Abstract
Background
We report the occurrence of palinopsia in patients with migraine and its correlation with migraine characteristics, triggers and allodynia.
Methods
This study included 153 consecutive patients with migraine and recorded their clinical details, including allodynia and migraine triggers and characteristics. Palinopsia was evaluated in migraineurs and 101 controls by using a questionnaire and a novel method.
Results
According to the questionnaire assessment, 9.8% migraineurs had palinopsia. According to the novel method, 57.5% of migraineurs and 12% of controls had palinopsia. Migraineurs most frequently had palinopsia to red color (51.6%), followed by yellow (49.7%), blue (47.7%), green (46.4%) and the least to white (30.7%). A similar pattern with a lesser frequency was noted in controls. The duration of palinopsia was longer in migraineurs than in controls (32.68 ± 20.24 vs. 5.92 ± 4.55 seconds; p < 0.001). Migraineurs with palinopsia differed from those without in terms of noise as a migraine trigger (p < 0.001) and allodynia as a migraine-associated phenomenon (p = 0.03). In multivariable analysis, predictors of palinopsia were the frequency (p = 0.003) and severity (p = 0.04) of headache and the presence of headache during examination (p = 0.0001).
Conclusion
Migraineurs had a pattern of palinopsia to different colors that was similar to the controls, but the palinopsia of migraineurs was more frequent and of longer duration, especially during headaches.
Introduction
Palinopsia, originating from the Greek words palin (again) and opsis (vision), is defined as the reappearance of a visual perception after the stimulating object has been removed from the visual field (1). This term has been interchangeably used with the terms “persistence of afterimage” or “perseveration of visual image” (2). Palinopsia has been described in association with various structural lesions localized to the right parieto-occipital region, psychoactive drugs, transient ischemic attack, epilepsy, retinal disease, psychiatric disorders and migraine (3–8). Both migraines with and without aura are associated with photophobia, phonophobia and allodynia. Various evoked-potential studies have reported hyperexcitability and impaired habituation of cortical neurons in patients with migraine (9–12). In a study on median somatosensory evoked potentials in migraine, allodynia was related to sensitization and impaired habituation of the N19 potential (9). Visual afterimage and palinopsia may be clinical markers of sensitization and impaired habituation of sensory neurons. Studies on palinopsia in migraine are limited. Certain studies have reported palinopsia in association with migraine (13–16), but only one study has reported the occurrence of palinopsia in patients with migraine (17). These reports are based on questionnaires, and none of these studies have documented palinopsia by using a standardized technique. In the present study, we report the frequency of palinopsia in patients with migraine using a novel method of documentation and compare patients with and without palinopsia in order to identify predictors of palinopsia in migraine.
Subjects and methods
This is a cross-sectional case–control study. Consecutive patients with migraine visiting the neurology out-patient department of Sanjay Gandhi Postgraduate Institute of Medical Sciences between July 2014 and January 2015 were included. Migraine was diagnosed based on the International Headache Society Criteria (IHS-2004) (18). The study design was approved by the Institutional Ethics Committee (no.2014-171-IP-EXP-PGI/BE/700/2014). Patients with seizures, stroke, transient ischemic attack or major psychiatric disorders and on antipsychotic, antiepileptic, psychoactive or sedative drugs or corticosteroids were excluded. In addition, patients who were pregnant or had a history of collagen vascular disease, malignancy, organ transplantation, liver or kidney failure, addiction to alcohol, marijuana or other drugs or uncontrolled diabetes or hypertension were excluded.
Clinical evaluation
Demographic characteristics, occupation, education level and duration of illness, along with frequency and severity of headaches in the previous month, were recorded. Functional disability and associated symptoms during migraine attacks were also recorded. The presence of aura was recorded, and migraine triggers were evaluated using a 19-point questionnaire (19). Moreover, allodynia was evaluated using a questionnaire (20,21). The time from onset of headache to peak severity was noted, and the severity of migraine was assessed on a scale of 0–4 (0 = none, 1 = mild, 2 = moderate, 3 = moderately severe and 4 = severe) (9). The overall severity of headache in the previous month was assessed using a 0–10 Visual Analog Scale (VAS). The presence of headache and its severity at the time of palinopsia evaluation were also enquired. Intake of abortive drugs within 24 hours and during the previous month was noted. The use of prophylactic drugs was recorded, and patients were advised to stop these drugs 15 days before enrollment. During this period, such subjects were prescribed non-steroidal anti-inflammatory drugs if needed.
