Abstract 188: Association Between Uric Acid and Stroke in the REgards Case-cohort Study

Abstract

Introduction: The association between hyperuricemia and stroke has been inconsistent especially after adjustment for cardiovascular risk factors. Previous studies were limited in the number of events and did not measure these associations in subgroups. Methods: We used a stratified case-cohort design within the REasons for Geographic and Racial Differences in Stroke study. Clinically adjudicated stroke events (n=903) were defined as focal neurologic deficit lasting >24 hours or non-focal neurological symptoms with brain imaging consistent with stroke. Controls (n=951) were a random stratified sample selected from baseline to ensure representation of high-risk groups. Uric acid was assayed using the baseline sample and categorized into three groups: < 6 (referent), 6-6.8, ≥ 6.8 mg/dl (hyperuricemia). Cox-proportional-hazard-models adjusted for demographic and clinical variables were fit to examine the association between uric acid levels and stroke. We repeated the analysis stratified by race, gender, and age (</≥ 65years) and finally, since uric acid is associated with hypertension, explored the count of antihypertensive treatment classes (nHTN) as a potential mediator. Results: Hyperuricemic individuals were more likely to be male and black. Hyperuricemia versus the referent category was significantly associated with stroke after adjustment for race, sex, age, and age*race interaction [model 1 HR(95%CI)= 1.42(1.12-1.80)], and after further adjustment for systolic and diastolic blood pressure [HR (95%CI)= 1.42(1.12-1.80)]. The associations was attenuated after full adjustment [HR (95%CI)= 1.22(0.91-1.63)]. Incremental adjustment by clinical variables suggested nHTN attenuated the association between hyperuricemia and stroke. Results were similar for hyperuricemia within subgroups defined by age, gender and race, except among men <65 years [HR (95%CI)= 2.72(1.25-5.93)]. Mediation analysis estimated that nHTN accounted for 43%(95%CI: 15 to 158%) of the relationship. Conclusion: We observed a significant association between hyperuricemia and stroke that was partially dependent upon hypertension severity (count of antihypertensive treatment classes). The association was strongest among males aged <65 years.