The interplay between climate change and ageing: A systematic review of health indicators

2. Studies that do not investigate the impact of climate change on health outcomes. 3. Studies that focus on animal or laboratory-based research. 4. Studies that do not report original research (e.g., review articles, editorials, comments, letters). 5. Studies that focused on future climate change scenarios projections of climate change or health outcomes. Interrater reliability Two reviewers were responsible for the data selection process, with a third reviewer (E.C.) resolving any discrepancies. The SPSS software v.28.0 (IBM Corp. Armonk, NY, USA) was employed to determine the level of agreement (interrater reliability) among the reviewers. The reliability of data collection is crucial, particularly in healthcare and clinical research studies, to ensure consistency when selecting articles for a systematic review or samples for a study. Cohen’s kappa statistics were utilized to measure the interrater reliability of the selected arti- cles for this study [37]. The article selection process for this systematic review was conducted in three phases: title selection, abstract selection, and full article selection, with Cohen’s kappa value calculated independently for all phases. Results The systematic search retrieved 3154 papers (Fig 1). In the end, after removing the duplicates and applying the above-mentioned inclusion and exclusion criteria, 19 studies were consid- ered suitable for this systematic review and were analyzed to identify health indicators of the impact of climate change on older adults [10, 38–55]. After reviewing 2682 titles, 424 abstracts

were identified for further analysis. Among these, 27 articles underwent full-text analysis, resulting in 19 articles being chosen for further examination. Cohen’s kappa values were inde- pendently calculated for each phase. A moderate kappa value of 0.553 was observed for inter- rater reliability during the title selection phase, while an almost perfect agreement was achieved during the abstract and full-text selection phases, with kappa values of 0.910 and 0.914, respectively. Most articles were ecological studies (n = 8) [40, 42, 43, 45, 48, 53–55] and the rest were ret- rospective cohort studies (n = 5) [38, 39, 47, 51, 52], case-crossover studies (n = 3) [41, 46, 50], prospective cohort studies (n = 2) [10, 49] and a case-control study (n = 1) [44]. In the included articles, the locations studied were dispersed across North America [10, 38, 41, 43, 44, 46, 48, 51, 54], Asia [39, 42, 47, 49, 52, 53], Europe [40, 45, 50, 54], and Oceania [55], and none in Africa or South America. The majority of the included articles covered devel- oped countries. The duration of the data collection ranged from 2 to 44 years. The studies investigated the effect of different climate change phenomena on the health of older adults, the most recurring being various forms of ambient temperature and its variability. Extreme tem- peratures were also a common exposure. Other climate change phenomena investigated included drought periods, air pollution from wildfires, strong winds, sea surface temperature variability and extreme El Niño-southern Oscillation (ENSO) conditions. The human health

outcomes were related to mortality, cardiovascular, respiratory, renal, and mental health risks. One study also investigated hip fracture incidence. Table 1 summarizes the characteristics and main findings of the studies in this systematic review. As the aim of this systematic review was to identify health indicators, the outcomes of each study were analysed and classified as morbidity or mortality (Table 1). Overall, 10 articles [10, PLOS ONE Climate change, ageing and health PLOS ONE | https://doi.org/10.1371/journal.pone.0297116 April 24, 2024 5 / 17