Skip to main content
NCBI home page
Journal of Diabetes and Metabolic Disorders logoLink to Journal of Diabetes and Metabolic Disorders
. 2025 Sep 15;24(2):202. doi: 10.1007/s40200-025-01717-2

Economic burden, health burden and predictive analysis of diabetes attributed to non-optimal temperature in global from 1990 to 2030

Xudong Li 1,#, Yajie Wang 1,#, Tiancheng Zhang 1, Weiqing Rang 1,✉, Yan Liu 1,✉
PMCID: PMC12436262  PMID: 40964211

Abstract

Aims

This study investigated the impact of epidemiological and demographic Changes on the health and economic burdens of type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) attributed to non-optimal temperature.

Methods

Mortality data were from the 2021 Global Burden of Disease (GBD) database, while economic data were sourced from multiple databases. The study analyzed global and regional trends in diabetes deaths, mortality rates, and years of life lost (YLL) due to non-optimal temperatures from 1990 to 2021, considering factors such as the Socio-demographic Index (SDI), gender, region, and age. The economic impact was assessed using YLL and labor market indicators in the 50 most populous countries.

Results

In 2021, 102,872 T2DM deaths globally were attributed to non-optimal temperatures, with low-temperature ASMR 1.61 times higher than high-temperature ASMR. From 1990 to 2021, Age-Standardized Mortality Rate(ASMR) for T2DM attributed to high and low temperatures increased significantly (84.17% and 24.06%) in low-to-middle SDI regions. Older adults had the highest mortality rate, and males faced higher risks than females. ASMR peaked at an SDI of 0.48 and decreased with increasing SDI. From 2022 to 2030, the female ASMR from T2DM attributed to non-optimal temperatures is projected to rise by 26.54% in Pakistan and 42.61% in Nepal.

Conclusion

Low temperatures remain a major mortality driver, with elderly males and low-SDI populations most at risk. Targeting young males with preventive measures can reduce future mortality. Countries with large populations and low SDI should prioritize temperature interventions to address climate change.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40200-025-01717-2.

Keywords: Diabetes mellitus, Cold temperature, Hot temperature, Global burden of disease, Economic burden

Introduction

Diabetes, a chronic metabolic disease, not only reduces life expectancy in patients but also leads to a variety of disabling and potentially life-threatening complications, such as cardiovascular diseases, kidney diseases, and cancer [1–3]. According to GBD2021 study, diabetes has become the seventh leading cause of disability worldwide, with its complications, ischemic heart disease and stroke, ranking second and fourth, respectively [4]. As reported by the International Diabetes Federation (IDF), the global prevalence of diabetes reached 537 million adults in 2021, representing 10% of the adult population. This significant prevalence resulted in health expenditure costs amounting to $966 billion globally. Looking ahead, it is anticipated that by 2045, the number of individuals with diabetes will escalate to 783 million, driving medical costs to an estimated $10.5 trillion [5].

To mitigate the burden of diabetes, researchers have explored numerous potential risk factors in recent years [6]. In addition to genetics and lifestyle, environmental factors have also emerged as significant contributors to diabetes risk [7, 8]. Diabetic patients are particularly vulnerable to the adverse effects of extreme temperatures due to impaired thermoregulation and increased risk of dehydration [9]. Both high and low temperatures pose significant threats to the health of diabetic patients [10–12]. In high-temperature environments, insulin action tends to peak more readily, thereby increasing the risk of hypoglycemia in type 1 diabetic patients [13]. Cold weather, especially cold spells, can elevate the incidence and mortality rates of cardiovascular diseases among diabetic patients [14, 15]. Moreover, physical activity in high temperatures may reduce heart rate variability and increase the risk of cardiac autonomic neuropathy in diabetic patients [16, 17].

Several studies have also highlighted the impact of non-optimal ambient temperatures on the health of diabetic patients. A study conducted in England and Korea demonstrated a significant association between non-optimal temperatures and diabetes-related hospitalizations and mortality [10, 18]. A global study revealed that between 1990 and 2019, deaths and disability-adjusted life years (DALYs) attributable to T2DM due to non-optimal temperatures increased by 136.13% and 122.26%, respectively [19]. In the United States, a preliminary estimate suggested that exposure to extreme ambient temperatures could lead to annual economic losses of at least 2.7 billion dollars in the Minneapolis-St. Paul region [20].

Despite the existing evidence linking non-optimal temperatures to the health of diabetic patients, several gaps remain. For instance, research on the global diabetes burden stratified by diabetes subtypes is very limited, and comprehensive economic analyses of the global impact of diabetes are lacking. These gaps restrict our understanding of the health and economic impacts of extreme temperatures on diabetic patients. This study comprises three parts. The first part describes the global diabetes burden attributable to non-optimal temperatures and further explores the relationships between mortality burden and gender, age groups, and geographical locations. Using restricted cubic spline (RCS) and quantile regression analyses, the study examines the association between SDI and ASMR across 21 regions. The second part employs a Bayesian age-period-cohort (BAPC) model to predict diabetes mortality trends from 2022 to 2030 in five countries representing different SDI levels (high SDI: United States; middle-high SDI: China; middle SDI: Brazil; middle-low SDI: Pakistan; low SDI: Nepal). We focus on the top 50 most populous countries globally, collecting data on per capita health expenditure, labor force participation rates, and age-specific YLLs due to diabetes and non-optimal temperatures. By estimating market and non-market productivity losses and applying attribution percentage methods, we quantify the economic burden of diabetes attributable to non-optimal temperatures as a risk factor. Thus, this study not only describes the mortality burden of diabetes caused by non-optimal temperatures but also explores influencing factors (gender, age, region, SDI) and calculates the indirect economic burden, providing data support to mitigate the disease burden and offering references for global public health policies.

