Regional and racial differences in the prevalence of physician-diagnosed essential tremor in the United States.

For reasons that are unclear, prevalence estimates of essential tremor (ET) differ considerably across the United States. Separate communities have never been sampled within the framework of the same study to substantiate these differences. We estimated the prevalence of physician-diagnosed ET in the elderly in four communities in the United States in whom the same screening questions were used, and examined whether this prevalence differed between Caucasians and African Americans. The Cardiovascular Health Study recruited a sample of Medicare beneficiaries >/=65 years of age from four communities in different regions of the United States. In 1998 to 1999, 3,494 participants (mean age, 80.0 years; range, 70-103 years) answered a 12-question screen for ET, including the question, "has a doctor diagnosed you as having familial tremor or benign essential tremor?" Fifty-four participants reported that a doctor had diagnosed them as having ET (1.5%; 95% confidence interval, [CI], 1.1-2.0%). Prevalence was similar across the four communities (1.1-2.0%). A larger proportion of Caucasians than African Americans reported a diagnosis of ET (1.7% vs. 0.4%; odds ratio = 4.9; 95% CI, 1.2-20.2; P = 0.028). In a logistic regression analysis, physician-diagnosed ET was associated with Caucasian ethnicity (P = 0.038) but not with age, gender, education, mental status or depression scores, income, smoking status, or alcohol consumption. When a standardized screening question was used, the proportion of participants with physician-diagnosed ET was similar across four communities, suggesting that the prevalence of this condition may be less variable than is often reported. Caucasians were five times more likely to have physician-diagnosed ET than were African Americans. This study does not provide an explanation for this difference, which deserves further study.