Urban park features and cardiovascular disease: a prospective analysis of 377,817 UK biobank participants

Greenspace benefits for cardiovascular disease (CVD) may depend more on quality than proximity, but feature-specific evidence is limited.

We linked objectively mapped park features to 377,817 urban UK Biobank participants and followed incident CVD (46,474 events) over a median 14.1 years. Network-based walk-distance catchments (400/800/1,600 m) were delineated for each home. We created an "any park" present indicator and feature-specific park indicators (e.g. tree canopy, water, cafes, toilets, attractions, sports) at each distance. Cox proportional hazards mixed-effects models adjusted for sociodemographic covariates, area deprivation, ambient natural landcover, and geographical clustering. Effect modification by sex, area-level deprivation, and age group (<65/≥65) was examined.

"Any park" was not associated with CVD, whereas feature-specific models indicated modest protection (≈3-6 % lower hazards). Tree canopy showed the strongest, distance-stable pattern with clearest support for ≥50 % at 800 m (HR 0.941, 95 % CI 0.912-0.992). Water features were protective at 400 m (HR 0.946, 95 % CI 0.907-0.986). Cafes were protective at 800 m (HR 0.971, 95 % CI 0.945-0.999), and 2+ public attractions at 1,600 m (HR 0.977, 95 % CI 0.957-0.997). Sports facilities at 800 m were associated with higher risk (HR 1.028, 95 % CI 1.004-1.057). Effect modification was limited: cafes at 800 m were more protective in high-deprivation areas; ≥50 % canopy at 1,600 m were more protective in women.

What's inside parks matters more than proximity. Higher tree canopy and nearby blue features appear broadly protective, while strategically placed amenities may support engagement.