Reliability of Self-Report Data on Age of Onset in Essential Tremor: Data from a Prospective, Longitudinal Cohort

Abstract

<b><i>Background:</i></b> Essential tremor (ET) is a highly prevalent neurological disease. Age of onset can occur anytime between childhood and advanced age. Tremor generally starts insidiously. Age of onset is a particularly important data item in clinical and epidemiological research. In general, these data are self-reported by ET cases. A fundamental question is whether ET cases reliably report their age of onset. <b><i>Methods:</i></b> In this prospective, epidemiological study of 125 ET cases, self-reported age of onset data were collected at regular 18 months intervals over four time points. <b><i>Results:</i></b> The correlation between self-reported age of onset was high – intra-class correlation coefficient = 0.972 (95% confidence interval = 0.962–0.980, <i>p</i> &#x3c; 0.001). However, agreement was not perfect. Approximately 20–25% of participant’s reports at different time intervals differed by as much as 10 years, and approximately 10% of participant’s reports differed by as much as 20 years. <b><i>Conclusions:</i></b> There was a robust correlation between self-reports of age of onset. Yet in a not-insignificant number of cases, there were considerable differences, some of which were substantial. These findings have broad implications for development of diagnostic algorithms, data stratification schemes, and analyses that assess correlations between biomarker data and clinical features (e.g., disease duration).