Ocular Exposure to UV-B in Sunlight: The Melbourne Visual Impairment Project Model
McCarty, C. A., Lee, S. E., Livingston, P. M., Bissinella, M., Taylor, H. R., Bulletin of the World Health Organization
Clinicians, public health researchers, and programme planners have recently become interested in personal exposure to sunlight and ultraviolet-B (UV-B) radiation as preventable risk factors for sunburn, melanoma, and a number of ocular disorders (1).(a) Epidemiological studies of the association between cataract and personal ocular UV-B exposure in different geographical locations should permit investigation of the relationship between a thinning ozone layer (and hence increased UV-B exposure) and higher prevalence of cataract and lens opacities. As a consequence of the Earth Summit in Brazil in 1992, WHO and the International Agency for Research on Cancer (IARC) proposed that a multicentre study of the effect of environmental change on solar UV-B radiation and health be set up, with one component being the measurement of the ocular effects of UV-B radiation (3). The information obtained could be used to develop policies and programmes for the prevention of adverse ocular effects arising from UV-B exposure.
Two large epidemiological studies of eye disease in the USA collected data on UV-B exposure and demonstrated a positive association between lens opacities and UV-B exposure (3, 4). Of these, the Chesapeake Bay Waterman Study quantified UV-B exposure in "Maryland sun years" using a model that includes information on history of work activity, leisure activity, wearing of spectacles, and hat use, along with field and laboratory measurements of UV-B radiant exposure (5), while in the Beaver Dam Eye Study, a sunlight exposure variable was calculated based on a "Wisconsin sun year" (3). A simplified model for assessing ocular exposure to UV-B radiation has also been developed and evaluated (6). This model includes ocular UV-B exposure during the middle of the day (09:00 to 15:00) over the northern hemisphere summer months (April-September). The results correlated highly with the full model (r = 0.98), while predicting 62% of total ocular exposure and requiring far less participant's time. Such a model is useful for epidemiological studies if relative rather than absolute UV-B exposure is the variable of interest, and if limited resources preclude the implementation of a longer, more complete model. The aim of the present investigation was to extend this work and to develop further a questionnaire and model for the assessment and quantification of UV-B exposure in participants in population-based epidemiological studies of eye disease. This work was undertaken specifically for the Melbourne Visual Impairment Project (Melbourne VIP). The model is based on the abbreviated model previously tested in the Chesapeake Bay Waterman study (6), but includes a measure of childhood exposure and more information about leisure-time behaviour.
The model for personal ocular exposure to UV-B (see Annex) is based on a questionnaire that includes the following questions (Table 1):
* For each period of working life greater than 6 months, where did you live, what was your occupation, and how many hours between 09:00 and 17:00 (daylight savings) did you spend outdoors on weekdays and at the weekend during the warmer months, from mid-spring to mid-autumn?
* Did you work over water during that period?
* During that time spent outdoors, how often did you wear a hat, sunglasses or prescription spectacles/ contact lenses on weekdays and at the weekend?
[TABULAR DATA 1 OMITTED]
One piece of information included in the abbreviated model was not present in the Melbourne VIP questionnaire. This was the number of days worked per week, because in the majority of cases, workweeks comprise 5 days regardless of which days.
Additional information on the length of time lived at all different locations was gathered for school life-periods. Other than location in childhood, it was believed that the majority of variation in cumulative lifetime ocular UV-B exposure is due to behaviours in adulthood and that childhood exposures are likely to be generally similar among people in a given environment. …