Prognostic Factors for Long-Term Risk of Stroke After Transient Ischemic Attack or Minor Stroke: A Systematic Review and Meta-Analysis
Clinical Snapshot
PICO Framework
| P — Population | Patients with transient ischemic attack (TIA) or minor stroke |
| I — Intervention | Various prognostic factors (age, sex, comorbidities, clinical features, imaging findings) |
| C — Comparator | Absence of prognostic factors or reference categories |
| O — Outcomes | Long-term risk of subsequent stroke (≥1 year follow-up) |
Bottom Line
This high-quality systematic review of 86,810 patients identifies key prognostic factors for long-term stroke risk after TIA or minor stroke. The strongest predictors include large-artery atherosclerosis (HR 2.19), cardioembolism (HR 2.16), and acute infarct on neuroimaging (HR 1.97). Modifiable factors like hypertension contribute 19.3% population attributable risk. Male sex, diabetes, smoking, and clinical presentations with aphasia or paresis also significantly increase risk. These findings support intensive risk factor modification and prolonged monitoring in high-risk patients. The ABCD2 score ≥4 remains a useful clinical tool (HR 1.59). Australian clinicians should use these evidence-based factors to stratify patients for targeted secondary prevention strategies, including optimised antiplatelet therapy, anticoagulation for atrial fibrillation, and aggressive cardiovascular risk factor management. The large international dataset and GRADE methodology provide high certainty evidence for clinical decision-making.
Key Findings
P Value: Not explicitly reported but implied significant given CI exclusion of 1.0
Effect Size: Hazard ratios ranging from 1.25 (male sex) to 2.19 (large-artery atherosclerosis)
Primary Outcome: Long-term stroke risk after TIA or minor stroke
Nnt Or Sensitivity: Population attributable fractions: male sex 13.0%, hypertension 19.3%, minor stroke vs TIA 28.0%
Confidence Interval: All 95% CIs exclude null value for significant predictors
Clinical Application
Highly feasible - all prognostic factors are routinely assessed in Australian clinical practice Directly applicable to Australian stroke prevention guidelines, supports TGA-approved anticoagulants for AF, aligns with RACGP cardiovascular risk assessment, and relevant for PBS-subsidised secondary prevention medications Australian patients with TIA or minor stroke presenting to emergency departments, stroke units, and neurology services
Abstract
BACKGROUND: Patients with a transient ischemic attack (TIA) or minor stroke have an increased risk of subsequent stroke that persists for at least 10 years. We aimed to identify prognostic factors associated with long-term risk of stroke in this patient group and estimate their population attributable fraction (PAF). METHODS: A systematic review was performed of MEDLINE, Embase, and Web of Science for cohort studies including patients with TIA or minor stroke that evaluated factors for subsequent stroke over a follow-up period of ≥1 year. We pooled hazard ratios adjusted for relevant confounders using random-effect meta-analysis and determined the PAF of factors based on their pooled prevalence and adjusted hazard ratio. We assessed certainty of evidence using the grading of recommendations, assessment, development, and evaluation approach. The study is registered in PROSPERO (CRD42023476551). RESULTS: From 14 732 identified citations, we included 28 cohort studies comprising 86 810 patients with TIA or minor stroke (median age, 69 years [interquartile range, 65-71]; 52%-60% male patients). Factors that had high certainty evidence of association with increased long-term risk of stroke included older age (adjusted hazard ratio, 1.04 per year increase [95% CI, 1.02-1.05]); male sex (1.25, 1.15-1.36; PAF 13.0% [95% CI, 7.8-18.7]); atrial fibrillation (1.34, 1.18-1.52; 3.8% [95% CI, 0.3-9.9]); diabetes (1.52, 1.32-1.75; 7.7%, 3.1-14.1); hypertension (1.60, 1.31-1.94; 19.3%, 8.4-31.6); ischemic heart disease (1.67, 1.28-2.18; 10.7%, 2.8-22.9); history of stroke or TIA before the index event (1.70, 1.43-2.02; 12.0%, 5.2-21.4); smoking (1.29, 1.05-1.60; 11.2%, 1.0-30.7); ABCD2 (age, blood pressure, clinical features, duration of TIA, and presence of diabetes) score of ≥4 (1.59, 1.31-1.94; 18.0%, 2.9-39.9); presence of acute infarct on neuroimaging (1.97, 1.41-2.74; 19.0%, 5.2-38.9), including diffusion-weighted imaging-positive lesions (1.86, 1.02-3.37; 14.0%, 7.0-25.0); minor stroke as index event (1.75, 1.35-2.27 versus TIA; 28.0%, 10.2-47.6); presentation with aphasia or dysarthria (1.45, 1.24-1.69; 19.2%, 0.2-53.1); presentation with paresis (1.45, 1.15-1.84; 22.0%, 3.8-43.4); and etiologic stroke subtypes, including cardioembolism (2.16, 1.53-3.05; 14.6%, 3.1-33.5), large-artery atherosclerosis (2.19, 1.68-2.86; 13.2%, 5.1-25.5), and small-vessel disease (1.69, 1.14-2.49; 16.8%, 5.0-34.3). CONCLUSIONS: These findings can help identify patients with a particularly enduring risk of stroke who are most likely to benefit from ongoing monitoring and treatment and facilitate the development and implementation of targeted stroke prevention strategies.
References
- 1.Khan, F., Yogendrakumar, V., Lun, R., Marx, C. E., Rochwerg, B., Tran, A., Fernando, S. M., Ganesh, A., Barber, P. A., Ögren, J., Ois, A., Giralt-Steinhauer, E., Khanevski, A. N., Leng, X., Tian, X., Leung, T. W., Verburgt, E., Verhoeven, J., de Leeuw, F. E., Ildstad, F., Fandler-Höfler, S., Aarnio, K., von Sarnowski, B., Lorenzetti, D. L., Coutts, S. B., Amarenco, P., Hankey, G. J., & Hill, M. D. (2026). Prognostic factors for long-term risk of stroke after transient ischemic attack or minor stroke: A systematic review and meta-analysis. *Circulation*. https://doi.org/10.1161/CIRCULATIONAHA.125.078763
Related Research
Sleep medicine reviews
Machine and deep learning in REM sleep behavior disorder: a scoping review and analysis of reporting quality
3 Aug 2026
Neurological research
Federated deep learning model for epilepsy seizure detection using electroencephalogram (EEG) signal
2 Aug 2026
Journal of psychiatric research
Systematic review of machine learning and deep learning models for EEG-based detection of depression.
2 Aug 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service