TGA indicates transient global amnesia; and TIA, transient ischemic attack.
A, Electroencephalogram (EEG) bipolar montage depicting focal left temporal slowing (black boxes), which was the most common EEG finding in individuals with single and recurrent TGA episodes. B, Diffusion-weighted (left) and apparent diffusion coefficient (right) magnetic resonance imaging (MRI) results showing a recurrent TGA episode with the classic finding of focal mesial hippocampal diffusion restriction (white arrowheads). C, T1-weighted postgadolinium MRI showing a large left temporal developmental venous anomaly (white arrowhead).
eFigure. Relationship Among TGA, Migraine, and Age
eTable. Sensitivity Analysis of Olmsted County Residents
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Morris KA, Rabinstein AA, Young NP. Factors Associated With Risk of Recurrent Transient Global Amnesia. JAMA Neurol. 2020;77(12):1551–1558. doi:10.1001/jamaneurol.2020.2943
Are there physiological, environmental, or other factors that increase risk for recurrence of transient global amnesia?
In this cohort study of 1044 patients with transient global amnesia, recurrence was associated with earlier age at the time of the initial episode and higher prevalence of both personal and family history of migraine, compared with isolated cases.
These findings may help clinicians when counseling patients with transient global amnesia on their risk of recurrent episodes.
Transient global amnesia (TGA) is usually considered a benign event with a low recurrence rate. However, recurrence rates vary considerably among studies and there are no known risk factors for TGA.
To examine risk factors for the recurrence of TGA.
Design, Setting, and Participants
This retrospective cohort study involved medical record review of patients with isolated or recurrent TGA presenting to the Mayo Clinic in Rochester, Minnesota, between August 1, 1992, and February 28, 2018. A total of 1491 cases were reviewed and 1044 met diagnostic inclusion criteria for TGA, with the remainder excluded owing to indeterminate or alternate diagnoses or limited information available in the medical record.
Single vs recurrent episodes of TGA.
Main Outcomes and Measures
Demographics, precipitating factors, migraine history, imaging and electrodiagnostic findings, and family history of TGA were collected. The main outcome measure was TGA recurrence.
Of 1044 included patients, 575 (55.1%) were male, and the mean (SD) age at inclusion was 75.0 (11.5) years. A total of 901 patients (86.3%) had a single episode of TGA and 143 (13.7%) had recurrent episodes of TGA. The 2 groups were similar in age at inclusion, sex, identifiable triggers, and duration of anterograde amnesia. The number of recurrences ranged from 1 to 9, with 137 individuals (95.8%) having 3 or fewer recurrences. The mean (SD) age at first episode of TGA was 65.2 (10.0) years for individuals with a single episode vs 58.8 (10.3) years for those with recurrent episodes (P < .001). There was a personal history of migraine in 180 individuals (20.0%) with a single episode of TGA and 52 individuals (36.4%) with recurrent episodes of TGA (P < .001), and a family history of migraine in 167 individuals (18.5%) with a single episode of TGA and 44 individuals (30.8%) with recurrent episodes of TGA (P = .001). There were no electroencephalographic findings associated with increased risk of TGA recurrence. Acute and subacute temporal lobe abnormalities on results of magnetic resonance imaging were seen rarely and did not require intervention. A family history of TGA was identified in 12 individuals (1.3%) with a single episode of TGA and 4 individuals (2.8%) with recurrent episodes of TGA.
Conclusions and Relevance
This study suggests that, in this large cohort of patients with TGA, recurrent TGA was associated with earlier age at the time of first TGA episode and higher prevalence of both personal and family history of migraine compared with isolated cases. These results can be used to counsel patients about risks of recurrence and may have implications for the understanding of TGA pathophysiology.
Quiz Ref IDTransient global amnesia (TGA) is a clinical syndrome characterized by sudden-onset anterograde amnesia that typically lasts 4 to 6 hours but can last up to 24 hours. The inability to form new memories during an episode of TGA often causes patients to ask questions repetitively, while their identity, consciousness, and other neurologic functions remain intact.1-3
Episodes of TGA often have an identifiable trigger, such as the Valsalva maneuver, sexual intercourse, exercise, or strong emotion.4 These triggers support the prevailing notion that TGA may be caused by transient intracranial venous hypertension.1 There is also a strong association of TGA with migraine, suggesting a potential shared genetic or pathophysiological mechanism.5-7 However, the exact association between TGA and migraine remains unclear. Although there is a strong familial predisposition for migraine, there are only rare case reports of familial clusters of TGA. Likewise, most individuals who experience migraine never have an episode of TGA.