Study design and methods

Research objective and data sources

The key data for this study were derived from the GBD 2021, the United Nations Department of Economic and Social Affairs Population Division, the International Labour Organization and the Global Health Expenditure Database [21–24]. These sources cover indicators such as the number of deaths, mortality rates, YLLs, population by age group, labor force participation rates, and per capita health expenditure.

The study employs the GBD modeling framework to systematically assess the disease burden. Data are collected in accordance with the ICD-10 coding standards and integrated through standardization to ensure accuracy [6]. The association between non-optimal temperatures and diabetes is established using the Burden of Proof Risk Function (BPRF) method combined with meta-regression analysis. Subsequently, the Population Attributable Fraction (PAF) is calculated to evaluate the impact of risk factors on disease burden [25]. Finally, the mortality is estimated using a causal mortality ensemble model and the DisMod-MR 2.1 tool. Monte Carlo simulations are conducted to calculate the 95% confidence intervals by performing 1,000 iterations and extracting the 2.5th and 97.5th percentiles for each estimate [26]. This comprehensive approach elucidates the impact of non-optimal temperatures on the burden of diabetes. The detailed methodology for estimating the burden of diabetes attributable to non-optimal temperatures has already been published in the GBD 2021 study [26, 27].

“Per 100,000 population” is a standardized rate used as a population base to calculate health indicators. In GBD studies, this metric quantifies disease burden and enables rate comparisons, ensuring data comparability across countries and regions [28]. It adjusts for variations in population density and structure, allowing accurate comparisons across different sizes and demographic groups.

SDI is a comprehensive indicator for assessing the level of socio-economic development in a region. It is calculated by taking the geometric mean of three key indicators: the total fertility rate among those under the age of 25, the average years of education for those aged 15 and above, and the income per capita distribution adjusted for temporal changes [29]. GBD2021 showed that the SDI quintiles categorize 204 countries and regions into five distinct groups: Low SDI (0 to 0.47), Lower-middle SDI (0.47 to 0.62), Middle SDI (0.62 to 0.71), Hight-middle SDI (0.71 to 0.81), and High SDI (0.81 to 1) [30].

Based on the GBD 2021 data, this study employs a variety of analytical methods to comprehensively assess the impact of non-optimal temperatures on global diabetes mortality. Descriptive analyses are used to explore differences in diabetes mortality rates by sex, heatmaps are utilized to visualize age-specific mortality distributions, and maps are employed to illustrate geographical variations in the global diabetes mortality burden. The study also examines the relationship between SDI and ASMR using RCS and quantile regression analyses. Additionally, the BAPC model is applied to predict diabetes mortality trends from 2022 to 2030. Furthermore, the study quantifies the economic burden of diabetes attributable to non-optimal temperatures by integrating data on YLLs, per capita health expenditure, labor force participation rates, and other relevant indicators. This study aims to provide scientific evidence for global public health policy-making, thereby contributing to the mitigation of the mortality impact and associated economic burden of non-optimal temperatures on diabetes.

Statistical analysis methods

Restricted cubic splines

RCS is a statistical method used to flexibly model non-linear relationships between variables within regression models. It is widely applied in fields such as survival analysis and epidemiological research, where non-linear relationships need to be addressed [31]. When modeling exposure-disease relationships with RCS, it is generally recommended to select between 3 and 5 knots. In this study, 3 knots were used to validate the model.

Quantile regression

Quantile regression is a statistical technique used to estimate the relationship between predictor variables and the response variable at different levels of probability, characterized by its robustness and flexibility. It is widely applied in various fields, especially in studying disease risk factors and income distribution. This method is particularly suitable for analyzing data with heterogeneity or non-normal distribution [32].

graphic file with name d33e413.gif

Inline graphic is the predicted quantile of Inline graphic given Inline graphic, with Inline graphic as the target quantile and β as the regression coefficients [33].

Bayesian age period cohort model

BAPC model is an advanced statistical tool that combines the effects of age, period, and cohort to predict disease incidence and mortality rates. The model employs a second-order stochastic walk to refine the initial mortality rates for age, period, and cohort factors, and utilizes the Integrated Nested Laplace Approximation technique to estimate the marginal posterior distributions. This approach enhances the model’s efficiency and stability when processing extensive datasets [34].

graphic file with name d33e446.gif

The observed health outcome Inline graphic is influenced by age (Inline graphic), period (Inline graphic), and cohort (Inline graphic) effects, each representing deviations from the overall mean Inline graphic. Inline graphic accounts for unexplained random variation [35]. For each estimate, we conducted 1,000 simulation draws. These simulation draws were generated using the Monte Carlo simulation method based on the distribution characteristics of the original data. Monte Carlo simulation is a statistical technique that uses random sampling to estimate the uncertainty of complex systems. From these simulation draws, we calculated the 2.5th and 97.5th percentiles, and the range between these two percentiles represents the 95% uncertainty interval [36].

Indirect economic burden

Indirect economic costs include market and non-market productivity losses, and the formula is revised as follows [37, 38]:

graphic file with name d33e492.gif

Inline graphic represent the YLLs value by gender and age group, Inline graphic represent the labor participation rate by gender and age group, andInline graphic represent the per capita annual income. Among them, was 23% in high-income countries and 35% in low-income countries. Considering the varying productivity levels across age groups, each age group is weighted accordingly [39]. Due to the lack of YLD data, the study only considered indirect economic burdens, which were based entirely on YLLs due to diabetes deaths.

In this study, we primarily utilized the R software packages rms, quantreg, BAPC, and INLA to conduct our analyses [40–43]. The R software version used was 4.3.3, and a p value less than 0.05 was considered to indicate statistical significance.