Evaluation of palinopsia
Questionnaire used for the evaluation of palinopsia.
Moreover, the aforementioned protocol was used in 101 healthy, age-matched (31.7 ± 11.4 vs. 34.3 ± 11.0 years; p = 0.07) and gender-matched (female, 71.9% vs. 61.4%; p = 0.10) controls who did not have migraine. The case–control ratio was set at 3:2.
Statistical analysis
The presence of palinopsia in the migraineurs versus controls was compared using Fisher’s exact test. The longest duration of palinopsia between migraineurs and controls was compared using the Mann–Whitney U test. Among patients, the duration of palinopsia triggered by each color was compared using analysis of variance. The number of patients and controls with palinopsia to different colors was compared using Fisher’s exact test. The clinical and migraine characteristics of patients with and without palinopsia were compared using parametric tests for categorical variables and non-parametric tests for continuous variables. Predictors of palinopsia were derived using binary logistic regression analysis by including variables with p-values <0.1 in the univariate analysis. A variable with a two-tailed p-value <0.05 was considered significant. The analysis was conducted using SPSS version 16 (Chicago, IL, USA) or GraphPad Prism 5 software (San Diego, CA, USA).
Results
In total, 153 migraineurs were included, with 71.9% females and a mean age of 31.7 years. Their median schooling was 10 (range, 0–23) years, and 87 (56.9%) patients resided in rural areas. Ninety-six patients were homemakers, 18 were employed, 16 were farmers, 13 were manual laborers and 10 were students.
Moreover, 146 (96%) patients had migraine without aura. The median duration of migraine was 7 (range, 1–35) years and the frequency of headache was 7 (range, 1–30) per month. Ninety (58.8%) patients had episodic migraine, whereas 65 (41.2%) patients had chronic migraine (>15/month). The number of abortive drugs taken in the previous month ranged between 0 and 30 (median, 5). Fifty-seven patients were on prophylactic medications until 15 days before examination; these included amitriptyline in 32, antiepileptic drugs in 10, flunarizine in 3 and beta-blockers in 14. No patient had received topiramate. The VAS score of severity of headache ranged between 5 and 10 (median, 9). On a severity scale of 0–4, 65 (42.5%) patients had grade 4, 48 (31.4%) had grade 3 and 40 (26.1%) had grade 2 headache. A majority of patients had photophobia (148, 96.7%) and phonophobia (147, 96.1%). Phonophobia was defined as a condition wherein normal sounds were painful for the patient during a headache and so he/she preferred a quiet environment. All patients reported triggers, and most reported multiple triggers. Bright sunlight and noise triggered migraine attacks in 106 (69.3%) and 88 (57.5%) patients, respectively. Noise was defined as loud, disturbing sounds that made it difficult to hear desired sounds. Noise was considered a trigger if headaches were precipitated by going to a noisy place (e.g., a marketplace, railway station or discotheque). Forty-five (29.4%) patients reported allodynia: mechanical and dynamic in 44 (28.8%) patients each and thermal in 12 (7.8%) patients.
Palinopsia
Fifteen (9.8%) patients reported palinopsia. The details are summarized in Table 2. Using our novel method of testing, 88 (57.5%) migraineurs and 12 (11.9%) controls reported palinopsia (p < 0.001). Migraineurs most frequently reported palinopsia to red color (51.6%), followed by yellow (49.7%), blue (47.7%), green (46.4%) and the least to white (30.7%). Similarly, the controls most frequently reported palinopsia to red (9.9%) and the least to white (3%) (Figure 1). The duration of palinopsia was significantly longer in migraineurs than in controls (32.68 ± 20.24 vs. 5.92 ± 4.55 seconds; p < 0.001) and was longest for red and shortest for white. Figure 2 presents the details of palinopsia in patients and controls. Considering the responses to various colors, 19 (22.6%) patients had only positive visual afterimages, 61 (72.6%) had both positive and negative visual afterimages and 4 (4.8%) had only negative visual afterimages. In controls, nine (8.8%) patients had only positive visual afterimages and three (3%) had both positive and negative visual afterimages.
