Association of climatic determinants with Type 1 and Type 2 Diabetes worldwide: Night length and photoperiod variation linked to T1D and sunshine to T2D

Nearly 500 million individuals are affected by diabetes worldwide. This very high prevalence is combined with a North-South gradient and a seasonality of diagnostics which all suggest the role of climate in diabetes etiology. However, only little is known about the impact of climate on diabetes. This article aims to understand the association of climatic variables with type 1 and type 2 diabetes (T1D and T2D) for 72 countries worldwide (1989-2021). T1D is, on average, more prevalent at extreme latitudes whereas T2D prevalence is higher near equator ( P < 0,001). Sunshine, temperature, solar irradiance and daylength (photoperiod) are negatively associated with T1D prevalence and positively associated with T2D in simple regression ( P < 0,001). Multicollinearity of climatic variables is considered as a challenge, and it is assessed with VIF and optimized with multiple regression. After adjustment, only photoperiod is associated with T1D prevalence (r 2 =0,45) and sunshine with T2D prevalence (r 2 =0,48). T1D monthly incidences are approximated with a cosine regression (RR=1,53) which is significantly associated with photoperiod along the year in Europe ( P < 0,05). The relation between photoperiod and T1D has never been reported before in an ecological study and a short review is developed in the discussion.