Background: Stroke is a leading cause of death and disability worldwide. The risk factors for stroke overlap those for cardiovascular disease. Atrial fibrillation (AF) is a particularly strong risk factor and is common, particularly in the elderly. Maccabi Healthcare Services (MHS) has maintained a vascular registry of clinical information for over 100,000 members, among them patients with heart disease and stroke.
Objectives: To determine the prevalence of stroke in MHS, and whether the association of AF and stroke, along with other risk factors, in the Maccabi population is similar to that in published studies.
Methods: Data on stroke and AF patients aged 45 and older were collected from the database for the year 2010, including age, previous transient ischemic attack (TIA), body mass index (BMI), prior myocardial infarction (MI), diabetes, hypertension, anticoagulation and dyslipidemia. A cross-sectional analysis was used to estimate stroke prevalence by AF status. A case-control analysis was also performed comparing a sample of stroke and non-stroke patients. This permitted estimation of the strength of associations for atrial fibrillation and various other combinations of risk factors with stroke.
Results: Stroke prevalence ranged from 3.5 (females, age 45–54 years) to 74.1 (males, age 85+) per thousand in non-AF members, and from 29 (males, age 45–54) to 165 (males, age 85+) per thousand for patients with AF. AF patients had significantly more strokes than non-AF patients in all age groups. Stroke prevalence increased with age and was significantly higher in males. Multivariable analysis revealed that male gender, increasing age, AF, hypertension, diabetes, and history of TIA were highly significant risk factors for stroke. In addition, for males, dyslipidemia and prior MI were moderately strong risk factors.
Conclusions: Analysis of the MHS vascular database yielded useful information on stroke prevalence and association of known risk factors with stroke, which is consistent with the epidemiological literature elsewhere. Further analysis of health fund data could potentially provide useful information in the future.