Quiz Ref IDResults of diagnostic workup, including electroencephalography (EEG) and brain and vessel imaging, are typically normal during and after a TGA episode. Electroencephalography findings are generally nonspecific, but rare epileptiform discharges may be present in those with concomitant epilepsy.8,9 Results of magnetic resonance imaging (MRI) may show a small focus of restricted diffusion initially, particularly in the hippocampus.10,11
Transient global amnesia is usually an isolated event, with variable recurrence rates reported in the literature. Studies with a mean duration of follow-up of at least 3 years have found recurrence rates ranging from 2.9% to 23.8%.12-20 The wide range of recurrence rates appears to reflect (1) variability of the strictness of the definition of TGA, with higher rates associated with more relaxed criteria for defining TGA; and (2) the relatively low numbers of patients with recurrent episodes, with a median of 12 patients with recurrent episodes across these studies. The differences in TGA recurrence rates among studies limit the ability of clinicians to discuss how likely it is that another event will occur. There are also no previously identified risk factors associated with TGA recurrence, to our knowledge.
Transient global amnesia is considered a benign event that is unlikely to recur. Studies suggest that TGA is not associated with an increased risk for development of other neurologic disorders, such as stroke, dementia, or epilepsy.13,14 Nonetheless, TGA can be emotionally distressing for both patients and family members, and can mimic neurologic emergencies such as stroke or seizures. It may also have implications for certain occupations.21 Therefore, it is important for clinicians to be able to counsel patients on their risk of TGA recurrence, including any factors that may increase the likelihood of another event.
We hypothesized that certain physiological, environmental, or genetic factors may increase the risk for TGA recurrence, potentially helping identify patients with a higher risk of recurrence and providing further insights into the pathophysiological mechanisms of TGA.
All patients with a diagnosis of TGA presenting to the Mayo Clinic in Rochester, Minnesota, between August 1, 1992, and February 28, 2018, were identified using the Advanced Cohort Explorer, a free-text search tool and data repository of outpatient and inpatient Mayo Clinic medical records. We reviewed records containing the term “transient global amnesia” as a diagnosis and/or in the “Impression, Report, and Plan” section of clinical notes. Patients were included in the study if they had experienced 1 (single cases) or more (recurrent cases) episodes of TGA meeting the proposed diagnostic criteria, specifically: (1) clear anterograde amnesia witnessed by an observer; (2) cognitive impairment limited to amnesia, without clouding of consciousness or loss of personal identity; (3) no accompanying focal neurologic symptoms, epileptic features or active epilepsy, or recent head injury; and (4) resolution within 24 hours.22,23 We excluded patients with alternative diagnoses, indeterminate spells without a clear cause, or insufficient information in the medical record to reach a diagnosis (Figure 1). All study protocols were approved by the Mayo Clinic institutional review board. All included patients provided consent for use of their medical records for research purposes.
A detailed medical record review was performed in May 2018, and we collected information from all visits of each patient since the time of their first TGA episode until the last follow-up. Quiz Ref IDData collected included demographic information, characteristics of TGA episodes, personal and family history of migraine, family history of TGA, and EEG and advanced imaging findings. Personal and family history data were collected from review of clinical notes and a personal and family history questionnaire, which is obtained on the patient’s initial visit and updated regularly by patients and clinicians.
Most patients were seen either in the outpatient setting on referral from the Mayo Clinic Health System or other local or regional health care systems or in the emergency department and/or inpatient setting at Mayo Clinic Hospital (Rochester, Minnesota). A minority of patients included national or international patients who receive regular medical care at Mayo Clinic or who were seeking a second opinion on an outside diagnosis of TGA or an associated condition. Mayo Clinic neurologists made or confirmed the diagnosis in 771 of 901 single cases (85.6%) and 138 of 143 recurrent cases (96.5%). In other cases, diagnoses were made by a Mayo Clinic emergency medicine physician, primary care physician, or clinician from another medical specialty.
Data were stratified into single and recurrent TGA cases for comparison. Data reported as mean (SD) values were analyzed using 2-tailed t tests when comparing single vs recurrent cases and using analysis of variance with Tukey honestly significant difference post hoc tests when comparing more than 2 groups. P < .05 was considered significant. Data reported as median and interquartile range were analyzed using 2-tailed Mann-Whitney tests. Data reported as proportions or percentages were analyzed using χ2 tests. All analyses were performed using SAS, version 9.4 (SAS Institute Inc).
