Increased levels of solar radiation are associated with reduced type-2 diabetes prevalence: A cross-sectional study of Australian postcodes

Type-2 diabetes is a leading cause of death and disability. Emerging evidence suggests that ultraviolet radiation or sun exposure may limit its development. We used freely available online datasets to evaluate the associations between solar radiation and type-2 diabetes prevalence across Australia. We extracted prevalence data for 1822 postcodes from the Australian Diabetes Map on 25 January 2020. Daily solar radiation data averaged over 30-years (1990-2019) were collated from online databases (Australian Bureau of Meteorology). Population-weighted linear regression models were adjusted for covariates at the postcode level including socioeconomic status (IRSAD), remoteness, mean age, gender, Aboriginal and Torres Strait Islander status, as well as mean annual ambient temperature (1961-1990) and rainfall (1981-2010). A consistent inverse correlation was observed between type-2 diabetes prevalence and solar radiation, after adjusting for these covariates (ss (coefficient of regression) = -0.045; 95% CI: -0.086, -0.0051; p = 0.027). However, the relative contribution of solar radiation towards type-2 diabetes prevalence was small (2.1%) in this model. Other significant correlations between type-2 diabetes prevalence and covariates included: socioeconomic status (ss = -0.017; 95% CI: -0.017, -0.016; p < 0.001), mean age (ss = 0.041; 95% CI: 0.028, 0.054; p < 0.015), remoteness (ss = -0.05; 95% CI: -0.088, -0.011; p < 0.001) and rainfall (ss = -0.0008; 95% CI: -0.00097, -0.00067; p < 0.001). In conclusion, in Australian postcodes, higher levels of solar radiation and rainfall was associated with reduced type-2 diabetes prevalence. Further studies are needed that consider lifestyle covariates such as physical activity.

Automated summary

Higher levels of solar radiation and rainfall in Australian postcodes were associated with reduced prevalence of type-2 diabetes. Solar radiation showed a small relative contribution to the prevalence, while other covariates such as socioeconomic status, mean age, remoteness, and rainfall also showed significant correlations with the prevalence.