Article Text

15 Capgras delusion: a meta-analysis of case reports in the english language
  1. Charalampos (Charis) Pandis,
  2. Norman Poole


Objective The Capgras delusion has an almost totemic status in the non-scientific understanding of psychopathology. The original paper describing Mme M’s horrific enmeshment in a logarithmic expansion of doubles reads more like a short story by Edgar Allan Poe than a medical case history. Traditionally was associated with functional psychiatric disorder, there has been accumulating evidence of neuropsychiatric aetiologies including dementias, toxic-metabolic conditions and localised structural lesions (Berson, 1983). Enoch’s review also said 25%–40% of cases occur in the context of diverse “organic disease”, and that neuropsychological abnormalities imply predominantly right and bilateral cerebral dysfunction. Due to its rarity, little is known of the prognosis and response to treatment or whether there are systematic differences between the Capgras delusion associated with functional and organic disorders. This review therefore aims to collect all case reports in the English language on the Capgras delusion to better understand the clinical profile of this uncommonly important psychopathological phenomenon. If the theorists have the clinical features wrong this carries serious implications for the robustness of their models.

Method Search and selection strategy for studies: All papers published in English since Capgras’ case report was published in 1923 were sought. We searched Medline (from 1950), PsychINFO (from 1806), Embase Classic (from 1947 to 1979), and Embase (from 1980 to January 2009). We used the key words ‘Capgras delusion’ OR ‘Capgras disorder’ OR ‘misidentification disorder(s)’ OR ‘misidentification syndrome’ AND ‘case report(s)’ AND ‘meta-analysis’ OR ‘systematic review’ OR ‘literature review’. The titles and abstracts online were reviewed by the authors and obtained copies of all publications that appeared relevant to the study question. The reference lists of all these publications were then hand searched for additional relevant studies. Papers were excluded if presented any other misidentification delusion but Capgras delusion. Data extraction and analysis: Two authors (N.A.P and C.P.) independently extracted all the relevant data from the papers and where they disagreed a third reviewer (N.A.) was consulted. The following information were collected systematically and entered into a Microsoft Excel database: title, journal, year of publication, age, gender, marital status, co-morbidities (medical and psychiatric), alcohol, substances, family psychiatric history, ilius relationship, animate/inanimate double, duration of delusion, comorbid psychiatric symptoms, other types of misidentification delusions, neuropsychological investigations and results, neuro- imaging results (type and location of lesion), EEG. Not all parameters were available in every case reported. The non-reported parameters were not included in the analysis. When information was incomplete we attempted to contact the authors. In addition to surveying the clinical features of the Capgras’ delusion with the use of frequency statistics, we sought to explore whether these differed between cases with an organic and those with a functional aetiology. All the variables were then associated with organic versus functional by cross-tabulation (Pearson’s chi-squared). Fischer’s exact test was used when one or more cells in the 2 × 2 table contained five cases or fewer. Odds ratios (ORs) and 95% CIs were calculated on Microsoft Excel for Windows and a p-value less than 0.05 was considered significant.

Results We identified 212 papers, 179 of which were found to meet the study inclusion criteria. The papers identified 252 cases presenting with the Capgras’ delusion. Of these, 143 reported the delusion occurring in the context of so-called “functional” psychiatric disorder with the remaining 109 having an identified an “organic” aetiology. The first case was reported in 1963 and the most recent in 2015. Male to female ratio was 1:1.4. The median age was 45 and the age of cases ranged from eight to 94 years old. Interestingly, there was no difference in either the median age or range of ages between the “organic” and “functional” groups. The breakdown of primary psychiatric disorders displays the most prominent diagnosis to be schizophrenia (89), followed by organic psychosis (45) and dementia (28). Interestingly there was no statistical significant difference in gender between organic and functional disorders as well as no difference in treatment response. The ilius relationship for the spouce and inanimate objects was statistically significant for the organic causes of the delusion (1.15–3.30* and 1.27–5.10**, 95% CI) whereas for a parent and multiple imposters was in favour of functional causes of the delusion (0.23–0.83**and 0.32–0.91*, 95% CI). Associated symptoms of aggression and homicide were statistically higher for the functional causes of the delusion (0.29–0.86* and 0.02–1.11*, 95% CI). With regard to perceptional abnormalities, visual hallucinations were statistically more frequent in organic causes of the delusion (1.21 to 5.56*, 95% CI), whereas auditory hallucinations were statistically more prominent for the functional causes of the Capgras delusion (0.17 to 0.60**, 95% CI). The table 3. shows differences in findings of neuropsychological testing, EEG and neuroimaging. Among different cognitive areas, memory and visuospatial were statistically impaired in organic causes of the delusion ((1.52 to 12.29* and 1.11 to 11.16*, 95% CI), with EEG and neuroimaging findings statistically significant for the organic causes of the Capgras delusion (3.30 to 19.75** and 1.66 to 6.14**, 95% CI). The graph 1. shows the neuroimaging investigation including X-Ray, CT, MRI, PET and SPECT scan has been used per number of cases over the years since 1960. There is a gradual increase of CT brain scan from 1980 onward followed by a steady increase of MRI scan from 2000 to present. 160 of all reported cases were treated with antipsychotic medications including clozapine (9), while 28 of these cases received additional treatment including antidepressant medication (27), benzodiazepines (10) lithium (9), anti-dementia medication (8), anti-epileptic mood stabilisers (16), ECT (17), insulin coma therapy (4), leucotomy (1), cingulotomy (2), IV immune-globin (1) and deep brain stimulation (1). The duration of the Capgras delusion in functional disorders ranged from one month to over ten years, while in organic conditions it was reported from three days to over ten years.

Conclusion This review of all identified cases published in English demonstrates that the Capgras delusion is found in a variety of neuropsychiatric and functional psychiatric disorders. The results reveal distinct clinical symptoms and neuro diagnostic findings between functional and organic causes of the delusion. The statistically significant high frequency of inanimate objects misidentified as imposters followed by multiple objects is in favour of organic causes of Cpgras delusions. This certainly throws doubt upon the traditional psychodynamic account while the preponderance of memory deficits on neuropsychological testing, undermines the model based on selective visuo-affective impairment proposed by cognitive neuropsychiatrists. Interestingly, more cases with a neuropsychiatric disorder than functional as a primary diagnosis, showed cognitive impairment with memory and visuo spatial impairment being statistically significant. The presence of Capgras delusion should be regarded as a risk factor for violence given the statistically significant difference between organic and functional when it comes to aggression and homicide. Thorough neuroimaging and neuropsychological investigations are warranted. All functional causes showed in imaging bilateral or right fronto-temporal involvement whereas the organic causes of the syndrome may involve left hemisphere as well as global atrophy. There was no statistical significant difference between functional and organic with regard to treatment response. Active treatment can be expected to result in a reasonably good prognosis with almost two thirds having a partial or good prognosis. The long term course of Capgras delusion remains unclear as only few case reports reported extended follow-up of the patients.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.