Is the Long-Term Prognosis of Transient Ischemic Attack or Minor Ischemic Stroke Affected by the Occurrence of Nonfocal Symptoms?
Background and Purpose—In patients with a transient ischemic attack or ischemic stroke, nonfocal neurological symptoms, such as confusion and nonrotatory dizziness, may be associated with a higher risk of vascular events. We assessed the relationship between nonfocal symptoms and the long-term risk of vascular events or death in patients with a transient ischemic attack or minor ischemic stroke.
Methods—We related initial symptoms with outcome events in 2409 patients with a transient ischemic attack (n=723) or minor ischemic stroke (n=1686), included in the Life Long After Cerebral ischemia cohort. All patients underwent a standardized interview on the occurrence of focal and nonfocal neurological symptoms during the qualifying event. The primary outcome was the composite of any stroke, myocardial infarction, or vascular death. Secondary outcomes were all-cause death, vascular death, cardiac death, myocardial infarction, and stroke. Hazard ratios were calculated with Cox regression.
Results—Focal symptoms were accompanied by nonfocal symptoms in 739 (31%) patients. During a mean follow-up of 10.1 years, the primary outcome occurred in 1313 (55%) patients. There was no difference in the risk of the primary outcome between patients with both focal and nonfocal symptoms and patients with focal symptoms alone (adjusted hazard ratio, 0.97; 95% confidence interval, 0.86–1.09; P=0.60). The risk of each of the secondary outcomes was also similar in both groups.
Conclusions—About one third of the patients with a transient ischemic attack or minor ischemic stroke has both focal and nonfocal neurological symptoms. Nonfocal symptoms are not associated with an increased long-term risk of vascular events or death.
Clinical Trial Registration—This trial was not registered because enrollment began before July 1, 2005.
The clinical diagnosis of transient ischemic attack (TIA) or stroke is based on the presence of focal neurological symptoms of sudden onset.1,2 Nevertheless, in hospital-based series, 22% to 53% of patients with a TIA or minor ischemic stroke had accompanying nonfocal symptoms, such as nonrotatory dizziness or confusion.3,4 In a large population-based study, typical TIAs were accompanied by nonfocal symptoms in 12% of patients.5
Several studies have suggested that nonfocal neurological symptoms may be less innocuous than often thought. In the Dutch TIA Trial (DTT), patients with atypical transient neurological symptoms had a lower risk of stroke but higher risk of cardiac death or nonfatal myocardial infarction than patients with focal symptoms alone during a mean follow-up of 2.6 years.3 In the population-based Rotterdam study, patients with transient nonfocal symptoms of sudden onset had a higher risk of stroke than patients without these symptoms, and patients in whom focal symptoms were accompanied by nonfocal symptoms had a higher risk of myocardial infarction or vascular death than those with focal symptoms alone.5
The aim of this study was to assess whether the occurrence of nonfocal neurological symptoms also increases the risk of major vascular events during long-term follow-up in patients who present to a neurologist or stroke physician because of TIA or minor ischemic stroke.
Materials and Methods
Study Design and Patients
The present study is a post hoc analysis of the Life Long After Cerebral (LiLAC) ischemia study, which in turn was based on the DTT. The design and results of both studies have been published.6,7 Briefly, patients with a TIA or minor ischemic stroke in the previous 3 months were randomly allocated to doses of 30 or 283 mg of aspirin in the DTT between February 1986 and March 1989. The diagnosis of TIA or minor ischemic stroke was made by the treating neurologist at one of the participating hospitals. All patients had to be able to walk unassisted (score on modified Rankin scale, ≤3).8 Patients with a cardiac source of embolism or clotting disorder were excluded from participation. In the LiLAC study, follow-up was extended up to the period between March 2001 and December 2003 for patients who were alive at the close-out of the DTT in 1990. For practical reasons, the LiLAC study included only patients from the 24 hospitals that enrolled ≥50 patients in the DTT (2473 of the original 3150 patients). Patients with nonfocal symptoms alone were excluded from the present study because these would now not be classified as TIA.1 The DTT had been approved by the Ethics Committees of the participating hospitals. All patients provided informed consent for inclusion. The LiLAC study was approved by the Ethics Committee of the University Medical Center Utrecht.
At baseline, the participating neurologists interviewed all patients with multiple choice questions about the nature and time course of the neurological symptoms during the qualifying event. In the current study, we divided these symptoms into focal and nonfocal symptoms. This classification was based on that of the Rotterdam Study (Table 1). In contrast to the Rotterdam Study, the nonfocal symptom unwell feeling was not recorded in the DTT, and paresthesias occurring in the affected limb were considered a focal symptom.
Demographic and clinical characteristics, including the presence of cardiovascular risk factors, history of vascular disease, current medication, and the results of physical examination, laboratory investigations, ECG and brain computed tomography, were also recorded.
Until the close-out of the DTT in 1990, patients visited their neurologist every 4 months. In the LiLAC study, follow-up was obtained between 2001 and 2003 from all patients alive at the end of the DTT. Follow-up on the occurrence of outcome events was obtained from the neurologist who included the patient in the DTT, the general practitioner of the patient, direct contact with the patient, the relative or acquaintance whose name had been given at baseline, or the municipality register. If a possible vascular event had occurred, we retrieved hospital discharge letters and results of relevant radiology examinations.
