Good point, it makes me think any causal effect could easily be in the reverse direction, that married men are more likely required to exercise. This observatio

Good point, it makes me think any causal effect could easily be in the reverse direction, that married men are more likely required to exercise. This observation does not explicitly assess pre-marriage fitness. It does not do a before and after marriage on a cohort.