Validity of probands' reports and self-reports of essential tremor: Data from a large family study in North America.

The search for genes for essential tremor (ET) is active. Researchers often depend on probands' reports or self-reports to assign disease status to relatives. Yet there are surprisingly few data on the validity of these reports. In two prior studies, with small sample sizes, validity was poor (sensitivity = 16.7-43.3%). In the current study, ET probands and their relatives were screened for tremor and then underwent a videotaped in-person neurological examination. One investigator then assessed the screening questionnaires and videotapes to assign diagnoses of ET, borderline tremor or other diagnosis. There were 98 probands and 243 relatives (105 with ET, 34 with borderline tremor). Educational attainment was high (15.6 ± 2.7 years). Probands failed to report tremor in 39/139 relatives with ET or borderline tremor; conversely, they reported tremor in 32/104 relatives without ET or borderline tremor. Thus, in total, there were 71/243 (29.2%) mis-identifications. Thirty six of 139 ET and borderline ET cases failed to self-report tremor; conversely, 30/104 relatives without ET or borderline tremor self-reported tremor. Thus, in total, there were 66/243 (27.2%) mis-identifications. In summary, in individuals with greater educational attainment, the validity of reported information on ET was considerably higher than previously reported. Despite this, even among well-educated individuals in North America, probands' reports and self-reports misclassified approximately 30% (i.e., one-in-three) of relatives.