I think this is just reacting to the headline though. The linked article isn’t engaging in such a conflation, it explicitly calls it out as a limitation of the study:

> Given the narrow demographic of the study, however, it's not possible to say whether these results would vary in non-White, non-American, LGBT+, and other populations.

And if you read the actual paper, it makes similar points in its “Limitations” section, albeit not as succinctly