J Neurol Neurosurg Psychiatry doi:10.1136/jnnp-2012-303520
  • Neuropsychiatry
  • Short report

Short report: is there anything distinctive about epileptic déjà vu?

  1. Adam Zeman2
  1. 1Research Department of Infection & Population Health, University College London, London, UK
  2. 2Department of Neurology, Peninsula Medical School, Peninsula College of Medicine and Dentistry, University of Exeter, Exeter, UK
  1. Correspondence to Professor Adam Zeman, Department of Neurology, Peninsula Medical School, Peninsula College of Medicine and Dentistry, University of Exeter, Room F01, St Luke's Campus, Magdalen Road, Exeter EX1 2LU, UK; adam.zeman{at}
  • Received 22 June 2012
  • Revised 23 August 2012
  • Accepted 11 December 2012
  • Published Online First 11 January 2013


Background Déjà vu can occur as an aura of temporal lobe epilepsy and in some psychiatric conditions but is also common in the general population. It is unclear whether any clinical features distinguish pathological and physiological forms of déjà vu.

Methods 50 epileptic patients with ictal déjà vu, 50 non-epileptic patients attending general neurology clinics and 50 medical students at Edinburgh University were recruited. Data were collected on demographic factors, the experience of déjà vu using a questionnaire based on Sno's Inventory for Déjà Vu Experiences Assessment, symptoms of anxiety and depression using the Hospital Anxiety and Depression Scale as well as seizure characteristics, anti-epileptic medications, handedness, EEG and neuroimaging findings for epileptic patients.

Results 73.5% of neurology patients, 88% of students and (by definition) all epilepsy patients had experienced déjà vu. The experience of déjà vu itself was similar in the three groups. Epileptic déjà vu occurred more frequently and lasted somewhat longer than physiological déjà vu. Epilepsy patients were more likely to report prior fatigue and concentrated activity, associated derealisation, olfactory and gustatory hallucinations, physical symptoms such as headaches, abdominal sensations and fear. After controlling for study group, anxiety and depression scores were not associated with déjà vu frequency.

Conclusions Déjà vu is common and qualitatively similar whether it occurs as an epileptic aura or normal phenomenon. However ictal déjà vu occurs more frequently and is accompanied by several distinctive features. It is distinguished primarily by ‘the company it keeps’.


Déjà vu—a ‘subjectively inappropriate impression of familiarity for a present experience (in relation to) an undefined past’1—is common in the general population2 ,3 but also occurs as an aura of temporal lobe epilepsy, and in some psychiatric conditions.3 The neural origin of ictal déjà vu has been localised to a neural network that includes medial temporal lobe structures and temporal neocortex.4–7 The neuroanatomical basis for déjà vu in healthy subjects is less well defined, although a recent neuroimaging study demonstrated a reduction in grey matter volume in a network of cortical (mostly mesiotemporal) and subcortical regions in healthy individuals with déjà vu.8 Despite this anatomical similarity, it remains uncertain whether pathological and physiological déjà vu share a common physiology or have comparable qualitative and quantitative features. Recent attempts have been made to improve classification of déjà vu experiences based on current models of memory function.9 In this study we examined the prevalence of déjà vu in neurological and control populations, and compared quantitative, experiential and associated features of epileptic and non-epileptic déjà vu.



We recruited three groups of participants: 50 patients known to neurology clinics in Southeast Scotland with a confirmed or clinically likely diagnosis of epilepsy with ictal déjà vu; 50 unselected patients attending neurology outpatient clinics at the Western General Hospital, Edinburgh without a diagnosis of epilepsy; 50 non-epileptic medical students attending a lecture at Edinburgh University.

Data collection

Participants completed a questionnaire adapted from Sno's Inventory of Déjà Vu Experiences Assessment10 (see online supplementary appendix A) containing items on the context, timing, frequency, duration and nature of the déjà vu experience including physical and emotional symptoms. Where treatment had abolished ictal déjà vu (n=7), frequency was recorded at pretreatment levels. A detailed example was requested to ensure that true déjà vu was being described. Participants completed the Hospital Anxiety and Depression scale and reported any diagnoses (current or past) of anxiety, depression or schizophrenia requiring treatment. Among epilepsy patients, information was also collected on features of typical seizures, antiepileptic medications, handedness and lateralisation of EEG and/or neuroimaging abnormalities.

