Background: This study analyzes Chicago-area weather effects on objectively measured physical activity over a 3-year period among a cohort of 241 participants in an on-going arthritis physical activity trial. Methods: Uniaxial accelerometer counts and interview data were analyzed for up to 6 weekly study waves involving 4823 days of wear. The effects of temperature, rainfall, snowfall and daylight hours were analyzed after controlling for participant characteristics, day of the week, and daily accelerometer wear hours in a mixed effects linear regression model. Results: Daylight hours, mean daily temperature < 20 or ≥ 75 degrees, and light or heavy rainfall (but not snowfall) were all significantly associated with lower physical activity after controlling for the significant effects of weekends, accelerometer wear hours, age, sex, type of arthritis, employment, Hispanic ethnicity, obesity, and SF36 physical and mental health scores. Conclusions: The cumulative effects of weather are reflected in a 38.3% mean monthly difference in daily counts between November and June, reflecting over 3 additional hours of sedentary time. Physical activity promotion programs for older persons with chronic conditions need lifestyle physical activity plans adapted to weather extremes.
Keywords: accelerometry, arthritis, physical activity measurement, meteorology