Objective Delusions of misidentification are complex, rare delusions in which patients believe that the identity of a person, object, or location has been altered or replaced. Previous research has identified distinct neuropsychiatric correlates of specific misidentification syndromes, as well as evidence of their variation based on the aetiology of the psychosis. However, no study has sought to examine all types of misidentifications and their underlying pathoaetiology collectively. The current study aimed to investigate the clinical, neuropsychological, and neuroanatomical features of delusional misidentification syndromes in patients with a primary psychiatric disorder compared to those which are secondary to an underlying medical cause.

Methods A patient-level meta-analysis of published cases of delusional misidentification syndromes was conducted. PubMed, PsychINFO, Medline, Embase, and Global Health were searched. Demographic characteristics, diagnosis, neuropsychiatric features, cognitive ability, neuroimaging, and treatment information were extracted. Cases were classified based on the aetiology of psychosis (primary or secondary). Chi-square tests were used to compare the frequency of each feature across the two aetiologies, and odds ratios (OR) were calculated.

Results A total of 422 cases were identified (214 females; median age 44 years), comprising 258 cases of Capgras Delusion, 93 cases of Fregoli Syndrome, and 71 cases of more rare misidentification subtypes. Delusions were attributed to a secondary psychosis in 190 cases (45.0%). Mixed or unknown aetiology was reported in 19 patients (4.5%). Patients with secondary psychosis were significantly less likely to experience misidentification delusions involving people (e.g. Capgras or Fregoli delusions) but were significantly more likely to misidentify places than those with primary psychosis (OR=0.79, 95% CI=0.03, 0.23; p<.001). Additionally, patients with secondary psychosis were more likely to experience cognitive impairment (OR=2.99, 95% CI=1.65, 5.41; p<.001), specifically impairments of memory (OR=4.17, 95% CI=2.14, 8.14; p<.001) and executive functioning (OR=4.02, 95% CI=2.00, 8.10; p<.001). Conversely, they were less likely to report additional neuropsychiatric symptoms, including persecutory features (OR=0.22, 95% CI=0.13, 0.36; p<.001) and hallucinations (OR=0.17, 95% CI=0.08, 0.33; p<.001). Neuroimaging abnormalities were more frequent in those with secondary psychosis (χ2=75.62, df=1, p<.001), although bilateral involvement was more common in primary psychosis (χ2=4.34, df=1, p=.037). No differences in treatment outcomes were observed.

Conclusions In the largest study investigating delusional misidentification syndromes to date, differences in neuropsychiatric presentation were observed based on aetiology. Findings further our understating of the clinical profile of delusional misidentification syndromes secondary to an underlying medical cause which may ultimately lead to advances in earlier detection and treatment.