Bar diagram showing the percentages of migraine patients and controls with palinopsia to different colors by using the novel method of testing. Error bar diagram showing the duration of palinopsia to different colors in patients with migraine versus controls. Characteristics of palinopsia on history reported by 15 migraineurs.

Correlation of palinopsia
Comparison of clinical characteristics of migraine patients with and without palinopsia.
CM: chronic migraine; VAS: Visual Analogue Scale.
Comparison of abortive and prophylactic drugs in migraine patients with and without palinopsia.
AED: antiepileptic drug (e.g., valproate or divalproate); CCB: calcium channel blocker (e.g., flunarizine); NSAID: non-steroid anti-inflammatory drug (e.g., ibuprofen or naprosyn).
Discussion
This is the first study on palinopsia in migraine that used a standardized documentation method and systematically reported predictors of palinopsia. Overall, 9.8% of enrolled migraineurs had palinopsia according to the questionnaire, while 57.5% of patients had palinopsia according to our novel evaluation method. The color red was associated with the highest incidence and longest duration of palinopsia in migraineurs compared to blue, yellow, green and white. In addition, a similar pattern was observed in the controls, with lower frequency and shorter persistence. Multivariable logistic regression analysis revealed that frequency, severity and presence of migraine headaches at the time of examination were independently associated with palinopsia.
Palinopsia in migraine has been reported in earlier studies (13–16). A questionnaire-based study on 200 migraineurs reported palinopsia in 12.8%, more frequently in migraineurs with aura compared to those without (14.2% vs. 6.6%, respectively). In another study, palinopsia was inversely proportional to the frequency of migraine (17). By contrast, in the present study, 57.5% of migraine patients had palinopsia according to our novel evaluation method, and all but 4% had migraine without aura. The low prevalence of migraine with aura is not unexpected because previous studies have also reported a lower frequency of aura (2.25–24.3%) in Asians (19,24–27).
In the present study, palinopsia was associated with a higher frequency and severity of migraine attacks, as well as with the presence of headache at the time of examination. The relationship of repeated painful stimulation with excitability of the brainstem and cortical neurons has been reported in both animal and human studies (28–30). In a rat model, repeated stimulation of the dura resulted in a decrease in the sound intensity required to induce a startle reaction and allodynia (28). Sensitivity to pain positively correlated with the severity of pain (29). Similarly, the intensity of photophobia and phonophobia has been correlated with headache severity (30). In our study, noise as a trigger of migraine attack correlated with palinopsia. Noise can not only trigger headaches, but also increase headache severity (31). According to the reported literature, headache attack was triggered by noise in 50–70% of migraineurs, by odor in 50% and by visual stimuli in 40% (32). The association of visual, auditory, olfactory and somatosensory hypersensitivity with migraine has resulted in the concept of the multisensory integration theory. There are cortical and subcortical neurons that are capable of receiving different sensory stimuli, which form the multisensory convergence zone. The neurons of the superior colliculus, basal ganglia and the premotor, posterior parietal, inferior prefrontal and posterior superior temporal cortices have multisensory convergence capability. The association of palinopsia with noise and allodynia in our study may be attributed to this multisensory convergence theory (32). The presence of photophobia, phonophobia and allodynia in migraine suggests hyperexcitability of cortical neurons. In the present study, palinopsia was related to both noise and allodynia, suggesting hyperexcitability of sensory and auditory cortical neurons. The cortical areas devoted to human vision extend from V1 to V8 (33). In migraine, cortical spreading depression (CSD) usually originates from V1 and spreads anteriorly (34). The CSD margin has an intense depolarizing electrical activity, which has been postulated as the cause of visual aura, photophobia, phonophobia, palinopsia and cutaneous allodynia (9,35). If the CSD margin extends to the parieto-temporal region, it may give rise to palinopsia, allodynia or phonophobia. The area behind the CSD has hypometabolism, which gives rise to scotoma, blurring of vision or other negative symptoms (35).
