Test–retest repeatability of assessing environmental and lifestyle factors in Parkinson's disease

Abstract

Epidemiological studies of environmental risk factors in Parkinson's disease (PD) are dependent on recollection of past exposures based on patients' self-reports. There are limited studies that have assessed the quality of such data. We conducted a prospective study to determine the test-retest repeatability of environmental and lifestyle factors, and medical data in a PD cohort of Asian ethnicity. A total of 150 consecutive PD patients were initially screened, and 100 were recruited and completed an initial interview. Eighty-three patients completed the second interview more than 6 months later. Lifestyle habits (such as smoking and coffee consumption) showed excellent agreement (kappa > 0.90). For the amount and duration of coffee, tea, alcohol, and cigarette smoking exposure, the total agreement in the response for these factors in the repeat interview were noted in 71.4%, 73.3%, 100%, and 90%, respectively (ICC > 0.83). Medical conditions for which the patients were on treatment, such as diabetes, hypertension, and stroke, revealed very high repeatability (kappa = 0.81-0.90). Environmental exposures like well-water consumption and prior farm-dwelling produced a moderately good repeatability (kappa = 0.66-0.77). In conclusion, our study demonstrates that even over long interval period of more than half a year, self-report lifestyle exposure information, personal and environmental exposure data can be collected with moderate-to-high repeatability from PD patients.