We identified 1044 patients (575 men [55.1%]; mean [SD] age at inclusion, 75.0 [11.5] years) with TGA. There were 901 patients (86.3%) with single episodes and 143 patients (13.7%) with recurrent episodes. Baseline characteristics are shown in Table 1. The geographical distribution by state of residence was similar between groups. Patient age at the time of medical record review (age at inclusion) was similar between groups, with a mean (SD) age of 75.1 (11.6) years for patients with a single episode of TGA and 74.3 (11.3) years for patients with recurrent episodes of TGA (P = .45). There was a slight male predominance in both groups (single episode, 497 [55.2%]; and recurrent episodes, 78 [54.5%]). There was a significant difference in mean (SD) age at first TGA episode (single episode, 65.2 [10.0] years vs recurrent episodes, 58.8 [10.3] years; P < .001). There was also a significant difference in duration of follow-up starting from the index episode of TGA, with a median of 8.5 years (interquartile range, 4.3-13.8 years) in patients with a single episode of TGA vs 13.8 years (interquartile range, 8.4-20.3 years) in patients with recurrent episodes of TGA (P < .001).
Table 2 shows characteristics of patients with single and recurrent episodes of TGA. Among the 143 patients with recurrent episodes, 95 (66.4%) had a single recurrence and 137 (95.8%) had 3 or fewer recurrences. There were only 6 patients (4.2%) with more than 3 TGA recurrences. The duration of TGA episodes was similar between patients experiencing single and recurrent episodes, with a mean (SD) duration ranging from 4.3 (5.8) hours to 6.0 (5.1) hours (P = .23). The mean (SD) interval between the initial TGA episode and the first TGA recurrence was 4.1 (5.2) years and between subsequent recurrences was 3.2 (3.2) years, with no significant difference between groups (P = .20). The percentage of cases with identifiable triggers was similar between patients experiencing the first TGA episode in those with a single episode (259 of 901 [28.8%]) and recurrent episodes (50 of 143 [35.0%]), as well as among patients experiencing recurrent episodes (117 of 361 [32.4%]) of TGA (P = .16). There was no overall association between single or recurrent episodes of TGA and types of triggers.
The association among TGA, migraine, and age is shown in Table 3 and the eFigure in the Supplement. Recurrence of TGA was associated with a personal history of migraine in 180 individuals (20.0%) with a single episode of TGA and 52 individuals (36.4%) with recurrent episodes of TGA (P < .001), with a family history of migraine in 167 individuals (18.5%) with a single episode of TGA and 44 individuals (30.8%) with recurrent episodes of TGA (P = .001), and with a personal and/or family history of migraine in 282 individuals (31.3%) with a single episode of TGA and 72 individuals (50.3%) with recurrent episodes of TGA (P < .001). Individuals with a personal history of migraine had an earlier mean (SD) age at TGA onset (61.1 [11.3] years with migraine vs 65.4 [9.7] years without migraine; P < .001), although this association was present among patients with single episodes of TGA (61.8 [11.6] years with migraine vs 66.1 [9.4] years without migraine; P < .001), but not in those with recurrent TGA (56.8 [12.8] years with migraine vs 60.0 [8.5] years without migraine; P = .26). Among patients with a personal history of migraine, migraine aura was present in 59 of 180 individuals (32.8%) with a single episode and 18 of 52 individuals (34.6%) with recurrent episodes, with no significant difference between groups (P = .94). Patients with migraine aura had similar mean (SD) ages at TGA onset compared to those without aura in those with single episodes (61.6 [11.4] vs 61.9 [11.7] years; P = .87) and those with recurrent episodes (56.3 [15.0] vs 57.1 [11.7] years; P = .83).
Electroencephalograms were performed in a significantly higher proportion of patients with recurrent episodes of TGA than those with single episodes (51 of 143 [35.7%] vs 236 of 901 [26.2%]; P = .02). The median interval between TGA onset and EEG testing was similar between individuals with single and recurrent episodes of TGA (14 days [range, 4.0-35.3 days] vs 21 days [range, 9.0-40.5 days]; P = .08). Among those in whom EEG was performed, the most common finding was temporal slowing (Figure 2A) in 34 of 236 individuals (14.4%) with a single episode and 8 of 51 individuals (15.7%) with recurrent episodes, with no significant difference between groups (P = .82). Among EEGs performed within 1 day of TGA onset, temporal slowing was present in 4 of 36 individuals (11.1%) with a single episode and 1 of 8 individuals (12.5%) with recurrent episodes. Epileptiform discharges were present in 7 of 236 individuals (3.0%) with a single episode and no individuals with recurrent episodes. Magnetic resonance imaging data were available for 323 individuals (35.8%) with a single episode of TGA and 70 individuals (49.0%) with recurrent episodes of TGA. The median time from TGA episode to undergoing MRI scanning was 7 days (interquartile range, 1.0-19.0 days) in individuals with a single episode of TGA, including 101 of 323 (31.3%) within 48 hours and 164 of 323 (50.8%) within 7 days of the TGA episode. The median time to MRI scanning was 12 days (interquartile range, 3.0-27.0 days) in recurrent cases, including 17 of 70 (24.3%) within 48 hours and 27 of 70 (38.6%) within 7 days of the TGA episode. Most MRI results were normal or showed nonspecific findings. Eleven of 323 patients (3.4%) with a single episode and 2 of 70 patients (2.9%) with recurrent episodes had MRI results indicating mesial temporal restricted diffusion (Figure 2B), with all MRIs performed within 3 days of the TGA episode. Two patients with recurrent episodes had large left temporal developmental venous anomalies (Figure 2C), which were not present in any patients with a single episode.
