Diagnosing mental health disorders is subjective on a variety of levels — just ask critics of the newly released DSM-V ; some of them claim that the addition of “new” diagnosis criteria over time threatens the credibility of the mental health diagnostic tool.

A new study published in JAMA Psychiatry suggests that general population surveys may be just as enigmatic when it comes to reporting mental health issues. Researchers looked at survey data taken from the Baltimore Epidemiological Catchment Area Study (ECA) and found that adults under report prior mental health diagnoses, while their recall of physical problems such as high blood pressure or type 2 diabetes remained fairly accurate over time.

The researchers analyzed a variety of mental health diagnoses among participants such as major depressive disorder, obsessive compulsive disorder, drug and/or alcohol dependence, etc. Prior studies have focused primarily on major depressive illness only.

The authors of the study, which included over 1,000 middle aged and older adults, assert that there could be several reasons for the inaccuracies, such as social stigma associated with mental illness, the fact that we tend to look at past events through rose-colored glasses, or that we forget many things as we age. However, they assert that this likely only accounts for a small number of recall variations given the fact that those surveyed had no problem remembering physical disorders.

They also emphasize the importance of cumulative reports gathered over time; they may give a more accurate assessment of mental health prevalence over the course of a lifetime.