Palinopsia is associated with visual snow in migraine, but none of our patients experienced this symptom (22). Palinopsia has been classically attributed to right parieto-occipital lesions (36,37). In a review of 46 patients with palinopsia, the location of lesions was right parieto-occipital in 17, left posterior hemispheric in 12, bilateral in 4, possible right sided in 5, possible left sided in 1 and of uncertain laterality in 4 patients. The laterality could not be tested in one blind patient (2). In these patients, palinopsia occurred mostly in the defective field. In another patient, palinopsia improved after removal of a left temporo-parieto-occipital glioblastoma (38). In our patients, by contrast, palinopsia was present in both the visual fields and was present even on eye closure.
Palinopsia may be a manifestation of visual after-sensation, sensory seizures, hallucinations or psychogenic elaboration of fantasies (1). Palinoptic seizure is usually accompanied by other symptoms and signs of seizure, does not have hemianopia and involves an image spread to the entire visual field, which is usually not affected by the surrounding environment (36,39). Hallucinatory palinopsia is a long-lasting, high-resolution, isochromatic image that is not affected by environmental light and motion (23). In our study, both migraineurs and controls experienced formed palinopsia.
Cummings et al. described palinopsia as a type of release visual hallucination resulting from impaired suppression of normal visual input (36). The Bayesian theory explains palinopsia as a capability of the brain to predict its visuospatial surroundings (40,41). In a previous study, migraineurs had increased sensitivity to white, red and blue light compared to those with tension-type headaches and control subjects (42). In our study, patients most frequently had palinopsia to red color, which was also the most persistent. White color was the least frequent and least persistent. These phenomena were observed in both migraineurs and controls, although migraineurs had more frequent and more persistent palinopsia compared to controls. The increased frequency of palinopsia in migraineurs may suggest an enhanced physiological function of cortical neurons in the migraine. The enhanced physiological function of the posterior cerebral cortex may be attributed to the impaired regulation of cortical neurons by the brainstem–thalamic serotonergic system or aminergic system (35). Previous studies have reported depletion of 5-hydroxytryptophan levels in migraineurs (43,44). This may be the basis of the improvement in headache frequency and severity following amitriptyline therapy in migraine patients (45). Topiramate-induced palinopsia has been attributed to impaired GABAergic transmission (46).
Previous studies have reported low serum endorphin levels in migraine, which have been shown to improve following high-rate repetitive magnetic stimulation. The improvement of endorphin levels is associated with a reduction in migraine frequency (47,48). Low serum endorphin levels may result in cortical excitability, which may explain the efficacy of various antiepileptic drugs in migraine (45,48).
The present study is limited by the lack of appropriate evaluation of other forms of palinopsia, such as those found in patients with trailing or visual snow. Moreover, we did not perform cranial imaging in all of the patients because they had classical migraine with no neurological deficit. We used our previously published questionnaire for allodynia assessment, which has not yet been validated (20,21). Migraine was diagnosed based on IHS-2004 and not on the International Classification of Headache Disorders (ICHD)-beta version because this classification is under field trial and has not yet been finalized (49). The strength of the study is the novel detection method for palinopsia and its correlation with migraine triggers and allodynia.
Our patients had the immediate type of palinopsia, wherein an image persists after the disappearance of the actual scene and usually fades over a period of several minutes. This type of palinopsia bears some similarity to the normal phenomenon of an afterimage experienced after prolonged viewing of a bright object (50). We conducted the test in a dark room and used similar brightness, size of screen and time of exposure for all patients and controls, as well as for all tested colors.
Palinopsia is common in migraine, especially during headaches and in patients with frequent and severe migraine attacks. Additional studies are needed in order to evaluate the basis of palinopsia by using neurophysiological and imaging techniques.
Clinical implications
A novel method of palinopsia evaluation has been reported. According to the questionnaire, 10% of migraineurs had palinopsia, whereas according to the novel method, 58% of migraineurs had palinopsia. Patients most frequently had palinopsia to red color and least frequently to white color. Palinopsia was related to noise and allodynia. The predictors of palinopsia were the frequency and severity of migraine and headache during examination.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethical approval
The study was approved by the Institutional Ethics Committee (no.2014-171-IP-EXP-PGI/BE/700/2014).
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Language editing support for this manuscript was provided by Cactus Communications and funded by Sanofi (India).
Acknowledgments
We thank UK Misra for his suggestions regarding the manuscript and SK Mandal, Scientist, Center of Biomedical Research, for his help with the statistical analysis.