Results

Temporal trends in ASMR

We collected the ASMRs of T1DM and T2DM attributable to non-optimal temperature at five different SDI quintiles from 1990 to 2021. For specific numbers of deaths, ASMR, YLLs, and age-standardized rates of YLLs in 1990 and 2021, please refer to e-Table 1.

Figure 1A presented that between 1990 and 2021, there was a notable decline in the global ASMR for T1DM attributed to non-optimal temperatures, with a decrease of 27.96% (from 0.06 to 0.04, per 100,000 people). This trend was not uniform across all SDI regions, as Hight-middle SDI regions experienced the most significant reduction in T1DM ASMR (49.40%, from 0.04 to 0.02, per 100,000 people), contrasting with the smallest decrease in Lower-middle SDI regions (5.74%, from 0.07 to 0.06, per 100,000 people). Figure 1D presented that For T2DM, the global ASMR due to non-optimal temperatures showed an opposite trend, increasing by 8.51% (from 1.16 to 1.26, per 100,000 people), with high SDI regions recording the most significant decrease (26.79%, from 0.99 to 0.72, per 100,000 people), and Lower-middle SDI regions showing the largest increase (47.49%, from 1.62 to 2.38, per 100,000 people).

Fig. 1.

Fig. 1

Open in a new tab

A, ASMR of T1DM attributable to non-optimal temperature at different SDI levels from 1990 to 2021. B, ASMR of T1DM attributable to high temperature at different SDI levels from 1990 to 2021. C, ASMR of T1DM attributable to low temperature at different SDI levels from 1990 to 2021. D, ASMR of T2DM ASMR attributable to non-optimal temperature at different SDI levels from 1990 to 2021. E, ASMR of T2DM ASMR attributable to high temperature at different SDI levels from 1990 to 2021. F, ASMR of T2DM ASMR attributable to low temperature at different SDI levels from 1990 to 2021. In addition, the confidence interval of death data of T1DM and T2DM caused by high temperature in high, middle and high SDI areas contains negative numbers, indicating that they may not be statistically significant or uncertain

Figure 1B presented that the impact of high temperatures on T1DM ASMR increased by 20.56% (from 0.01 to 0.02, per 100,000 people) globally, with high SDI regions experiencing the most pronounced increase (29.23%, from 0.00 to 0.01, per 100,000 people). Conversely, low SDI regions showed a slight decrease of 1.92% (from 0.03 to 0.03, per 100,000 people). Figure 1E presented that in the case of T2DM, the ASMR attributed to high temperatures rose significantly, with an overall increase of 76.21% (from 0.28 to 0.49, per 100,000 people). The impact was most substantial in Lower-middle SDI regions, where there was an 84.17% increase (from 0.06 to 0.11, per 100,000 people), while low SDI regions saw a 43.98% rise (from 0.77 to 1.11, per 100,000 people).

Figure 1C presented that the ASMR for T1DM due to low temperatures saw a significant decrease of 43.50% (from 0.04 to 0.02, per 100,000 people) globally, with Hight-middle SDI regions experiencing the steepest decline (53.35%, from 0.04 to 0.02, per 100,000 people) and Lower-middle SDI regions a more modest reduction (16.09%, from 0.03 to 0.02, per 100,000 people). Figure 1F presented that for T2DM, the ASMR attributed to low temperatures decreased by 11.60% (from 0.90 to 0.79, per 100,000 people), with high SDI regions recording the most significant reduction (33.18%, from 0.93 to 0.62, per 100,000 people). In contrast, Lower-middle SDI regions observed an increase of 24.06% (from 0.79 to 0.97, per 100,000 people), indicating a contrasting trend in the impact of low temperatures on T2DM ASMR across different SDI regions.

Gender-specific trends in ASMR

As illustrated in Fig. 2A-F, from 1990 to 2021, the globally associated ASMR for T1DM attributed to non-optimal temperatures declined for both males and females. The ASMR for males decreased by 22.62% (from 0.06 to 0.04, per 100,000 people), while females experienced a more significant reduction of 32.82% (from 0.06 to 0.04, per 100,000 people). In the SDI region analysis, males saw the greatest reduction in Hight-middle SDI regions, with a 40.50% decrease (from 0.04 to 0.03, per 100,000 people), and the smallest reduction in Lower-middle SDI regions, with an 8.95% decrease (from 0.07 to 0.06, per 100,000 people). Females also showed the greatest reduction in Hight-middle SDI regions, with a 56.99% decrease (from 0.05 to 0.02, per 100,000 people), and the smallest reduction in Lower-middle SDI regions, with a 2.75% decrease (from 0.07 to 0.07, per 100,000 people). In 2021, males generally had higher ASMR than females, except in middle SDI and Lower-middle SDI regions.

Fig. 2.

Fig. 2

Open in a new tab

A-F, ASMR of diabetes attributable to non-optimal temperature at different SDI levels from 1990 to 2021. (A-F) Type 1 Diabetes: (A) Global, (B) High SDI, (C) High-middle SDI, (D) Middle SDI, (E) Low-middle SDI, and (F) Low SDI; (G-L) Type 2 Diabetes: (G) Global, (H) High SDI, (I) High- middle SDI, (J) Middle SDI, (K) Low-middle SDI, and (L) Low SDI

As illustrated in Fig. 2G-L, contrary to the trend for T1DM, the ASMR associated with non-optimal temperatures for T2DM showed an increasing trend globally, with a 15.11% increase for males (from 1.14 to 1.31, per 100,000 people) and a 6.20% increase for females (from 1.08 to 1.15, per 100,000 people). In the SDI region analysis, males and females in high SDI and Hight-middle SDI regions exhibited a decreasing trend, while those in middle SDI, Lower-middle SDI, and low SDI regions showed an increasing trend. Males in Lower-middle SDI regions had the largest increase, with a 46.16% rise (from 1.62 to 2.37, per 100,000 people), and the greatest decrease in high SDI regions, with a 12.16% drop (from 0.98 to 0.86, per 100,000 people). Females in Lower-middle SDI regions had the largest increase, with a 54.24% rise (from 1.48 to 2.28, per 100,000 people), and the greatest decrease in high SDI regions, with a 39.14% drop (from 0.88 to 0.54, per 100,000 people).