A family history of TGA was reported in 12 individuals (1.3%) with a single episode of TGA and 4 individuals (2.8%) with recurrent episodes of TGA, with no significant difference between groups (P = .45). Most patients had only a single first-degree relative with TGA, but there were 2 patients (1 with a single episode and 1 with recurrent episodes) who reported multiple first-degree relatives with TGA. Among patients with family history of TGA, there was also a personal history of migraine in 6 of 12 individuals (50.0%) with a single episode and 2 of 4 individuals (50.0%) with recurrent episodes and a family history of migraine in 5 of 12 individuals (41.7%) with a single episode and 3 of 4 individuals (75.0%) with recurrent episodes.
The eTable in the Supplement shows a sensitivity analysis limited to Olmsted County, Minnesota (where the Mayo Clinic is located). Baseline characteristics were similar to those of the larger cohort; the association of migraine with TGA recurrence was not significant, despite similar differences in prevalence, owing to a lack of statistical power.
The results of this study show that, compared with isolated episodes of TGA, recurrence is associated with earlier age at the time of initial TGA episode and higher prevalence of both personal and family history of migraine. Most patients with recurrent TGA had 3 or fewer recurrences separated by 3 to 4 years. In comparison, individuals with single and recurrent episodes of TGA were similar in age at study inclusion, sex, duration of anterograde amnesia, identifiable triggers, and electrodiagnostic findings. This study identifies potential risk factors for TGA recurrence and can help clinicians when counseling patients on the risk of recurrence.
A recent retrospective study found that patients with a personal history of migraine were more likely to have recurrence of TGA, while other clinical factors, results of bloodwork, and results of imaging studies were not associated with recurrence.20 Although there was no apparent association between age at initial TGA presentation and recurrence, the study was limited by the relatively low number of individuals with recurrent episodes and short duration of follow-up. Other studies with data on TGA recurrence were not dedicated to comparing single vs recurrent episodes, and were also limited by relatively low numbers of individuals with recurrent episodes.12-19 To our knowledge, the present study is the largest analysis of individuals with recurrent episodes of TGA, with sufficient power and follow-up duration to assess significant differences from individuals with isolated episodes of TGA.
The rate of TGA recurrence in this analysis was 13.7%, which is near the median of recurrence rates estimated from prior studies. Nearly two-thirds of individuals with recurrent episodes had only a single recurrence and more than 95% had 3 or fewer recurrences, further supporting that multiple recurrences are rare. In a previous population-based study of residents of Olmsted County, Minnesota, the rate of recurrence was 14.4%, including 9.0% of individuals with episodes before and 5.4% of individuals with episodes after the index TGA event.13
There were no significant differences in the proportion or types of triggers between individuals with single and recurrent episodes of TGA, consistent with a prior study.19 However, the same trigger often precipitated multiple TGA recurrences in individual patients, suggesting a particular susceptibility and helping explain why certain patients had high numbers of recurrences. For example, in the sole patient who had 6 TGA episodes, 5 of the 6 were precipitated by intense straining due to singing. Two patients had 4 TGA episodes each, all precipitated by sexual intercourse. There are also several case reports in which a particular trigger has precipitated recurrent episodes of TGA in the same patient, including high altitude,24 sexual intercourse,25 and certain medications.26
Our study indicates that patients with migraine present with TGA at an earlier age compared with those without migraine, a finding in agreement with a previous study.27 Besides age, migraine is the only known risk factor associated with TGA, although the underlying mechanism remains unclear. Some authors suggest that a similar mechanism of cortical spreading depression might be involved in both migraine aura and TGA.6 However, in the present study, patients with or without migraine aura had similar age at TGA onset and similar rates of TGA recurrence. Ascertainment of all cases of aura, however, was limited by the retrospective study design.
