The NPR article seems to confuse causation and correlation.

The actual paper doesn't, and merely implies correlation. Which is fascinating (and well-known) and might still prove useful in one way or another.

[1] https://jamanetwork.com/journals/jama-health-forum/fullartic...