Statistical analysis

Analyses were done using Stata V.11.0. Characteristics of epilepsy patients were compared first with neurology patients and then with students. Between-group differences in demographic characteristics, presence or absence of déjà vu, jamais vu, dissociative symptoms, psychiatric diagnoses and median anxiety and depression scores were assessed using χ2 or Mann-Whitney U tests as appropriate. Features of déjà vu were compared between groups using χ2 or χ2 tests for trend. A Bonferroni correction was used for multiple comparisons; based on comparing 23 features of déjà vu between epilepsy patients and two control groups, the p value needed for significance was 0.05/(23×2)=0.0011. In a sensitivity analysis, responses from epilepsy patients with a clinical diagnosis and no witnessed seizures (n=11) were excluded. The relationship between anxiety and depression scores and déjà vu frequency was investigated using linear regression.


Characteristics of study participants

A hundred and fifty participants were recruited (50 epilepsy patients, 50 neurology patients and 50 students). Median age across all groups was 36 years (IQR 24–50) and 63% of participants were female. Medical students had significantly lower anxiety and depression scores than epilepsy (p<0.001) and neurology patients (p=0.001). They were also less likely to have received treatment for anxiety or depression.

Features of epilepsy

Of 50 patients with ictal déjà vu, 24 (48%) had a history of a witnessed blackout or tonic-clonic seizure, 10 (20%) a witnessed complex partial seizure, 4 (8%) an unwitnessed blackout and 12 (24%) described déjà vu with temporal lobe phenomena alone (eg, olfactory or gustatory hallucinations, depersonalisation or derealisation). Nine (18%) had an epilepsy diagnosis confirmed by clinical, EEG and imaging findings, 16 (32%) were diagnosed by EEG and clinical features, 3 (6%) by imaging and clinical features and 22 (44%) had a clinical diagnosis. Most EEG foci were in the temporal lobes (eight right-sided, five left-sided, five bitemporal and seven abnormalities were either generalised or at an unspecified/non-temporal site). Seven of 12 potentially epileptogenic imaging abnormalitiesi were located in the temporal lobes (two right-sided, four left-sided, one bitemporal) and a further five abnormalities were either generalised or non-temporal.

Table 1

Characteristics of participants and features of the déjà vu experience

Features of déjà vu

All subjects with epilepsy (by definition) had experienced déjà vu at least once, compared with 36 neurology patients (73.5%) and 44 students (88%). Jamais vu was markedly less common—26% of epilepsy patients, 26.5% of neurology patients and 10% of students had ever experienced jamais vu. Déjà vu was reported in several contexts and sensory modalities across all groups, with triggers including places, speech, music and actions. Déjà vu occurred more frequently in epilepsy patients than in either control group (p<0.001) (figure 1). Epilepsy patients were more likely to report prior fatigue than students (p=0.001) but not neurology controls (p=0.01). Déjà vu lasted longer in epilepsy patients than students (p=0.001), but not neurology controls (p=0.67). Patients were less likely to remember the original occasion of which they were being reminded during ictal déjà vu than students (p<0.001) or, to some extent, neurology patients (p=0.005). Neither the experiences of precognition nor depersonalisation during déjà vu differed between groups. However epilepsy patients were somewhat more likely to report derealisation with déjà vu than students (p=0.002) but not neurology patients (p=0.02). The déjà vu of epilepsy patients had more physical associations including headaches (p=0.001), blackouts (p<0.001) and concentrated activity (p<0.001) than that of students and more blackouts (p<0.001) than that of neurology controls. They were more likely to report fear than students (p<0.001) and slightly more likely than neurology controls (p=0.009; table 1). Only epilepsy patients reported olfactory and gustatory hallucinations or abdominal sensations with déjà vu. Some descriptions of ictal déjà vu are given in box 1.

Figure 1

Déjà vu frequency in epilepsy patients, neurology patients and students.

Box 1

‘At work (in the factory) I was listening to the line leaders when suddenly I felt they'd had the same conversation before. I felt flushed, my ears popped and I tried to shut out the horrible tingly feeling.’

‘I was standing in front of the mirror with the tap running and (suddenly) the curtains, the smell…everything seemed so vivid and familiar’.

‘At work I was on the phone and a colleague asked ‘are you ok?’ The conversation seemed familiar; I felt warm and had a metallic taste in my mouth. I had to sit down’.

