I guess there are tons of other confounders possible and quite difficult to control for (admittedly did not read the original study). I guess if you would take a green space and pour concrete on it just for a study to measure it against a controll, it would for sure also trigger police violence as the protesters will most likely also be quite unhappy.

Per abstract, the effect is stronger in poor regions (precisely "across five levels of social deprivation"). They did control for affluence.

There are studies in Japan that showed train stations with blue lights installed have lower instances of attempted suicides than stations without blue lights [1]. There are also studies that indicate time in nature is associated with lower stress levels and blood pressure, regardless of income [2].

[1] https://www.bbc.com/future/article/20190122-can-blue-lights-... [2] https://www.sciencedaily.com/releases/2024/02/240207120426.h...