The primary outcome was the composite of any stroke, myocardial infarction, or vascular death, whichever occurred first. Secondary outcomes were all-cause death, vascular death, cardiac death, myocardial infarction, and stroke. On the computed tomographic scan, infarcts were classified as any infarct, symptomatic lacunar infarct, symptomatic cortical infarct, or posterior circulation infarct.
We calculated the relative risk with corresponding 95% confidence intervals (CIs) for the presence of selected demographic and clinical characteristics in relation to nonfocal symptoms as part of the qualifying event. The cumulative risk of the outcome events in patients with or without nonfocal symptoms was estimated by means of the Kaplan–Meier method. Hazard ratios (HRs) were calculated by performing Cox proportional regression analyses. In multivariable analyses, adjustments were made for the 3 baseline factors that most affected the crude HR. Subgroup analyses were performed for patients presenting with a minor stroke versus a TIA as the qualifying event. Furthermore, we calculated the crude and adjusted HR for each outcome for patients with or without a specific nonfocal symptom (eg, confusion).
We excluded 64 (2.6%) patients from the LiLAC cohort who presented with nonfocal symptoms alone, leaving 2409 patients (Figure 1). The mean age of the patients was 65.2 years (SD, 10.2). Focal symptoms were accompanied by nonfocal symptoms in 739 (30.7%; 95% CI, 28.9–32.6) of the 2409 patients (Table 1). Characteristics of patients with or without nonfocal symptoms are presented in Table 2. Most baseline characteristics were equally divided between the 2 groups. Symptomatic lacunar infarct on computed tomography and smoking were more prevalent among patients with focal symptoms alone, whereas higher age (>66 years) and posterior circulation infarction were more frequent among those with nonfocal symptoms (Table 2).
Follow-up was complete for all but 19 (0.8%) participants.7 The mean duration of follow-up was 10.1 years (SD, 4.8) resulting in 24 319 patient-years of observation. The primary outcome occurred in 1313 (54.5%; 95% CI, 52.5–56.5) patients (Table 3); 1458 patients (60.5%; 95% CI, 58.6–62.5) died.
Kaplan–Meier curves and univariable Cox proportional hazard analysis show that patients with both focal and nonfocal symptoms had the same risk of the primary outcome during follow-up as patients with focal symptoms alone (HR, 0.99; 95% CI, 0.88–1·12; P=0.898; Figure 2; Table 4). After adjustment for age, history of intermittent claudication, and history of myocardial infarction, this risk remained essentially the same (adjusted HR, 0.97; 95% CI, 0.86–1.09; P=0.599). The risk of all-cause death (HR, 0.97; 95% CI, 0.87–1.09; P=0.248) and each of the other secondary outcomes was also similar in both groups in univariable and in multivariable analyses. Subgroup analyses, in which we compared patients presenting with either a minor ischemic stroke (n=1686) or a TIA (n=723), resulted in essentially the same results, with minimal differences in risk ratios and broadly overlapping 95% CIs (data not shown).
In univariable Cox proportional hazard analysis for the different nonfocal symptoms separately, patients presenting with confusion tended to have a higher risk of death, vascular death, and cardiac death than patients without confusion (Table 4). By contrast, patients presenting with paresthesias tended to have a lower risk of all-cause death and vascular death than patients without paresthesias. In multivariable analysis, only the risk of all-cause death (adjusted HR, 1.28; 95% CI, 1.02–1.60; P=0.030) and cardiac death (adjusted HR, 1.94; 95% CI, 1.04–3.61; P=0.037) in patients with confusion remained significantly increased.
We found that about one third of the patients with a TIA or minor ischemic stroke presents with a combination of focal and nonfocal neurological symptoms. The occurrence of nonfocal symptoms was not associated with a different risk of vascular events or death during a mean follow-up of more than 10 years. Only patients presenting with confusion had an increased risk of death from a cardiac or other cause.
Table II in the online-only Data Supplement shows an overview of the differences in design, inclusion, and exclusion criteria, definitions of nonfocal symptoms, and outcome for the original DTT, the Rotterdam study, and the current study. The proportion of patients with nonfocal symptoms during a TIA or minor ischemic stroke in the current study is lower than in a previous hospital-based study4 but considerably higher than in the population-based Rotterdam study.5 In the hospital-based study, 53% of the patients had accompanying nonfocal neurological symptoms.4 However, half of the patients in that study had posterior circulation TIA or stroke, and half of the patients had a stenosis of ≥50% of the carotid or vertebral artery. Both in our current study and in previous studies, nonfocal or atypical neurological symptoms were more frequent in patients with TIA or stroke in the posterior circulation than in the anterior circulation.4,9–11 In addition, nonfocal symptoms were more frequent in the presence of a stenosis in one of the supplying arteries.4 Both in our study and in previous studies, the occurrence of nonfocal symptoms was associated with a higher age.5,12
In the population-based Rotterdam study, focal symptoms were accompanied by nonfocal symptoms in just 12% of all patients with focal neurological symptoms. The fact that the combination of focal and nonfocal symptoms was considerably less frequent than in our study may be explained first of all by differences in the classification and timing of the assessment of symptoms. In the DTT, all patients were interviewed by a neurologist shortly after the TIA or stroke, whereas in the Rotterdam study, a research physician screened patients for the occurrence of neurological symptoms every 4 years, making this study more prone to recall bias. In addition, paresthesias occurring in the affected limb were considered a focal symptom in the current study but a nonfocal symptom in the Rotterdam study. In contrast to the Rotterdam Study, the nonfocal symptom unwell feeling was not recorded in the current study.