Age-specific mortality analysis

As illustrated in Fig. 3. Between 1990 and 2021, deaths attributed to diabetes caused by non-optimal temperatures showed significant changes in age distribution, with a marked positive correlation between age and mortality rates. Notably, individuals with T1DM exhibited abnormal increases in the 40–44 and 75–79 age groups. Among different age groups, the oldest age group, 95 and above, displayed the highest age-specific mortality rates, especially in Lower-middle SDI regions, where the death rates for T1DM and T2DM reached up to 0.36 (per 100,000 people, 95% UI 20.81 to 99.30) and 56.78 (per 100,000 people, 95% UI 6.08 to 28.89), respectively.

Fig. 3.

Fig. 3

Open in a new tab

A: Age-specific Mortality Rates (per 100,000 People) for Type 1 Diabetes Attributable to Non-optimal Temperatures at different SDI Levels, 1990–2021. B: Age-specific Mortality Rates (per 100,000 People) for Type 2 Diabetes Attributable to Non-optimal Temperatures at different SDI Levels, 1990–2021

Geographical disparities in the burden

As shown in Fig. 4A, geographically, in 2021, the mortality burden of T1DM due to non-optimal temperatures was most severe in Central and South Asia, with Pakistan having the highest mortality rate, reaching 0.21 (per 100,000 people, 95% UI 0.06 to 0.40). Between 1990 and 2021, deaths attributed to T1DM from non-optimal temperatures showed a clear upward trend in Oceania, Central America, and Sub-Saharan Africa, with the Solomon Islands having the highest estimated annual percentage change (EAPC) of 3.66 (95% UI 3.04 to 4.28).

Fig. 4.

Fig. 4

Open in a new tab

A: 2021 ASMR and 1990–2021 global EAPC for type 1 Diabetes Mortality Attributed to Non-Optimal Temperatures. B:2021 ASMR and 1990–2021 global EAPC for type 2 Diabetes Mortality Attributed to Non-Optimal Temperatures

As depicted in Fig. 4B, geographically, in 2021, the mortality burden of T2DM due to non-optimal temperatures was most severe in North Africa, Western Asia, and South Asia, with Bahrain having the highest mortality rate, reaching 16.72 (per 100,000 people, 95% UI 6.08 to 28.89). Between 1990 and 2021, deaths attributed to T2DM from non-optimal temperatures showed a clear upward trend in Africa, Oceania, Central America, South America, and Northern Asia, with Samoa experiencing the highest EAPC of 4.93 (95% UI 0.5 to 9.55).

Relationship between SDI and ASMR

Figure 5 shown that The RCS model for T1DM did not show statistical significance. The RCS analysis for T2DM revealed that at an SDI value of approximately 0.5, the ASMR for T2DM was at its apex, primarily due to non-optimal temperature, signifying that this corresponds to the Lower-middle SDI quintile level. The results of the quantile regression indicate that for T1DM, the association at the 25th percentile is statistically significant, with ASMR increasing as SDI increases. However, at other percentiles, the impact of SDI on ASMR is not statistically significant. For T2DM, statistically significant results are observed at the 5th, 25th, 50th, and 75th percentiles, with ASMR decreasing as SDI increases. At other percentiles, the influence of SDI on ASMR is not statistically significant. The quantile results for YLLs and SDI show no clear trend of change, and most are not statistically significant.

Fig. 5.

Fig. 5

Open in a new tab

The relationship between the age-standardized mortality rate (per 100,000 people) and the age-standardized YLLs rate (per 100,000 people) for both type 1 and type 2 diabetes, attributable to non-optimal temperatures and the SDI from 1990 to 2021 across 21 regions worldwide, as classified by the global burden of disease. Each colored line in the figure traces the temporal trend within its respective region, with each data point corresponding to a specific year. The dotted lines indicate the outcomes of quantile regression analyses, arranged from top to bottom as follows: P95 = 95th percentile; P75 = 75th percentile; P50 = 50th percentile; P25 = 25th percentile; P5 = 5th percentile; SDI = Socio-demographic Index; YLLs = years of life lost

Projections of ASMR

It is projected that from 2022 to 2030, the mortality burden of T1DM due to non-optimal temperatures will increase for males in the United States, Brazil, and females in Pakistan, while it will decrease for males in China and Pakistan. Among them, Brazilian males are expected to have a significant increase in ASMR, projected to rise from 0.01, per 100,000 (95UL% 0.01–0.02) in 2022 to 0.02, per 100,000 (95UL% 0.00-0.03) in 2030, an increase of 36.11%. The ASMR for females in China is expected to decrease from 0.01, per 100,000 (95UL% 0.01–0.02) in 2022 to 0.01, per 100,000 (95UL% 0.00-0.01) in 2030, a decrease of 30.27%. Due to small sample sizes causing model estimation errors, the BAPC projections for females in Brazil and Nepal showed impractical negative values and were therefore omitted from e-Figure 1.

It is projected that from 2022 to 2030, the mortality burden of T2DM due to non-optimal temperatures will increase in Brazil, Pakistan, and Nepal. Among them, Nepalese females are expected to have a significant increase in ASMR, rising from 2.70, per 100,000 (95UL% 2.28–3.13) in 2022 to 3.86, per 100,000 (95UL% 1.63–6.08) in 2030, an increase of 42.61%. Meanwhile, the corresponding increase for females in Pakistan is projected to be 26.54%. Brazilian males will have the smallest increase, with ASMR rising from 0.87, per 100,000 (95UL% 0.76–0.97) in 2022 to 0.87, per 100,000 (95UL% 0.38–1.36) in 2030, an increase of 0.42%. The BAPC projection results for the global level, the United States, and China showed negative values, which are not meaningful in practical application and thus are not displayed in e-Figure 2.

