If anything, it appears the bans were associated with slightly worse mental health:

> In the full sample, the estimated postpolicy association was 0.65 (95% CI, 0.01-1.30; P = .05) for depressive symptoms, 0.21 (95% CI, −0.08 to 0.50; P = .16) for anxiety, 0.23 (95% CI, −0.12 to 0.59) for hostility, and −0.52 (95% CI, −1.08 to 0.03; P = .06) for mental well-being.

> Although a small number of estimates were suggestive of worse outcomes before multiple testing correction, none remained statistically significant after prespecified correction. Results were similarly null in gender-stratified analyses.