Abstract 3559: Differential Effect of High Cholesterol by Intracerebral Hemorrhage Location

Abstract

Background: High cholesterol has been associated with a reduced likelihood of intracerebral hemorrhage (ICH). Lobar ICHs that arise from brain parenchyma near the cortex are predominantly related to amyloid angiopathy and thus have a different risk factor profile than non-lobar ICHs for which the main risk factor is hypertension. We were interested in whether high cholesterol was associated with reduced likelihood of all ICH, or whether there was a differential effect by hemorrhage location. Methods: The Genetic and Environmental Risk Factors for Hemorrhagic Stroke Study is a prospective case-control study of hemorrhagic strokes in the Greater Cincinnati/Northern Kentucky region. Charts are abstracted by nurses and verified by physician review to identify cases. Controls selected by random-digit dialing are matched to each case by age (±5 years), race, and sex. Data is collected through structured interview by study nurses within 90 days of onset; history of high cholesterol is determined by self-report. Hemorrhage location is classified by physician review as either lobar or non-lobar (deep gray matter, brainstem, cerebellum, or pure intraventricular hemorrhage). Results: Of the 2,850 ICH cases identified from 5/98 to 11/06, 995 died prior to nurse contact, 396 declined interview, and 784 were not contacted within 90 days. To restrict the analysis to spontaneous, first-ever ICH cases among whites and blacks, 78 subjects were excluded due to prior ICH, structural cause of ICH, or “other” race. The remaining 597 subjects underwent direct interview and were matched to 1-3 controls base on ethnicity, age and gender. Frequency of high cholesterol history was 34% for all ICH cases vs. 43% in controls; 40% for lobar cases vs. 43% in their controls; and 31% in non-lobar cases vs. 42% in their controls. Multivariable logistic regression, controlling for hypertension, warfarin use, first-degree relative with ICH, education level, prior ischemic stroke, and Apolipoprotein E genotype, revealed odds ratios and 95% confidence intervals for high cholesterol of 0.57 (0.45-0.72) for all ICH, 0.45 (0.34-0.61) for non-lobar ICH, and 0.85 (0.57-1.26) for lobar ICH. Conclusions: High cholesterol is associated with a lower risk of ICH, and this appears to be driven primarily by its association with non-lobar ICHs. This novel finding suggests that hypercholesterolemia may be associated with reduced risk of hypertensive ICH.