Well, if you're approaching this as a hard-scientist that's not the case (for psychology).
Because "disorder" may be on a spectrum, because "negative" and "impact" may be on a spectrum. Conditions like depression follow a bell-curve, there's no clear line between depressed and not, and we can draw a line in the sand, but statistically there's no justification for where that line falls.
One way to try to draw a line is "the point at which a certain intervention is no longer effective."
For example studies find CBT therapy reduces depression scores for depressed people ~.8 standard deviations (for a while).
But why not try CBT therapy on happy people? Maybe it's even more effective on people who are already happy. Well the reason why not is likely just because of the clumsy nature of healthcare -- interventions are thought of as "treatments" for "conditions", even though that lens doesn't always make sense.
There’s a meaningful line in the sand for treatments with major side effects. CBT therapy may be “fine” for normal people, but the most effective treatment for depression is ECT which has major side effects. Including a ~1 in 50k chance of death.
Saying something is a bell curve distribution is an approximation, it doesn’t mean there’s actually a continuous function out to infinity and negative infinity.
When to resort to ECT is a subjective decision of doctor and patient. It is clear it should come after other treatments failed. But there is no clear line.
Spectrums extend continuously from normal to disorders. So if you believe depression is a spectrum you must also believe that no treatment is necessary for some people with depression. However if depression is a disorder there may be some cases that are on the margins that aren’t quite depression that still warrant some forms of treatment.
DSM V has depression as depressive disorders, but lists “Schizophrenia Spectrum” and “Autism Spectrum” so I invite you to consider what distinction for spectrum is being used.
Because "disorder" may be on a spectrum, because "negative" and "impact" may be on a spectrum. Conditions like depression follow a bell-curve, there's no clear line between depressed and not, and we can draw a line in the sand, but statistically there's no justification for where that line falls.
One way to try to draw a line is "the point at which a certain intervention is no longer effective."
For example studies find CBT therapy reduces depression scores for depressed people ~.8 standard deviations (for a while).
But why not try CBT therapy on happy people? Maybe it's even more effective on people who are already happy. Well the reason why not is likely just because of the clumsy nature of healthcare -- interventions are thought of as "treatments" for "conditions", even though that lens doesn't always make sense.