Low cholesterol, statins and outcomes in patients with first-ever acute ischemic stroke

Background: High cholesterol has been assiated with better stroke outcomes. Conversely, a protective effect of prestroke statin use in the acute phase of stroke has been reported. The effect of low cholesterol on outcome in patients with and without prestroke statin treatment has not been studied. We assessed the association between low cholesterol and ischemic stroke short- and long-term outcomes and studied potential interactions with statin treatment in patients with a first-ever ischemic stroke in a prospective national registry. Methods: Ischemic stroke patients in the National Acute Stroke Israeli (NASIS) registry with a first-ever stroke and no previous disability, dementia or cancer admitted in all hospitals nationwide during 2-month periods in 2004, 2007 and 2010 were included (n = 1,895). Cholesterol levels ≤155 mg/dl (1st quintile) were defined as low cholesterol and patients treated with statins for at least 7 days before stroke onset were categorized as prestroke statin treated. Severe stroke (NIHSS ≥11), total anterior circulation infarction, poor functional outcome (defined as discharged to a nursing facility or modified Rankin Scale >3 or death), and mortality at discharge and at 3 years were the study outcomes. Associations between low cholesterol and outcomes at discharge were assessed separately in patients with and without prestroke statin treatment using multiple logistic regression analyses. Mortality at 3 years was assessed in a subset of 681 patients with Cox proportional hazard models. Results: Patients were 67.4 ± 13.5 years old on average; 43.1% were women. Low cholesterol was associated with higher rates of stroke risk factors. Controlling for age, sex, hypertension, diabetes, current smoking, ischemic heart disease, congestive heart failure and atrial fibrillation, low cholesterol was significantly associated with severe stroke, total anterior circulation infarction and poor functional outcome in patients with and without statin treatment. There were no interactions between low cholesterol and prestroke statin therapy in association with outcomes. Short- and long-term mortality rates were increased for patients with low cholesterol (5.2% at discharge and 35% at 3-years) compared with higher levels (2.5% at discharge and 20.5% at 3 years). Adjusted mortality risks were increased for patients with low cholesterol; however, this finding was statistically significant only for patients not on statins before the stroke. Conclusions: Low cholesterol is associated with increased stroke severity and poorer functional outcome in patients with and without prestroke statin use. Low-cholesterol statin-naive patients show increased risks of mortality. 'Reverse epidemiology' in the association between cholesterol and outcome is possible in patients with ischemic stroke.