Combining Novel With Traditional Risk Factors Improves Stroke Prediction In Older Americans: A Systematic Review

Background: The Framingham Risk Score, one validated model for predicting stroke risk in middle age does not perform as well in older cohorts who may have different/novel stroke risk factors. Methods: Two independent raters systematically reviewed (through July 31, 2016) PubMed and Embase, identifying studies including peer-reviewed, original, clinical studies, randomized controlled trials and cited references to assess independent risk factors for first ischemic stroke in older, (u003e 65 years) Americans using PRISMA guidelines. Results: 467 relevant studies were identified. We extracted data from 42 papers and categorized their aims into 6 sections: Serologic/Diagnostic, Traditional, Psychosocial, Genetic, Cognitive, and Miscellaneous risk factors. Some studies fell into more than one category. Serologic/Diagnostic (11 studies) demonstrated: levels of serum androgens, C-reactive protein, advanced glycation endproducts, thrombin generation and increased left ventricular mass were significant risk factors; adipokines and plasma free fatty acids were not. Traditional (11 studies) risk factors included: age, hypertension (both systolic and diastolic), atrial fibrillation, diabetes, peripheral vascular disease, waist/hip ratio in women, and BMI in men. Two studies showed that risk associated with hypertension and diabetes diminishes significantly over age 75. One study showed the protective effect of physical activity increased with age. Psychosocial (13 studies): 6 studies found depression strongly associated with stroke risk. Also, positive affect and purpose in life protected against stroke. Socioeconomic status did not influence risk in 3 studies. Genetic (5 studies) included: Apolipoprotein E, coagulation factor XII, and phosphodiesterase 4D polymorphisms – all with equivocal results. Cognitive (4 studies): Lower cognitive functioning, increased stroke rates. Miscellaneous (4 studies): Decreased level of functioning and increased cadmium exposure were risk factors. Conclusion: Some traditional and other more novel risk factors emerged as increasing stroke risk in older individuals. Only a few studies evaluated each risk factor. Further studies are warranted to investigate risk factors for stroke in older cohorts.