Transient Global Amnesia - Risk Factors and Putative Background
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.288979
Source of Support: None, Conflict of Interest: None
Keywords: Autoimmune thyroiditis, migraine, transient global amnesia, vascular risk factor
Transient global amnesia (TGA) is a temporary short-term memory loss reversible within 24 hours. Typical features of TGA include anterograde amnesia (from a few to several hours, most often from 4 to 8) accompanied by repetitive questioning and disorientation with respect to time and place, but the preserved ability to recognize oneself and other people., On examination of TGA patients, no symptoms and signs of neurological deficit are found., The onset of symptoms is sudden and unpredictable. Approximately, one third of patients perform complex activities before an episode. Typical repetitive questioning (e.g., Where am I? What is the time? What is happening?) must be witnessed by an observer to fulfill the TGA criteria. During TGA patients present impairment in learning new verbal material and disorientation in time. However, the ability to speak, visual-motor coordination, working memory, executive functions, and the ability to perform complex tasks (e.g., driving or cooking) remain intact., The clinical picture of TGA is distinctive but its recognition requires good interviewing and observation skills.
The incidence of this entity in the general population is estimated at 3.4–10.4/100,000/year. TGA patients are usually middle-aged or elderly with the predominance in women. Based on published studies, the mean age of an episode ranges from 61 to 67.3 years., TGA episodes usually subside completely, but in 2.9%–26.3% patients there is a possibility of recurrence.
Etiology of TGA remains unclear with various hypotheses. Migraine, epilepsy, cerebral ischemia (associated with hypotension or vasospasm), and psychogenic background are considered as related or underlying conditions, but none of these hypotheses has been confirmed., Associations between TGA and cerebrovascular disease/stroke and their risk factors have been frequently discussed in the literature with contradictory results.,,, The main differentials for TGA are transient epileptic amnesia (TEA) and transient ischemic attack (TIA).
The aim of this study was to analyze clinical characteristics, neuroimaging, and electrophysiological findings as well as comorbidities in TGA patients with regard to possible background of the syndrome.
The retrospective medical records analysis of 56 patients with TGA hospitalized from January 2008 to December 2016 in the Department of Neurology, Wroclaw Medical University was conducted. Diagnosis of TGA was made according to Hodges and Warlow's criteria [Table 1]. In our Department, a standard protocol for taking history is used, based on the above criteria, which made it possible to identify patients retrospectively.
The exclusion criteria were as follows: Alteration in consciousness, loss of personal identity, recent history of head trauma, active epilepsy, presence of focal neurological deficits, duration longer than 24 hours. Data included in the analysis constituted a routine part of hospital diagnosis. Course of the episode, its circumstances, and potential triggers were analyzed, as well as neuroimaging and electrophysiological findings, blood tests' results and comorbidities (with a special regard to cerebrovascular disease risk factors: hypertension, diabetes, hyperlipidemia, ischemic heart disease, and smoking habit). All patients underwent detailed cognitive assessment including assessment of retrograde memory. The data on atmospheric pressure and mean temperature of the air at the day of TGA occurrence and the change in barometric pressure and temperature since the preceding day were collected from The Institute of Meteorology and Water Management – National Research Institute.
Ethics approval was obtained from the Commission of Bioethics at the Wrocław Medical University. During their hospitalization all the subjects provided informed consent concerning all the diagnostic tests and procedures, as well as processing their personal and medical data for clinical, didactic, and scientific purposes.
The 56 patients with TGA included 42 women (75%) and 14 men (25%) whose mean age was 66 ± 7.8 years (ranging from 50 to 84 years). Demographic characteristics of the study group are presented in [Figure 1].
A total of 52 of patients (92.9%) underwent their first-ever episode of TGA, 3 patients (5.4%) – their second, and one patient (1.8%) – their third.
Duration of amnesia in 29 patients (51.8%) ranged from 1 to 5 hours, in 13 cases (23.2%) from 5-14 hours, 14 patients (25%) could not rate this precisely but in all cases amnestic syndrome resolved within 24 hours. In 32 patients (57.4%) TGA symptoms occurred in the morning (from 6:00 to 12:00), in 7 patients (12.5%) between 12:00 and 18:00, and 17 patients (30.4%) experienced TGA after 18:00.
Neurological examination revealed no abnormalities during TGA attack and 24 hours afterwards. 34 patients (60.7%) did not associate anything with the amnestic episode. 22 patients (39.3%) noted potential specific triggers or events before the onset of TGA, with stressful event as the most common [Table 2].
Seasonal and atmospheric variation in the incidence of TGA
Most patients had TGA episodes in summer (n = 20; 35.7%) and winter months (n = 15; 26.8%). The most common month of episode occurrence was June and December [Figure 2]; the most common weekdays were Sunday (25%) and Wednesday (21.4%). The daily mean ambient temperature in June was 19°C (min. 9°C, max. 27°C), in December was 2°C (min. –8°C, max. 12°C). The daily mean atmospheric pressure in June was 1015 hPa (min. 999 hPa, max. 1025 hPa) and in December was 1025 hPa (min. 1010 hPa, max. 1042 hPa). Atmospheric pressure and mean temperature of the air at the day of occurrence of each TGA were not statistically related with TGA episodes. Temperature and pressure values did not differ from the average for a given month and there were no fluctuations and changes in temperature and pressure recorded on these days and the preceding ones.