Economic burden of diabetes

As shown in e-Table 2 for 2021, the United States, India, and China had the highest indirect economic burdens due to T1DM deaths from non-optimal temperatures, with $669.78 million, $295.41 million, and $180.37 million respectively. Madagascar had the lowest at $0.45 million. The same ranking applies to T2DM, with the United States at $2,818.55 million, India at $2,313.92 million, China at $1,711.47 million, and Madagascar at the bottom with $4.00 million.

Discussion

The ASMR of T2DM is higher in low to middle SDI areas. While cold temperatures are still the main death risk factor, their impact is reducing, and the risk from heat is growing. Males are more vulnerable to temperature extremes in diabetes, and older individuals face higher risks. Regions in South Asia, North Africa, and West Asia may face higher mortality risks from diabetes due to non-optimal temperatures. The RCS model shown that the highest ASMR for T2DM, attributable to non-optimal temperature, occurred when the SDI was 0.48. The results of the quantile regression showed that, in most cases, ASMR decreased as SDI increased for T2DM. The BAPC model forecasts increased T2DM deaths in Brazil, Pakistan, and Nepal due to non-optimal temperatures, especially among women in Pakistan and Nepal. T2DM causes greater economic burden than T1DM, particularly in lower SDI countries. The economic burden of T2DM is more substantial in populous countries such as the United States, India, and China.

A significant amount of data has indicated that diabetes-related fatalities are more likely to occur in low-temperature conditions [44, 45]. Additionally, a study encompassed diabetes among 17 mortality causes demonstrated that the burden attributed to low temperatures surpasses that from high temperatures, with a concurrent increase in the risk from high temperatures [26]. This study investigated the association between extreme temperatures, both high and low, and diabetes-related mortality, which is consistent with our conclusions.

A global analysis indicates that male patients with T2DM have a higher mortality burden than females [36]. In addition, some scholars have concluded that the mortality burden of men is higher than that of women in diabetes and kidney diseases caused by non-optimal temperature [29]. These conclusions align with our findings, as our study further explores the gender differences in mortality rates of T1DM and T2DM affected by non-optimal temperatures. The latest report shows that global warming has led to more frequent and serious extreme weather events [46]. A study on urban-rural differences reveals that women in rural areas and with lower levels of education are more susceptible to the effects of heatwaves [47]. This finding may account for the BAPC model’s forecast of heightened mortality among women in Nepal and Pakistan.

Older adults individuals with diabetes are more susceptible to the effects of extreme temperatures [48, 49]. A recent report indicated that among people aged 65 and above, the mortality rate for T1DM increases with age [50]. However, this study focused on all-cause mortality, as our research further investigated the impact of age on diabetes mortality caused by non-optimal temperatures. A study with a global perspective reported that the peak mortality for Type 1 diabetes occurred in the 40–44 age group [36]. These conclusions are consistent with our results. This emphasizes the need for special attention to temperature management for elderly diabetic patients.

A study published also found that extreme temperatures, especially high heat, increase the risk of death in several regions worldwide, with particularly significant impacts in sub-Saharan Africa and South Asia [26]. Our study further found that diabetic patients in South Asia, West Asia, and North Africa bore a greater burden due to non-optimal temperatures.

Data from the International Diabetes Federation indicates that China, India, and the United States have the highest number of people with diabetes globally [51]. Our research reveals that, despite their large populations, the attributable proportion of diabetes in these countries is not particularly high. This suggests that their large population base and high prevalence of diabetes, resulting in a significant number of diabetic patients, which leads to substantial medical and productivity losses [5]. Furthermore, the high medical costs in the United States, China’s aging population and uneven distribution of medical resources, and India’s uneven distribution of medical resources further exacerbate the economic burden of diabetes [52]. In contrast, many low SDI areas exhibit a larger attributable proportion but bear a lower overall economic burden. Research indicates that the diabetes burden in low- to middle-income countries is more significantly influenced by public awareness, policy, economic development, and levels of medical innovation [53]. A global study of low-income and middle-income countries shows that these countries are more susceptible to environmental factors, including extreme temperatures, due to limited medical resources, insufficient distribution of health resources, and poor health awareness among patients [54]. This susceptibility may be attributed to the lack of advanced medical technology, insufficient public health measures, limited income affecting healthcare spending, and low treatment rates, which result in poor health outcomes but a relatively lower overall economic burden.

Research indicates that as global warming progresses, the frequency and intensity of extreme heat events are increasing, posing a significant threat to the health of individuals with diabetes. For instance, studies have shown that each 1-degree Celsius rise in temperature can lead to over 100,000 new cases of diabetes annually [55]. Moreover, both heatwaves and cold snaps can increase the metabolic burden on individuals with diabetes, raising the risk of complications and leading to higher mortality rates [6]. Air pollution, particularly fine particulate matter (PM2.5) and ozone, has been linked to an increased incidence and mortality rate of diabetes. Research indicates that under high-temperature conditions, the production of air pollutants increases, which can further exacerbate the health risks of diabetes by triggering inflammatory responses and insulin resistance [9]. Studies predict that in a high-emission scenario, the attributable fraction of diabetes deaths due to heatwaves in the 2090 s will reach 9.3%, significantly higher than the 4.6% in a low-carbon scenario [56]. Therefore, climate adaptation strategies should prioritize vulnerable groups, including diabetes patients, and enhance the integration of low-carbon transitions with public health systems to reduce the risk of chronic disease deaths during extreme weather events.