Sensitivity analysis

Results remained similar after excluding responses from epilepsy patients with a clinical diagnosis and unwitnessed seizures. In particular déjà vu frequency remained significantly higher in epilepsy patients than either control group (p<0.001), as did the experience of prior fatigue, lack of memory of the original occasion, association with blackouts, concentrated activity and fear. A marginally significant association with derealisation (p=0.002) remained. The length of déjà vu and the association with headaches became somewhat less significant than before (p=0.005 for each).

Relationship with anxiety and depression

Anxiety (p=0.011) and depression scores (p=0.006) were positively correlated with frequency of déjà vu in univariable analysis. However no significant association with either anxiety (p=0.15) or depression (p=0.16) remained after adjusting for study group.


Déjà vu in epilepsy occurred more frequently, lasted somewhat longer and was associated with more prior fatigue, concentrated activity and symptoms of headaches, blackouts and fear than physiological déjà vu. There was a trend towards greater reporting of derealisation in epileptic déjà vu. There was no difference in timing, contexts or features such as pervasiveness, precognition or depersonalisation in epileptic compared with physiological déjà vu.

The prevalence of reported déjà vu in the neurology and student control populations (73.5% and 88%) is in keeping with previous findings.11 The slightly higher prevalence in students may reflect underlying age differences as an inverse correlation between age and déjà vu has previously been noted.12–14 While the term ‘déjà vu’ may be ambiguous, reflecting either experiences that are simply familiar or recurrent hallucinations or intrusive memories, use of a detailed example helped to establish that true déjà vu was being described. Although we used Sno's Inventory of Déjà Vu Experiences Assessment10—originally designed as a survey tool rather than to aid clinical classification—it has satisfactory reliability and face validity and was found to be user-friendly in a sample of whom half had a low educational level.10

Ictal déjà vu occurred more frequently than physiological déjà vu described by controls and was unsurprisingly related to seizure control. Though not formally assessed, many epilepsy patients reported being able to distinguish between seizure-related and non-seizure related déjà vu, a phenomenon that has previously been described.15 Ictal déjà vu was associated with more unpleasant emotions and physical symptoms than physiological déjà vu, which is in keeping with previous findings.1 ,7 ,15 ,16

Epilepsy and neurology control patients had significantly higher anxiety and depression scores than student controls. Using the Hospital Anxiety and Depression Scale (HAD) scale, we did not intend to diagnose patients formally but rather to describe trends in symptoms of anxiety and depression between groups. Evidence from a similar population shows that almost half of new referrals to general neurology clinics meet criteria for a psychiatric diagnosis.17 It is therefore possible that some déjà vu experiences occurring in epilepsy or neurology groups were in fact related to psychiatric conditions. However while psychogenic non-epileptic seizures could not be excluded in 11 patients with a clinical epilepsy diagnosis and unwitnessed seizures, omitting their responses made little difference to the main findings. In neurology controls, misclassification of ‘psychiatric’ déjà vu as physiological déjà vu would reduce, rather than enhance, any apparent differences between ictal and physiological déjà vu, and the effect should be mitigated by use of a second (student) control group. A next step might be to probe further the experience of déjà vu in psychiatric populations.

In conclusion epileptic déjà vu was similar to physiological déjà vu, but was distinguished from it by its frequency and its associated features—the ‘company it keeps’.1


We thank Drs Aylward, Mumford and Lueck for allowing us to enrol their patients; Steff Lewis for help with statistical analysis; Stephanie Sharp, Karen Robertson and Jeanette Robb for administrative assistance.


  • Contributors AZ conceived the idea, designed the study, assisted with identifying epilepsy patients, revised drafts of the manuscript and provided supervision. CW-G identified and recruited participants, administered questionnaires, extracted and analysed data and wrote drafts of the paper.

  • Competing interests None.

  • Ethical approval The study was approved by the University of Edinburgh, Faculty of Medicine.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • i Abnormalities seen on neuroimaging were dilatation of temporal horns of ventricles (a soft sign for mesial temporal sclerosis) more marked on the left, right lateral temporal focal atrophy, post-traumatic gliotic change affecting left temporoparietal region, vascular anomaly of left temporal pole, borderline left hippocampal sclerosis, a focal area of hypoperfusion in right posterolateral cortex on interictal Single-photon emission computed tomography, demyelination of temporal lobes due to possible previous infection, numerous periventricular white matter lesions, basal ganglia calcification, diffuse previous trauma with gliosis, minimal left frontal atrophy and a cavernous hemangioma (site unspecified).


Relevant Article

Visit the full archive of podcasts for JNNP here >>

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.