Chronic diseases, cooccurring in 92.9% patients, included hypertension (60.7%), dyslipidemia (48.2%), autoimmune thyroiditis (17.9%), and ischemic heart disease (14.3%). One patient suffered from migraine (1.8%). 22 patients smoked regularly [Table 3]. 34 patients (60.7%) had multiple comorbidities. Among patients with autoimmune thyroiditis - 7 patients (12.5%) had Hashimoto and 3 patients (5.4%) – Graves–Basedow disease. All these patients were euthyroid during TGA. The diagnosis of Hashimoto and Graves–Basedow disease was made by the endocrinologist based on current guidelines.
All the patients underwent brain imaging study using computed tomography (CT): 32 patients had subsequent (57%) magnetic resonance imaging (MRI) and 24 subjects (43%) had only CT. Brain imaging abnormalities were reported in 24 patients (42.6%); the most common were multifocal small white matter hyperintensities reflecting small vessel disease in 18 (32.1%), leukoaraiosis in 4 (7.1%), 2 patients (5.3%) had prior ischemic stroke. Doppler ultrasonography of carotid arteries showed abnormalities in 29 patients (51.8%) [Table 4]. Hemodynamically significant stenosis of carotid vessels was defined as 51–70% and severe stenosis > 70% according to the criteria Society of Radiologist in Ultrasound Consensus. Although abnormalities in EEG were observed in 17.6% of patients, they included occasional groups of theta waves during hyperventilation or photostimulation (which may occur in healthy persons, without clinical significance) or generalized slowing of background activity. No paroxysmal/epileptiform discharges were recorded in any of the patients [Table 4].
The age and sex distribution of the studied TGA patients were comparable with other studies.,, The clinical picture of TGA episodes in our subjects was typical and thus fulfilled Warlow's diagnostic criteria. It comprised anterograde amnesia accompanied by variable retrograde amnesia, repetitive questions concerning present circumstances, with intact autopsychic orientation, distorted allopsychic orientation, and preserved consciousness. There were no automatisms or focal neurological deficits. In all patients, the symptoms resolved within 24 hours.
In our study group, TGA most often occurred in the morning and least often in the afternoon (between noon and 6 p.m.). Similar relationships were observed in other studies.,
22 patients (39.3%) noted specific triggers prior to the onset of TGA. In other studies, the most commonly reported ones included emotional stress, physical exercise, sudden ambient temperature changes, and sexual activity. Bartsch and Deuschl found some gender differences, with TGA incidence related to physical effort in men and emotional stress in women. There are a few case reports describing TGA after angiography of cerebral vessels.
Less is known about seasonal occurrence of TGA. Consistent with our findings, Sabau and Comanescu also reported a higher frequency of TGA attacks in spring and summer. Others claimed that the incidences are more common in autumn and winter , or spring and autumn. In prior studies, the analyses of impact of temperature and atmospheric pressure on the incidence of TGA were ambiguous. While in some papers, the higher temperature was related to increased risk of TGA occurrence, the others (similar to our results) did not confirm such a relationship., In prior studies, the impact of temperature and atmospheric pressure on the incidence of TGA was not confirmed. Similarly, we also did not find a correlation between these atmospheric conditions and TGA.
The etiology of TGA has still not been clearly established. Various pathophysiological mechanisms are contemplated primarily by cerebral arterial ischemia in CA-1 area of the hippocampus. It is a structure particularly sensitive to metabolic stress and hypoxia., Other studies suggest that potential causes of TGA may be associated with venous insufficiency, deep vein thrombosis, and impaired blood flow in the hippocampal region. It has been reported that patients with TGA may present high signal intensity within one or both hippocampi on diffusion-weighted imaging (DWI) as well as on T2-weighted and fluid-attenuated inversion recovery (FLAIR) images. MR examination should be performed within 48 hours of symptoms onset to confirm these findings; however, there are some studies reporting no signal changes although imaging was done early. Furthermore, even when there are signs of diffusion restriction within hippocampi on DWI, these changes are reversible and disappear also on T2-weighted and FLAIR images on follow-up examination. It has been hypothesized that such findings may suggest that reported signal changes are not related to ischemia or ischemic changes are reversible and do not lead to infarction.
