Firstly, by "exercise" I am strictly talking about short or long high-intensity workouts, e.g. lifting weights or cardio at the gym.
Occasional stretching and taking walks is not what I have in mind and not what most studies explore.
I've been watching this video discussing studies correlating drinking wine with longevity.
The long and short of it is that the J-curve of the old studies depicted in this graph failed to account for factors that may act as causal relationships on their own.
Basically, people who live good lives can afford to drink every now and then.
Since most studies on exercise also experience similiarly shaped J-curves (moderate exercise > no exercise > high exercise), shouldn't missing factors apply here, too?
Here are my suggestions for potentially missing correlative factors:
>People who started exercising because they are unhealthy
>(People who won't exercise because they're) unhealthy
>" mentally unwell
>" socially isolated (many people love exercising in pairs or groups)
>" poor (gym membership, private trainers, expensive equipment etc.)
>" having a stressful life