cohort study found that older adults with more vegetation within 100 metres of home had a lower adjusted rate of incident dementia. The association survived adjustment for individual and environmental factors, but it does not prove that adding greenery prevents dementia.
CohortWhat the researchers followed
The analysis included 2,993 dementia-free adults aged 60 or older from Stockholm's Kungsholmen district. Over as long as 16 years, with a mean follow-up of 9.7 years, 477 participants developed clinically diagnosed dementia.
Dementia assessment used examinations by trained clinicians, review by a second physician and neurological adjudication when needed. Medical records and Sweden's cause-of-death register supplemented follow-up.
ExposureWhat satellite greenness actually measures
Landsat imagery supplied a vegetation index called NDVI for each residential address. The team calculated rolling exposure in 100-, 250- and 500-metre circles. NDVI detects green foliage; it does not measure park quality, access, safety, time outdoors or whether residents could see and use the vegetation.
ResultThe signal was narrow and close to home
| 100 m, previous 5 years | HR 0.79 (95% CI 0.64-0.96) per 0.18 NDVI increase; the primary adjusted association. |
|---|---|
| 100 m, 5-10 years earlier | HR 0.86 (0.70-1.06); statistically inconclusive. |
| 250 m, previous 5 years | HR 0.88 (0.71-1.09); statistically inconclusive. |
| 500 m, previous 5 years | HR 0.94 (0.76-1.15); statistically inconclusive. |
The 0.79 hazard ratio means roughly 21% lower relative hazard for one interquartile-range increase in NDVI. It is not a 21-percentage-point fall in absolute risk and not a comparison of a treeless home with a home beside a park.
The study found one credible proximity-specific association; it did not establish a universal dose of greenery that prevents dementia.
SubgroupsThe women-versus-men result needs restraint
The estimate was stronger in women, HR 0.69 (0.54-0.89), than in men, HR 0.93 (0.71-1.24). But the interaction p-value was 0.082 under a relaxed 0.10 threshold. That is a hypothesis-generating difference, not proof of benefit in women and no benefit in men.
The apparent difference between people with and without APOE epsilon-4 was not supported by a significant interaction test. Age, fine-particle pollution, road noise and nearby water also did not significantly change the association.
LimitsWhy association is not prevention
| Residual confounding | Greener neighbourhoods can differ in income, housing, mobility, social contact and baseline health despite extensive adjustment. |
|---|---|
| Reverse causation | Early cognitive or health changes may influence activity and residential decisions before a dementia diagnosis. |
| Exposure mismatch | Air-pollution data covered the period before baseline, not the same changing window used for every other exposure. |
| Generalizability | Participants came from a generally affluent urban district in Stockholm; other cities and populations may differ. |
| Multiple analyses | The clearest result appeared for one buffer and time window, while related larger-buffer and earlier-window estimates were inconclusive. |
The broader literature is mixed. A 2022 systematic review found no overall highest-versus-lowest greenness association with dementia and highlighted exposure error and unmeasured confounding. A 2025 review reported favorable associations across older-adult health outcomes, including Alzheimer's disease, but pooled heterogeneous observational evidence.
MeaningWhat the finding can support now
Urban planners have another reason to test and preserve usable greenery close to homes, especially in ageing neighbourhoods. Researchers now need replication across socioeconomic and urban settings, direct measures of access and use, and natural experiments or interventions that can address causality.
For individuals, nearby nature may help physical activity and wellbeing, but this study does not justify moving house or replacing established dementia-risk care. Continue evidence-based cardiovascular and lifestyle measures and seek clinical assessment for cognitive concerns.
Primary sourcesAlzheimer's & Dementia study: Precision residential greenspace, urban environmental exposures, and incident dementia among older adults in Sweden·PubMed record 42805935·PubMed Central full text PMC13619732·Crossref DOI record·2022 systematic review and dose-response meta-analysis of greenness and dementia·2025 systematic review and meta-analysis of green space and older-adult health·WHO guidelines on risk reduction of cognitive decline and dementia·Telegram post 1526 (CNSmydream)