People coded in English GP records with mixed dementia were more likely than those coded with Alzheimer's disease to receive both an annual care plan and a medication review, but were also more likely to be prescribed drugs flagged as potentially inappropriate, according to a retrospective cohort study of 571,663 patients published in Age and Ageing on 4 April 2026. The study, by Charlotte Morris and colleagues at the University of Manchester and Newcastle University, used the Clinical Practice Research Datalink (Aurum) covering 1,490 English general practices from 1 January 2006 to 30 June 2024, and split dementia into eight coded subtypes, including two mixed categories.

Across the whole cohort, 72.1 per cent had a care plan coded and 79.4 per cent a medication review coded within 24 months of diagnosis. Compared with Alzheimer's disease, mixed dementias were the only subtypes more likely to receive a care plan (hazard ratio 1.29, 95 per cent confidence interval 1.26 to 1.32 for mixed including Alzheimer's and Lewy body dementia; HR 1.37, 1.32 to 1.43 for mixed non-Alzheimer's/Lewy body). All other subtypes were less likely. Mixed groups also led on medication reviews (HR 1.28 and HR 1.35), ahead of vascular dementia, Lewy body dementia and unspecified dementia. On the four prescribing indicators the direction reversed: compared with Alzheimer's, every other subtype, mixed included, was more likely to be prescribed high anticholinergic burden drugs, z-drugs, benzodiazepines or antipsychotics at least three times after diagnosis.

That gap matters to families because the review is the moment the drug list is supposed to be examined. The authors call the greater likelihood of potentially inappropriate prescribing in non-Alzheimer's dementias a novel finding, and note that new Alzheimer's treatments could widen the disparities they measured. NICE guideline NG97, published on 20 June 2018, warns that in dementia with Lewy bodies and Parkinson's disease dementia antipsychotics can worsen motor features and in some cases cause severe antipsychotic sensitivity reactions.

The fair counterpoint is that this is an observational study of coded records, and higher prescribing rates may partly reflect the symptoms that come with these diagnoses, including hallucinations, disturbed sleep and distress, rather than worse care alone. A coded care plan or medication review also tells researchers that an entry was made, not what was discussed or changed. The authors adjusted for age, sex, deprivation, ethnicity, comorbidity score and learning disability, with standard errors clustered by practice, but adjustment cannot remove all differences between the groups.