Our study reveals that the burden of high temperatures is increasing while that of low temperatures is decreasing, necessitating a shift in focus towards heat-related health risks. Elderlymenhave been identified as a high-risk group, and it is recommended that preemptive interventions target middle-aged men [49]. Regions such as South Asia, West Asia, and North Africa face disproportionate mortality burdens and require enhanced temperature control infrastructure [54]. Additionally, the rising trend of T2DM burden among women in Pakistan and Nepal calls for early diabetes screening and prevention in similar regions. In countries with high diabetes prevalence, such as the United States, India, and China, subsidizing medical costs and providing free health check-ups can help alleviate the significant economic burden [20]. These targeted strategies can effectively reduce diabetes mortality and economic pressure while enhancing public health resilience.

This study boasts several salient strengths. Firstly, this study conducted an in-depth analysis of T1DM and T2DM, separately examining their impacts on SDI regions from 1990 to 2021, and comparing the health and economic burdens associated with each subtype. Secondly, our study offers a comprehensive view using robust statistical methods, including RCS, quantile regression, and BAPC models, ensuring the precision and reliability of our findings. Additionally, this research establishes a scientific basis for the formulation of global health policies by assessing indirect economic burdens and quantifying the contributions of key risk factors, enhancing our comprehensive understanding of the disease burden and economic impacts of diabetes due to non-optimal temperatures, and thus providing strong support for the prioritization and implementation of public health intervention measures. This study, however, exhibits certain limitations. Firstly, due to the limitations of the original data, we could not fully account for the impact of factors such as seasonality, genetics, and culture on diabetes. Secondly, the incidence of T1DM is relatively low, which poses a challenge to achieving a sufficiently large sample size to ensure the universality of the research results. Furthermore, due to the lack of original data, our study focused mainly on the mortality burden, omitting the disability burden, which likely underestimates the comprehensive economic impact of non-optimal temperatures on diabetes. To enhance the efficacy and precision of this research, it is recommended that subsequent studies utilize more advanced statistical methods and strengthen data integration, overcoming these limitations through interdisciplinary collaboration.

Conclusion

The mortality burden from diabetes due to non-optimal temperatures is notably higher in T2DM, with low temperatures being the main cause of death. Factors contributing to this burden include gender, age, geographical location, and SDI. Targeting young males with preventive measures can reduce future mortality. Countries with large populations and low SDI should prioritize temperature interventions to address climate change.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (DOCX 787 KB) (787.2KB, docx)
Supplementary Material 2 (DOCX 3.39 MB) (3.4MB, docx)
40200_2025_1717_MOESM3_ESM.docx (18.2KB, docx)

Supplementary Material 3 (DOCX 18.1 KB )

Supplementary Material 4 (DOCX 23.5 KB) (23.5KB, docx)
Supplementary Material 5 (DOCX 2.53 MB) (2.5MB, docx)

Acknowledgements

We would like to express our sincere gratitude to the following organizations for providing the datasets used in this study: Institute for Health Metrics and Evaluation (IHME) for the Global Burden of Disease Study 2021 (GBD 2021) Data Resources.International Labour Organization (ILO) for the ILOSTAT labour statistics.United Nations Department of Economic and Social Affairs (UN DESA) for the World Population Prospects.World Health Organization (WHO) for the Global Health Expenditure Database (GHED).International Diabetes Federation (IDF) for the IDF Diabetes Atlas.

Abbreviations

ASMR

Age-Standardized Mortality Rate

BAPC

Bayesian Age-Period-Cohort

BARF

Burden of Proof Risk Function

DALYs

Disability-Adjusted Life Years

GHED

Global Health Expenditure Database

GBD

Global Burden of Disease

IDF

International Diabetes Federation

ILOSTAT

International Labour Organization Statistics

PM2.5

particularly fine particulate matter

PAF

Population Attributable Fraction

RCS

Restricted Cubic Splines

SDI

Socio-Demographic Index

T1DM

Type 1 Diabetes Mellitus

T2DM

Type 2 Diabetes Mellitus

UNPD

United Nations Population Division

YLL

Years of Life Lost

Author contributions

All authors agreed on the final version of the manuscript. X. L. acquired the data, performed the analysis of data, and wrote the manuscript. Y. W. contributed to the coding of the statistical analysis. T. Z. contributed to the language revision of the manuscript. W. R. and Y. L. designed and evaluated the whole work. Other contributions: We thank the GBD, the United Nations Population Division, ILO and the World Bank Group, and GHED for providing data resources.

Funding

This research was supported by Key Projects in Education Department of Hunan Province (Grant No. 2324JY102) for the study titled"Optimization of the Attribution System for Liver Cancer Burden Caused by Cyanotoxins and Construction of a Visualization Data Platform”.

Data availability

The datasets analyzed in the current study are available in the [GBD2021, https://ghdx.healthdata.org/gbd-2021], [ILOSTAT, https://ilostat.ilo.org/], [UNPD, https://population.un.org/wpp/downloads? folder=Documentation&group=Documentation], [GHED, https://apps.who.int/nha/database/Home/Index/en/], [IDF Diabetes Atlas, https://diabetesatlas.org/atlas/tenth-edition/]. We will strictly follow the principles of legality and transparency of data use and ensure the correct use and interpretation of research data. We will avoid data misunderstanding and abuse as much as possible and use the data only for research purposes.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Xudong Li and Yajie Wang contributed equally to this work.

Contributor Information

Weiqing Rang, Email: rwqktz2018@126.com.

Yan Liu, Email: 2015002062@usc.edu.cn.