A second potential explanation for the difference in frequency of nonfocal symptoms between the Rotterdam study and the DTT is that the first was population based, whereas the second included patients referred to a neurologist because of a TIA or ischemic stroke, possibly resulting in selection bias toward patients with relatively serious symptoms.
We found no increased risk of vascular death or cardiovascular events in patients with nonfocal symptoms, in contrast to previous studies, partly from the same cohort.3,5,13 In a previous study, based on the full cohort of the DTT, patients presenting with a TIA or minor ischemic stroke and with atypical neurological symptoms had a higher risk of major cardiac events and a lower risk of stroke than patients with typical symptoms alone during a mean follow-up of 2.6 years.3 In the population-based Rotterdam study, patients with a combination of focal and nonfocal neurological symptoms had a higher risk of ischemic heart disease and vascular death than patients with focal symptoms alone.5
The most likely explanation for the difference in prognosis between patients with atypical TIAs in the full cohort of the DTT and those with the combination of nonfocal and focal symptoms in the present study is the difference in definition. In the first study, a tired or heavy sensation in ≥1 limbs, sensory symptoms alone, a gradual spread of sensory symptoms, and an isolated disorder of uncoordinated movements were considered atypical symptoms, whereas we now classified these symptoms as focal.5 In addition, patients with isolated nonfocal symptoms were excluded from the present study to limit the analysis to patients without a real TIA or minor stroke, whereas these were included in the first study. Considering that some isolated nonfocal symptoms, such as nonrotatory dizziness, may have a cardiac cause, the increased risk of cardiac events observed in the first study is not surprising. Nonfocal symptoms that occur in combination with focal symptoms are much less likely to have a cardiac cause other than embolism, which is supported by the fact that these were not associated with an increased risk of cardiac events in our study.
The difference in the risk of cardiac events between our study and the population-based Rotterdam study may be explained by the factors described above as potentially responsible for the difference in frequency of nonfocal symptoms.
It is remarkable that in our study, patients presenting with confusion had a higher risk of all-cause death and cardiac death than patients without confusion. Because of the large number of secondary analyses, this may be a chance finding, but a biological basis also seems conceivable. Confusion is clearly one of the hallmark symptoms of delirium, which in turn has been associated with an increased risk of death in patients with acute stroke and in elderly patients in general.14,15 An alternative explanation is that confusion may be a harbinger of hypoperfusion caused by cardiac dysfunction.
Strengths of our study are the large study population, the well-defined inclusion criteria, and the detailed and structured interview all patients had on the nature and time course of the neurological symptoms during the qualifying TIA or minor ischemic stroke. In addition, the observation period was more than 10 years on average, with few patients being lost to follow-up.
This study is inherently limited by the retrospective data collection of events after the close-out of the DTT. In addition, follow-up ended in 2003 and since then more aggressive regimes of medical management have been implemented. However, this should probably have equally affected both patients with and without nonfocal symptoms.
In contrast to older hospital-based studies and a recent population-based study, we found no evidence that nonfocal symptoms in patients with a TIA or minor ischemic stroke seen in a hospital are associated with an increased long-term risk of vascular death, stroke, or myocardial infarction.
We gratefully acknowledge the contribution of the Dutch TIA Trial and Life Long After Cerebral ischemia Study Groups.
Sources of Funding
Drs Compter and van der Worp are supported by grants from the Dutch Heart Foundation (2007/B045 and 2010T075, respectively). The Life Long After Cerebral (LiLAC) study was supported by the Dutch Heart Foundation (99/160) and the Hersenstichting (10F02/14).
Dr van der Worp has received speaker’s fees from Sanofi Aventis and GlaxoSmithKline and has served as a consultant to Bristol-Myers Squibb.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.004360/-/DC1.
- Received December 4, 2013.
- Revision received February 18, 2014.
- Accepted February 19, 2014.
- © 2014 American Heart Association, Inc.
- Easton JD,
- Saver JL,
- Albers GW,
- Alberts MJ,
- Chaturvedi S,
- Feldmann E,
- et al
- van Swieten JC,
- Koudstaal PJ,
- Visser MC,
- Schouten HJ,
- van Gijn J
- Bots ML,
- van der Wilk EC,
- Koudstaal PJ,
- Hofman A,
- Grobbee DE
- Shi Q,
- Presutti R,
- Selchen D,
- Saposnik G