Late-onset schizophrenia

Definitions and terminology (International Late-Onset Schizophrenia Group consensus). Late-onset schizophrenia (LOS) = onset between 40 and 59 years. Very late-onset schizophrenia-like psychosis (VLOSLP) = onset at 60 years or above. The older European term for the latter is late paraphrenia. Both require the exclusion of dementia, delirium, mood disorder and organic causes.

Epidemiology. Onset of schizophrenia occurs after 50 years in about 13% of hospitalised patients with schizophrenia, after 60 in 7%, and after 70 in 3%. There is a marked female predominance in late-onset cases, in contrast to the male predominance of early-onset disease.

Risk factors. Female sex, sensory impairment — particularly deafness and visual loss, social isolation and living alone, never-married or childless status, premorbid paranoid or schizoid personality traits, migrant status, and a weaker family history of schizophrenia than in early-onset disease.

Clinical features — how it differs from early-onset schizophrenia. Persecutory delusions dominate, often with partition delusions (the belief that people, objects or radiation pass through walls, floors or ceilings) — highly characteristic. Multimodal hallucinations (auditory, visual, tactile, olfactory) are common. In contrast, formal thought disorder, negative symptoms and affective flattening are rare, and premorbid function, personality and occupational history are relatively well preserved. Cognitive impairment is milder and non-progressive.

Differential diagnosis and workup. Exclude delirium, dementia (especially DLB and Alzheimer's disease), psychotic depression, mania, delusional disorder, drug-induced psychosis and Charles Bonnet syndrome. Investigate as for any first-episode late-life psychosis: CBC, sodium (mEq/L), calcium (mg/dL), glucose (mg/dL), renal and liver function, TSH, vitamin B12, VDRL/HIV, urine culture, ECG, cognitive testing, audiometry, visual assessment and neuroimaging.

Management. Engagement and trust are the first therapeutic task, as insight is absent and patients resist the suggestion that they need treatment. Correct sensory deficits — hearing aids and cataract surgery — and reduce social isolation. Atypical antipsychotics are preferred over conventional agents, at much lower doses than in younger patients. Combine with psychosocial rehabilitation — social skills training, cognitive remediation, supported accommodation — plus assertive community treatment and case management, which greatly increase success. Optimally control vascular risk factors (obesity, hypertension, diabetes, dyslipidaemia), which are highly prevalent. Adherence is poor — consider long-acting injectables. Patients typically remain stable but symptomatic and functionally impaired, with higher mortality than earlier-onset schizophrenia, though sustained remission can occur even in chronic illness.

The ATLAS trial remains the only randomised, placebo-controlled evidence in VLOSLP (Lancet Psychiatry) — amisulpride 100 mg daily was superior to placebo in reducing psychosis symptoms over 12 and 24 weeks, with a large effect size and good tolerability despite a modest increase in extrapyramidal symptoms. Reviews through 2025 continue to cite it as the standard, while the 2023 AGS Beers Criteria permit antipsychotic use where the indication is schizophrenia or bipolar disorder, and xanomeline–trospium (2024) offers a non-dopaminergic alternative under study.