Identifying heat thresholds for South Africa towards the development of a heat-health warning system

Thandi Kapwata; Nada Abdelatif; Noah Scovronick; Michael T Gebreslasie; Fiorella Acquaotta; Caradee Y Wright

doi:10.1007/s00484-023-02596-z

Identifying heat thresholds for South Africa towards the development of a heat-health warning system

Int J Biometeorol. 2024 Feb;68(2):381-392. doi: 10.1007/s00484-023-02596-z. Epub 2023 Dec 29.

Authors

Thandi Kapwata^{1

2}, Nada Abdelatif³, Noah Scovronick⁴, Michael T Gebreslasie⁵, Fiorella Acquaotta⁶, Caradee Y Wright^{7

8}

Affiliations

¹ Environment and Health Research Unit, South African Medical Research Council, Johannesburg, 2028, South Africa. Thandi.Kapwata@mrc.ac.za.
² Department of Geography, Geoinformatics and Meteorology, University of Pretoria, Pretoria, 0028, South Africa. Thandi.Kapwata@mrc.ac.za.
³ Biostatistics Research Unit, South African Medical Research Council, Durban, 4001, South Africa.
⁴ Gangarosa Department of Environmental Health, Rollins School of Public Health, Emory University, Atlanta, GA, 30322, USA.
⁵ School of Agriculture, Earth, and Environmental Sciences, University of KwaZulu-Natal, Durban, 3629, South Africa.
⁶ Department of Earth Science, University of Turin, Turin, Italy.
⁷ Department of Geography, Geoinformatics and Meteorology, University of Pretoria, Pretoria, 0028, South Africa.
⁸ Environment and Health Research Unit, South African Medical Research Council, Pretoria, 0084, South Africa.

Abstract

Exposure to heatwaves may result in adverse human health impacts. Heat alerts in South Africa are currently based on defined temperature-fixed threshold values for large towns and cities. However, heat-health warning systems (HHWS) should incorporate metrics that have been shown to be effective predictors of negative heat-related health outcomes. This study contributes to the development of a HHWS for South Africa that can potentially minimize heat-related mortality. Distributed lag nonlinear models (DLNM) were used to assess the association between maximum and minimum temperature and diurnal temperature range (DTR) and population-adjusted mortality during summer months, and the effects were presented as incidence rate ratios (IRR). District-level thresholds for the best predictor from these three metrics were estimated with threshold regression. The mortality dataset contained records of daily registered deaths (n = 8,476,532) from 1997 to 2013 and data for the temperature indices were for the same period. Maximum temperature appeared to be the most statistically significant predictor of all-cause mortality with strong associations observed in 40 out of 52 districts. Maximum temperature was associated with increased risk of mortality in all but three of the districts. Our results also found that heat-related mortality was influenced by regional climate because the spatial distribution of the thresholds varied according to the climate zones across the country. On average, districts located in the hot, arid interior provinces of the Northern Cape and North West experienced some of the highest thresholds compared to districts located in temperate interior or coastal provinces. As the effects of climate change become more significant, population exposure to heat is increasing. Therefore, evidence-based HHWS are required to reduce heat-related mortality and morbidity. The exceedance of the maximum temperature thresholds provided in this study could be used to issue heat alerts as part of effective heat health action plans.

Keywords: Heat-health warning systems; Mortality; Temperature metric; Thresholds.

MeSH terms

Cities / epidemiology
Hot Temperature*
Humans
Mortality*
Seasons
South Africa / epidemiology
Temperature