The relationship between ambient ultraviolet radiation (UVR) and objectively measured personal UVR exposure dose is modified by season and latitude

Despite the widespread use of ambient ultraviolet radiation (UVR) as a proxy measure of personal exposure to UVR, the relationship between the two is not well-defined. This paper examines the effects of season and latitude on the relationship between ambient UVR and personal UVR exposure. We used data from the AusD Study, a multi-centre cross-sectional study among Australian adults (18-75 years), where personal UVR exposure was objectively measured using polysulphone dosimeters. Data were analysed for 991 participants from 4 Australian cities of different latitude: Townsville (19.3°S), Brisbane (27.5°S), Canberra (35.3°S) and Hobart (42.8°S). Daily personal UVR exposure varied from 0.01 to 21 Standard Erythemal Doses (median = 1.1, IQR: 0.5-2.1), on average accounting for 5% of the total available ambient dose. There was an overall positive correlation between ambient UVR and personal UVR exposure (r = 0.23, p <0.001). However, the correlations varied according to season and study location: from strong correlations in winter (r = 0.50) and at high latitudes (Hobart, r = 0.50; Canberra, r = 0.39), to null or even slightly negative correlations, in summer (r = 0.01) and at low latitudes (Townsville, r = -0.06; Brisbane, r = -0.16). Multiple regression models showed significant effect modification by season and location. Personal exposure fraction of total available ambient dose was highest in winter (7%) and amongst Hobart participants (7%) and lowest in summer (1%) and in Townsville (4%). These results suggest season and latitude modify the relationship between ambient UVR and personal UVR exposure. Ambient UVR may not be a good indicator for personal exposure dose under some circumstances.