The most often discussed possible risk factors for TGA include migraine and other cerebrovascular risk factors, epilepsy, and psychogenic origin.,
The data on relationships between TGA and cerebrovascular disease are often contradictory. In many studies, no evidence of cardiovascular risk factors was found in patients with TGA.,,, However, Romero et al. demonstrated that the incidence of vascular risk factors was comparable in TGA and TIA. Jang et al. in turn revealed in TGA patients a higher prevalence of ischemic heart disease and dyslipidemia than in those with TIA and a higher prevalence of hyperlipidemia and ischemic heart disease compared to an age-matched control group. In our study group, dyslipidemia and hypertension were the most common comorbidities. Almost half of the group have had atherosclerosis of carotid arteries or multifocal small white matter hyperintensities reflecting small vessel disease. In our group, the incidence of vascular risk factors did not differ significantly from that in the general population in Poland except for diabetes (8,9% vs 4.4–5.8%) and small vessel disease (32.1% vs 25%) - more frequent in patients with TGA.,, Although the sample size of our group might have biased the significance of these findings, association between TGA and cerebrovascular dysfunction seem relevant. Thus, the background of TGA in these patients seems associated with cerebrovascular dysfunction. On the contrary, the clinical picture (isolated memory loss, duration of several hours) allowed to differentiate the episodes from TIA, and no signs of recent- acute/subacute brain ischemia were found on DWI in those patients who had MRI performed.
The variety of recurrence rates in different studies ranges from 2.9% to 23.8%., Four (7.2%) of our patients experienced a prior episode of TGA. All of them had hypertension, diabetes, and hemodynamically nonsignificant atherosclerosis of carotid arteries; two of them also had dyslipidemia and the female patient with the third TGA episode suffered from autoimmune thyroiditis. These data indicate an increased risk of recurrent TGA episode in patients with cardiovascular risk factors, which again may suggest vascular background of TGA.
Other potential risk factors of TGA include migraine and epilepsy. A history of migraine has been reported in only one of our patients and none of them presented migrainous features during an acute TGA event. Our findings are not consistent with studies considering shared pathophysiologic mechanisms of TGA and migraine., A possible explanation may be cortical spreading depression (CSD) and temporary hippocampal dysfunction triggered by emotional stress that causes glutamate release in the hippocampi. Although animal-based studies suggest that CSD could be related to migraines, it has not yet been disclosed in humans. Evans and Lewis  on the contrary claimed that a common background for TGA and migraine is less probable, considering the older age of TGA patients, different triggers for both conditions, and recent MRI findings. Some studies have focused on venous congestion in the migraine, indicating that intracranial venous reflex and jugular valve insufficiency can occur in TGA patients.
The possible relationship between TGA and epilepsy, although often discussed, has not yet been clarified. Some theories view TGA as a form of temporal lobe seizure or a condition similar to postictal Todd's paresis. On the contrary, clear consciousness during TGA and its low recurrence rate (usually a single event over a lifetime) contradict this hypothesis. TEA has to be considered in the differential diagnosis of TGA. TEA tends to be shorter and recurrent. Moreover, temporal lobe seizure features such as oral automatisms or olfactory hallucinations may accompany TEA but not TGA., In our material, none of the patients had a history of epileptic seizures. Although abnormalities in EEG were observed in 17.6% of patients, they did not include paroxysmal/epileptiform discharges. Minor EEG abnormalities without epileptic features were also observed by Quinette et al. in 18.3% (26 of 142) of TGA patients.
Few papers highlight the role of psychogenic origin in TGA, with psychiatric disorders (depression, personality, or anxiety disorders) as a possible background., This relationship is suggested especially in younger and female patients. Psychogenic amnesia as a part of a defence mechanism should be considered in those patients who exhibit autobiographical memory deficits and stay indifferent to the experienced memory loss, which is not typical for TGA., Although almost 20% of our patients experienced a TGA episode after a stressful situation, their memory loss did not cover these or preceding events. Moreover, none of them had had a history of depression or anxiety disorders.
Among comorbidities which might be potentially linked with TGA, autoimmune thyroiditis drew our attention (especially Hashimoto's disease, recognized in approx. 13% of patients). The most serious complication of Hashimoto's thyroiditis is Hashimoto encephalopathy (HE), a still poorly understood condition with a range of unspecific symptoms of varying severity., Two types of HE have been described so far: the first one with recurrent stroke-like incidents due to transient cerebral ischemia and with frequent cognitive dysfunction and the second one with progressive dementia, psychotic symptoms, and epileptic seizures., In the patients with the second type of HE, MR revealed bilateral hippocampal or thalamic lesions. TGA might be perhaps considered as an early manifestation of encephalopathy in the course of autoimmune thyroiditis. In recent years, the recognition of thyroid diseases in the general population has increased significantly, which might be plainly reflected within the studied a group of patients with TGA. However, the possible links between TGA and encephalopathy in the course of autoimmune thyroiditis seem interesting and deserve further investigation.
Despite the rarity of TGA and diagnostic difficulties, we managed to characterize a relatively large group of patients. We also aimed at the possibly multidimensional analysis of their data, including clinical picture, comorbidities, and imaging and electrophysiological tests' results.
The limitations of the study include its retrospective and cross-sectional mode. Retrospective analysis of data resulted in diversity of neuroimaging performed in the patients: 57% had an MRI and 43% only a CT brain scan. Prospective observation of the patients might have provided more relevant findings on links between TGA and cerebrovascular disease as well as prognostic factors and we plan to conduct the follow-up of these patients as a further study in this field.
Analysis of clinical characteristics, comorbidities, and neuroimaging findings in the patients having experienced TGA suggests a putative cerebrovascular background of this condition. No evidence has been provided for the association between TGA and epilepsy or migraine. [40,41]
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]