Quiz Ref IDThe most common EEG finding was nonspecific, predominantly left-sided temporal slowing, which is a common finding in healthy elderly individuals.28 There was no association between temporal slowing and TGA recurrence or proximity to the TGA event. There were no epileptiform discharges seen in individuals with recurrent episodes, further supporting that TGA is not an epileptic phenomenon and that those with recurrent episodes more consistent with seizures (in whom the diagnosis of transient epileptic amnesia should be strongly considered) were appropriately excluded from this study.
Only 11 individuals with a single episode and 2 individuals with recurrent episodes of TGA showed the classically described focus of hippocampal-restricted diffusion on MRI findings. This is likely because of the relatively long time lag between clinical onset and imaging, during which most diffusion-weighted lesions would have resolved. Although 1 study suggested that reversible changes seen on diffusion-weighted imaging may be more common in individuals with recurrent compared with single episodes of TGA, this assertion was based on a relatively small data set with only 5 individuals with recurrent episodes.29 Other studies point to vascular abnormalities on MRI findings being potentially important in the pathogenesis of recurrent episodes of TGA. For instance, recurrent TGA has been associated with jugular vein thrombosis,30 vertebral artery compression,31 and dural arteriovenous fistula.32 In our study, 2 patients with recurrent TGA had large left temporal developmental venous anomalies, which may have caused local venous congestion.33 Results of EEG and MRI brain evaluation did not lead to a change in patient management, suggesting that routine performance of these studies for patients with clinically typical TGA may not be necessary.
The role of genetic factors in TGA recurrence is unknown, although there are multiple case reports of recurrent TGA within families.34 In the present study, a family history of TGA was rare and not associated with TGA recurrence. The small number of familial cases could be because family history of TGA is not routinely investigated by clinicians or known by patients. A role for genetic factors in TGA is supported by our findings that family history of migraine was an independent risk factor for TGA recurrence. However, more long-term follow-up studies are needed to address this issue.
Our study has several limitations. The retrospective design is subject to bias, including the potential for missing data given that some patients travel to the Mayo Clinic from long distances and then undergo follow-up locally. Key factors associated with the risk of TGA may have been missed because they were not asked about during clinical encounters. Further prospective studies in which investigators inquire specifically about migraine, family history of TGA, and environmental factors are needed to help validate these findings.
There is the potential for ascertainment bias given that patients with recurrent TGA may have had more clinical encounters compared with patients with a single episode of TGA, and thus more opportunities for obtaining pertinent historical details. There is also the possibility of referral bias because patients may be more likely to seek an opinion at the Mayo Clinic after experiencing multiple episodes of TGA as opposed to an isolated episode. However, the similar geographical distribution between individuals with single and recurrent episodes, results of the sensitivity analysis of Olmsted County residents, and the similar TGA recurrence rate compared with a previous population-based study in our region argue strongly against a major referral bias affecting the results. Given the nature of the finding that individuals with recurrent TGA had an earlier age at initial presentation compared with those with isolated episodes, individuals with recurrent episodes had an inherently longer duration of follow-up. However, individuals with a single episode of TGA still had a median duration of follow-up of 8.5 years, which should be sufficient to detect most recurrences.
Transient global amnesia is a common clinical syndrome encountered by emergency department clinicians, neurologists, and primary care clinicians. Common practice is for clinicians to reassure patients about the benign nature of the syndrome and its low risk of recurrence. Quiz Ref IDOur study, combined with others, indicates that TGA recurrence is not rare, and that recurrence rates of around 10% to 15% within 5 years are reasonable to cite. When the clinical presentation of the TGA recurrence is typical, evaluation with EEG and MRI may not be necessary because the results are very unlikely to change management or affect prognosis. Patients with a personal and/or family history of migraine and those who are younger at the time of a first episode of TGA (ie, <60 years) have a higher risk for TGA recurrence.
Accepted for Publication: June 5, 2020.
Corresponding Author: Ken A. Morris, MD, PhD, Department of Neurology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (email@example.com).
Published Online: August 31, 2020. doi:10.1001/jamaneurol.2020.2943
Author Contributions: Dr Morris had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Morris.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Morris, Young.
Administrative, technical, or material support: Morris.
Supervision: Rabinstein, Young.
Conflict of Interest Disclosures: None reported.
Meeting Presentation: This study was presented at the Annual Meeting of the American Academy of Neurology; May 6, 2019; Philadelphia, Pennsylvania.
Additional Contributions: Jay Mandrekar, PhD, and Prabin Thapa, MSc, Department of Biomedical Statistics and Informatics, Mayo Clinic, assisted with statistical testing. They were not compensated for their contributions.