Abstract
New Zealand has one of the highest mortality rates from cutaneous malignant melanoma in the world. Skin cancer has been linked to sunburn and excessive exposure to UVR. For this reason, both national and regional health promotion programmes have been implemented in New Zealand since 1988. The aim of these campaigns was to decrease harmful exposure to the sun by encouraging the public to engage in sun safe behaviours. The Cancer Society of New Zealand Inc. and the Health Sponsorship Council have undertaken the Triennial Sun Protection Survey series (Sun Survey), to inform and guide the New Zealand skin cancer control programme, since 1994. The Sun Survey was an ongoing computer assisted telephone interview (CATI) survey of residents (15–69 years) in five New Zealand metropolitan areas (Auckland, Wellington, Hamilton, Christchurch, and Dunedin), conducted every three years during the southern hemisphere summer. Interviews were conducted on either the Monday or Tuesday evening in randomly selected households following selected survey weekends. Data were obtained from 6,195 respondents: 1994 (n=1,243), 1997 (n=1,188), 1999/00 (n=1,250), 2002/03 (n=1,250), and 2005/06 (n=1,264). A conceptual model was developed to aid the research, based on both the literature review and a previous conceptual model developed by Hill and Boulter (1996). Using the Sun Survey data in conjunction with climate and UVR data from the National Institute of Water & Atmospheric Research, matched to the respondents' interview day, this thesis had two main research foci: (1.) to describe the patterns of the outcome variables (duration outdoors, body coverage, and sunburn) across the survey years, 1994–2005/06 and (2.) to investigate predictors of those outcome variables described in (1.) using statistical modelling techniques, controlling for the respondents' current weather conditions, individual characteristics and behaviours. These analyses were undertaken on a reduced dataset of survey years 1999/00, 2002/03, and 2005/06. Attempts to tan have decreased and sunscreen utilisation has increased. However, no meaningful patterns that indicated the SunSmart programme goals have been achieved regarding reduced outdoor UVR exposure, increased body coverage, or decreased sunburn occurrence during peak UVR in summer months were found. Consistent with the literature review and in support of the conceptual model, the predictive models indicated that variables showing common associations with all three outcome variables were survey month, main activity undertaken, and utilisation of available shade. The contribution this thesis makes is: (1.) the development of a revised conceptual model of sunburn, (2.) the description of patterns across the survey years (1994–2005/06) in the areas of outdoor duration, body coverage and sunburn and (3.) the identification of predictors in these areas that may be addressed in future public health campaigns. In conclusion, this New Zealand study was the first assessment of patterns and predictors of duration outdoors, body coverage, and sunburn among respondents aged 15–69 years. Caution needs to be taken regarding generalisability of the results, due to the sampling technique, low response rates and missing data. However, this thesis may elucidate new research challenges and help guide future efforts to reduce skin cancer incidence.