Diagnostic categories in mental health are revised, merged, split and retired between editions of the manuals that define them. The revisions look like scientific corrections but are better understood as decisions about where to place boundaries.
The manuals are consensus documents
Diagnostic manuals are written by working groups. Each group reviews literature, argues about criteria and produces wording that the wider body then approves.
That process resembles standard-setting more than discovery. The output is an agreed description that clinicians in different places can apply consistently.
Consistency is the primary goal, and it is a real achievement. Without shared criteria, two clinicians examining the same person can reach descriptions that cannot be compared at all.
But consensus wording carries the assumptions of the people at the table, and those change between generations of clinicians and researchers.
Thresholds are choices, not findings
Most criteria sets require a number of features present for a duration. Five features for two weeks, for instance, is a threshold rather than a natural boundary.
Distress does not arrive in discrete units. The features being counted vary continuously across a population, and the cut point is placed where clinicians judge intervention becomes warranted.
Move the cut point slightly and the number of people who qualify changes substantially, because most of a population sits near the middle of any continuous measure.
Working groups know this, which is why threshold changes are argued so heavily. A small edit to a duration requirement has consequences far beyond its apparent size.
Field trials test whether clinicians applying the new wording agree with each other. They test reliability of application rather than whether the boundary is in the right place.
Categories and dimensions pull in opposite directions
Clinical work needs categories. A person either meets the criteria for a service or does not, and a treatment protocol is written for a named condition.
Research increasingly describes the same phenomena dimensionally, as positions along traits that vary in degree. The two descriptions do not map onto each other neatly.
Manuals sit between these demands. They keep categories because the systems around them require categories, while incorporating severity ratings that acknowledge the underlying gradient.
Each revision moves the compromise slightly, which is one reason boundaries keep shifting rather than settling.
Comorbidity is a symptom of the categories
It is common for one person to meet criteria for several conditions at once. In the manual's logic these are separate diagnoses that happen to co-occur.
An alternative reading is that the categories have been cut too finely, and that overlapping diagnoses indicate one underlying difficulty described from several angles.
Working groups have responded in both directions across editions: sometimes merging related categories into a spectrum, sometimes separating a category into more specific ones.
Neither move settles the question, because the evidence supports both readings depending on which population is examined and which method is used.
Labels carry administrative weight
A diagnostic label does more than describe. It determines access to services, eligibility for accommodations at school or work, and in some systems whether treatment is reimbursed.
That gives revisions consequences outside the clinic. Retiring a category can remove the entitlement attached to it, even where the person's difficulties are unchanged.
Conversely, adding a category creates a route to support that did not previously exist, which is often precisely why advocates campaign for its inclusion.
Working groups therefore weigh administrative disruption alongside evidence, and revisions are sometimes deferred because the surrounding systems cannot absorb the change.
This is a normal feature of any classification embedded in payment and eligibility systems, and it is not unique to psychiatry.
Culture shapes what counts as a disorder
Some categories have been removed because the behaviour they described stopped being regarded as pathological. The evidence did not change; the judgement did.
Others have been added as forms of distress became recognisable, often after affected people described a pattern that clinicians then formalised.
Expression varies across cultures too. The same underlying difficulty can present predominantly as bodily complaint in one setting and as reported mood in another.
Manuals now include notes on cultural presentation for this reason, acknowledging that criteria written in one clinical tradition travel imperfectly.
What revisions mean for people already diagnosed
When a category is redrawn, existing patients are not usually reassessed. Records carry the old label, and clinicians translate informally between editions.
This produces long transition periods where two vocabularies coexist. Older notes, current services and research literature can each be using different versions of the same term.
People understandably experience the change as instability. A label that organised an explanation of their experience can be described afterwards as outdated wording.
Clinicians generally handle this by treating the label as an administrative key rather than an identity, and focusing conversation on the specific difficulties being addressed.
Why a stable classification is unlikely
Classifications stabilise when the thing being classified has clear natural joints. Chemical elements stabilised once the underlying structure was understood.
Mental health categories are defined by reported experience and observed behaviour, without a confirmatory test that sits underneath the description.
Until descriptions can be anchored to something measurable outside the description itself, boundaries will keep being argued, because there is nothing external to settle the argument.
The practical consequence is that a manual should be read as the current working agreement of a profession, which is a useful thing to have and a different thing from a map of nature.