References

  • 1.Borén J, Öörni K, Catapano AL. The link between diabetes and cardiovascular disease. Atherosclerosis. 2024;394:117607. DOI ARTN 11760710.1016/j.atherosclerosis.2024.117607. [DOI] [PubMed] [Google Scholar]
  • 2.Stedman M, Davies M, Livingston M, Alshames R, Lunt M, Rayman G, Gadsby R, Heald AH. Estimating life years lost to diabetes: outcomes from analysis of National Diabetes Audit and Office of National Statistics data England. Diabetologia. 2020;63:S134–5. 10.1097/XCE.0000000000000210. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Chan JCN, Lim L-L, Wareham NJ, Shaw JE, Orchard TJ, Zhang P, et al. The lancet commission on diabetes: using data to transform diabetes care and patient lives. Lancet. 2020;396:2019–82. 10.1016/s0140-6736(20)32374-6. [DOI] [PubMed] [Google Scholar]
  • 4.Diseases GBD, Injuries C. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403:2133–61. 10.1016/S0140-6736(24)00757-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. IDF diabetes atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183:109119. 10.1016/j.diabres.2021.109119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Collaborators GBDRF. Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403:2162–203. 10.1016/S0140-6736(24)00933-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.ElSayed NA, Aleppo G, Bannuru RR, Bruemmer D, Collins BS, Ekhlaspour L, Gaglia JL, Hilliard ME, Johnson EL, Khunti K, et al. 2. Diagnosis and classification of diabetes:standards of care in Diabetes—2024. Diabetes Care. 2024;47:S20–42. 10.2337/dc24-S002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications. Nat Rev Endocrinol. 2018;14:88–98. 10.1038/nrendo.2017.151. [DOI] [PubMed] [Google Scholar]
  • 9.Ratter-Rieck JM, Roden M, Herder C. Diabetes and climate change: current evidence and implications for people with diabetes, clinicians and policy stakeholders. Diabetologia. 2023;66:1003–15. 10.1007/s00125-023-05901-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kim KN, Lim YH, Bae S, Kim JH, Hwang SS, Kim MJ, Oh J, Lim H, Choi J, Kwon HJ. Associations between cold spells and hospital admission and mortality due to diabetes: a nationwide multi-region time-series study in Korea. Sci. Total Environ. 2022;838:156464. 10.1016/j.scitotenv.2022. [DOI] [PubMed] [Google Scholar]
  • 11.Wang SY, Lei YT, Wang XL, Ma K, Wang C, Sun CH, Han TS. Association between temperatures and type 2 diabetes: a prospective study in UK Biobank. Diabetes Res Clin Pract. 2024;215:111817. 10.1016/j.diabres.2024.111817. [DOI] [PubMed] [Google Scholar]
  • 12.Xu ZW, Tong SL, Cheng J, Crooks JL, Xiang H, Li XY, Huang CR, Hu WB. Heatwaves and diabetes in brisbane, australia: a population-based retrospective cohort study. Int J Epidemiol. 2019;48:1091–100. 10.1093/ije/dyz048. [DOI] [PubMed] [Google Scholar]
  • 13.Al-Qaissi A, Papageorgiou M, Javed Z, Heise T, Rigby AS, Garrett AT, et al. Environmental effects of ambient temperature and relative humidity on insulin pharmacodynamics in adults with type 1 diabetes mellitus. Diabetes Obes Metab. 2019;21:569–74. 10.1111/dom.13555. [DOI] [PubMed] [Google Scholar]
  • 14.Fan JF, Xiao YC, Feng YF, Niu LY, Tan X, Sun JC, Leng YQ, Li WY, Wang WZ, Wang YK. A systematic review and meta-analysis of cold exposure and cardiovascular disease outcomes. Front. cardiovasc. med. 2023;10:1084611. 10.3389/fcvm.2023.1084611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Dal Canto E, Ceriello A, Rydén L, Ferrini M, Hansen TB, Schnell O, Standl E, Beulens JWJ, Aboyans V, Cosentino F. Diabetes as a cardiovascular risk factor: an overview of global trends of macro and micro vascular complications. Eur J Prev Cardiol. 2019;26:25–32. 10.1177/2047487319878371. [DOI] [PubMed] [Google Scholar]
  • 16.Macartney MJ, Notley SR, Herry CL, Seely AJE, Sigal RJ, Kenny GP. Cardiac autonomic modulation in type 1 diabetes during exercise-heat stress. Acta Diabetol. 2020;57:959–63. 10.1007/s00592-020-01505-9. [DOI] [PubMed] [Google Scholar]
  • 17.De Barros JA, Macartney MJ, Peoples GE, Notley SR, Herry CL, Kenny GP. The impact of age, type 2 diabetes and hypertension on heart rate variability during rest and exercise at increasing levels of heat stress. Eur J Appl Physiol. 2022;122:1249–59. 10.1007/s00421-022-04916-4. [DOI] [PubMed] [Google Scholar]
  • 18.Hajat S, Haines A, Sarran C, Sharma A, Bates C, Fleming LE. The effect of ambient temperature on type-2-diabetes: case-crossover analysis of 4 + million GP consultations across England. Environ Health. 2017;16:73. 10.1186/s12940-017-0284-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Liu Y, Wang D, Huang X, Liang R, Tu Z, You X, et al. Temporal trend and global burden of type 2 diabetes attributable to non-optimal temperature, 1990–2019: an analysis for the global burden of disease study 2019. Environ Sci Pollut Res Int. 2023;30:82686–95. 10.1007/s11356-023-28225-1. [DOI] [PubMed] [Google Scholar]
  • 20.Liu Y, Saha S, Hoppe BO, Convertino M. Degrees and dollars - Health costs associated with suboptimal ambient temperature exposure. Sci Total Environ. 2019;678:702–11. 10.1016/j.scitotenv.2019.04.398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Institute for Health Metrics and Evaluation (IHME). Global Burden of Disease Study 2021. https://vizhub.healthdata.org/gbd-results/. Accessed 10 Jun 2025.
  • 22.World Health Organization (WHO). Global Health Expenditure Database. https://apps.who.int/nha/database. Accessed 10 Jun 2025.
