Behavioural Science & Sociocultural Psychiatry · Paper A
Personality disorder types
Every personality disorder question is a sorting problem, not a spotting problem — the exam has already told you someone has an enduring, pervasive pattern; your only job is which of ten named patterns it is. ICD-10 and DSM-5 group the ten into three clusters by flavour, not cause: Cluster A reads as odd or eccentric, Cluster B reads as dramatic or erratic, Cluster C reads as anxious or fearful. Land the cluster first from the vignette's overall emotional temperature, then use one or two defining features to pick the specific type inside it — the same two-step move every time, on ten different names.
Imagine ten students sit the same personality questionnaire, and every one of them scores as having an enduring, unusually rigid way of relating to other people — a pattern that has been there since early adulthood and shows up everywhere they go, not just at a bad moment. A tutor sorting the ten scripts does not ask 'how severe is this?' first. She asks 'what FAMILY does this belong to?' — because the ten patterns cluster naturally into three groups by flavour, and once you know the family, picking the exact member inside it only takes one or two more clues. This page builds that sorting system from zero.
A personality disorder is a pervasive (shows up in nearly every area of life), enduring (present since adolescence or early adulthood, not a recent change) pattern of inner experience and behaviour that deviates markedly from what the person's culture expects, causes real distress or impairment, and is not better explained by another mental illness. Both of the two major classification systems — ICD-10 (the WHO's system) and DSM-5 (the American system) — recognise the same ten named types and group them into three clusters (a cluster is simply 'a set of types that share a family resemblance'), based on which flavour of behaviour dominates: Cluster A is odd or eccentric, Cluster B is dramatic or erratic, and Cluster C is anxious or fearful.
Here is the trap almost every exam candidate falls into: spotting one striking feature (suspicion, drama, anxiety) and jumping straight to a named type without first checking which of the OTHER types in the same cluster it might actually be. Three different Cluster A patients can all look like 'a loner' from a distance — but a genuinely indifferent loner (schizoid), a fearful loner who secretly wants friends (technically Cluster C, not A — proof that surface behaviour alone misleads even the cluster call), and an odd loner with strange beliefs (schizotypal) need three different names. The cluster narrows the field; it never finishes the job on its own.
Start from what a personality disorder actually is: not a single symptom, but a whole STYLE of relating to people and to oneself that has calcified into a fixed pattern. Some styles calcify around distrust and detachment (Cluster A: the person keeps others at arm's length, whether through suspicion, indifference, or outright strangeness). Some calcify around intense, unstable emotion and self-presentation (Cluster B: the person's inner world and relationships are dramatic, unpredictable, or self-focused). Some calcify around anxiety and control (Cluster C: the person manages fear either by avoiding people, clinging to them, or controlling their environment obsessively).
Every one of the ten named types is simply a more specific answer to 'HOW does this person keep others at arm's length / behave dramatically / manage their anxiety?' — which is why learning the cluster first, then the one or two features that separate its members, covers all ten with far less memorising than learning ten disconnected checklists.
Take three patients who each say 'I don't really have close friends.'
Patient one shows no distress about it at all — praised or criticised, his face and his answer stay the same. This is genuine indifference, so step one places him in Cluster A, and step two (indifference with no suspicion and no oddity) names him schizoid personality disorder.
Patient two says she desperately wants friends but is terrified they will laugh at her if she tries. This is fear, not indifference, so step one places her in Cluster C, and step two (avoids closeness from fear of rejection) names her avoidant personality disorder.
Patient three has no friends, believes he can read people's thoughts through walls, and dresses in clothes he says have 'special properties'. Step one places him back in Cluster A because of the oddity, and step two (indifference PLUS odd beliefs and eccentric behaviour) names him schizotypal personality disorder.
Same opening sentence, three different answers — because the exam is never testing the opening sentence, only the specific feature underneath it.
| Looks like | How to tell it apart |
|---|---|
| Schizoid vs avoidant | Schizoid = genuine indifference; avoidant = wants closeness but fears rejection |
| Schizoid vs schizotypal | Schizotypal adds odd beliefs or eccentric behaviour on top of the same withdrawal |
| Borderline vs histrionic | Borderline has abandonment-driven instability and self-harm; histrionic just wants attention |
| Narcissistic vs antisocial | Narcissistic exploitation serves ego/admiration; antisocial breaks rules with no remorse |
| Anankastic PD vs | Anankastic PD is ego-syntonic (feels correct); OCD's distress the person |
| Avoidant vs dependent | Avoidant withdraws from closeness out of fear; dependent clings to it from the same fear |
There is no blood test or scan for any of the ten types. Diagnosis rests on a longitudinal history: a trained clinician confirms the pattern is pervasive (present across relationships, work and social settings, not just one context), enduring (traceable back to adolescence or early adulthood, not a recent change following an illness or life event), and causes real distress or functional impairment, then matches the specific combination of features against ICD-10 or DSM-5 criteria for one of the ten named types (or, if the picture is genuinely mixed, records it as a mixed or unspecified personality disorder rather than forcing a poor-fitting single label). A single dramatic presentation, on its own, is never enough — the history has to show the pattern was already there.
Management is talking-therapy-led for every one of the ten types, and NICE is explicit that medication is not a treatment for the personality disorder itself.
For borderline personality disorder, NICE CG78 (2009) recommends structured psychological therapy — dialectical behaviour therapy or mentalisation-based treatment are the best-evidenced options — delivered by a team trained in the approach, with drug treatment reserved only for a genuinely separate co-occurring condition (such as a depressive episode), never prescribed to treat borderline traits or crises themselves.
For antisocial personality disorder, NICE CG77 (2010) centres on structured group-based cognitive and behavioural programmes, and places real weight on early intervention in childhood conduct problems, since the personality disorder itself becomes progressively harder to shift once fully established in adulthood.
Across the remaining eight types, the evidence base is thinner, but the same principle holds: the treatment targets the pattern of relating, not a chemical imbalance, so structured psychotherapy — not a prescription — is the first-line answer whenever a vignette asks 'what next?'
Whenever a vignette hands you a personality-disorder question, resist naming a specific type on your first read. Ask the cluster question first — odd, dramatic, or anxious? — and only then hunt for the one or two features that pick the exact type inside it. Before you move on, test yourself: could you explain to a friend why 'has no close friends' is true of at least four different personality disorders, and name the one extra fact that separates each pair?
Sort the cluster first by overall flavour — odd, dramatic, or anxious — then pick the specific type inside it from one or two defining features; the surface behaviour alone never tells you which of the ten names is correct.