I don't know if this claim is true or false but I'm pretty sure that doing a study on this kind of thing is an extremely fraught enterprise. Edit: more so than even simpler psychology studies, many of which are coming back with problems.

How does one determine that someone is really "feeling bad about feeling bad" or whether they're satisfying some expectation of the experimenter or doing something else? And moreover, cause and effect can be slippery, maybe those with more extreme problems tend to be caused to do the thing the experimenter thinks is a mistake rather what they do having an effect.