This question is inspired by answers and comments to the question "Should a postdoc talk about his depression with his mentor?" on this website.
Suppose that an applicant to a graduate program suffered from some mental health issues during the completion of his previous degrees, and that this situation had a noticeable impact on his academic performance (for example, significantly lower grades or part-time status for one or several semesters).
Most graduate school applications have sections dedicated to providing explanations for such special circumstances, and it is of course strongly recommended that any applicant with unusual drops in academic performance should mention something about this in the special circumstances section. Otherwise, the admission committee will be left to guess what might have happened, which is likely to decrease the odds of being admitted (indeed, it seems a consensus that the more tangible information an admission committee has about an applicant, the more confident they can be about the fact that admitting him will be a sound investment).
Many answers/comments to the question "Should a postdoc talk about his depression with his mentor?" seem to recommend being very careful about revealing details on one's mental health conditions, given that there is still a lot of stigma attached to mental health conditions. However, in the case of PhD applications, not mentioning anything is not an option.
This then leads to the following question: If not making any mention of health issues is not an option (such as in PhD applications), how much details should one go into? To avoid stereotyping associated to any particular disease (such as depression=unreliable, learning disability=not smart) one could limit the description to "health problem", but could this lack of details be seen as suspicious and/or still a situation where the admissions committee will have to do guesswork?