  • 23.International Labour Organization (ILO). Labour Force Surveys and Estimates. https://www.ilo.org/global/statistics-and-databases/lang--en/index.htm. 10 Jun 2025.
  • 24.United Nations Department of Economic and Social Affairs, Population Division. World Population Prospects: The 2023 Revision. https://population.un.org/wpp/. 10 Jun 2025.
  • 25.Murray CJL, Aravkin AY, Zheng P, Abbafati C, Abbas KM, Abbasi-Kangevari M, et al. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396:1223–49. 10.1016/s0140-6736(20)30752-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Burkart KG, Brauer M, Aravkin AY, Godwin WW, Hay SI, He J, et al. Estimating the cause-specific relative risks of non-optimal temperature on daily mortality: a two-part modelling approach applied to the global burden of disease study. Lancet. 2021;398:685–97. 10.1016/s0140-6736(21)01700-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Diseases GBD, Injuries C. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396:1204–22. 10.1016/S0140-6736(20)30925-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Collaborators GBDCD. Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403:2100–32. 10.1016/S0140-6736(24)00367-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Liu Y, Wen HY, Bai JJ, Shi F, Bi R, Yu CH. Burden of diabetes and kidney disease attributable to non-optimal temperature from 1990 to 2019: A systematic analysis from the Global Burden of Disease Study 2019. Sci. Total Environ. 2022;838:156495. 10.1016/j.scitotenv.2022. [DOI] [PubMed] [Google Scholar]
  • 30.Ferrari AJ, Santomauro DF, Aali A, Abate YH, Abbafati C, Abbastabar H, Abd ElHafeez S, Abdelmasseh M, Abd-Elsalam S, Abdollahi A, et al. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the global burden of disease study 2021. Lancet. 2024;403:2133–61. 10.1016/s0140-6736(24)00757-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Marrie RA, Dawson NV, Garland A. Quantile regression and restricted cubic splines are useful for exploring relationships between continuous variables. J Clin Epidemiol. 2009;62:511–7. 10.1016/j.jclinepi.2008.05.015. [DOI] [PubMed] [Google Scholar]
  • 32.Konstantopoulos S, Li W, Miller S, van der Ploeg A. Using quantile regression to estimate intervention effects beyond the mean. Educ Psychol Meas. 2019;79:883–910. 10.1177/0013164419837321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Huang Q, Zhang H, Chen J, He M. Quantile Regression Models and Their Applications: A Review. Journal of biometrics & biostatistics 2017.
  • 34.Riebler A, Held L. Projecting the future burden of cancer: bayesian age-period-cohort analysis with integrated nested Laplace approximations. Biom J. 2017;59:531–49. 10.1002/bimj.201500263. [DOI] [PubMed] [Google Scholar]
  • 35.Fosse E. Bayesian age–period–cohort models. In: Bell A, editor. Age, period and cohort effects: statistical analysis and the identification problem. London: Routledge; 2020. pp. 84–116. [Google Scholar]
  • 36.Dong CP, Wu GF, Li H, Qiao Y, Gao S. Type 1 and type 2 diabetes mortality burden: predictions for 2030 based on bayesian age-period-cohort analysis of China and global mortality burden from 1990 to 2019. J Diabetes Investig. 2024;15:623–33. 10.1111/jdi.14146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Behera DK, Viswanathan PK, Mishra S. Effects of air pollution on global health: evidence from the global burden of disease study in the BRICS countries. Int Arch Occup Environ Health. 2024;97:813–32. 10.1007/s00420-024-02087-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Li C, Qi J, Yin P, Yu X, Sun H, Zhou M, et al. The burden of type 2 diabetes attributable to air pollution across China and its provinces, 1990–2021: an analysis for the global burden of disease study 2021. Lancet Reg Health. 2024;53:101246. 10.1016/j.lanwpc.2024.101246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Azizi S, Hadi Dehghani M, Nabizadeh R. Ambient air fine particulate matter (PM10 and PM2.5) and risk of type 2 diabetes mellitus and mechanisms of effects: a global systematic review and meta-analysis. Int J Environ Health Res. 2024. 10.1080/09603123.2024.2391993. [DOI] [PubMed] [Google Scholar]
  • 40.Harrell, FE. rms: Regression Modeling Strategies. https://CRAN.R-project.org/package=rms. (2023). Accessed 10 Jun 2025.
  • 41.Koenker, R. quantreg: Quantile Regression. https://CRAN.R-project.org/package=quantreg. (2023). Accessed 10 Jun 2025.
  • 42.Held L, Rue H. BAPC: Projection of Cancer Incidence and Mortality Data Using Integrated Nested Laplace Approximations (INLA). https://CRAN.R-project.org/package=BAPC. (2022) Accessed June 10, 2025.
  • 43.Rue H, Martino S, Chopin N. Approximate bayesian inference for latent Gaussian models using integrated nested Laplace approximations (with discussion). J Royal Stat Soc Ser B. 2009;71:319–92. 10.1111/j.1467-9868.2008.00700. [Google Scholar]
  • 44.Davis RE, Driskill EK, Novicoff WM. The association between weather and emergency department visitation for diabetes in Roanoke, Virginia. Int J Biometeorol. 2022;66:1589–97. 10.1007/s00484-022-02303-4. [DOI] [PubMed] [Google Scholar]
  • 45.Lai SW, Chang WC, Lin CL, Chou IC, Tsai FJ, Lai YJ. Low ambient temperatures correlate with increased risk of hypoglycemia in patients with type 2 diabetes an ecological study in Taiwan. Medicine (Baltimore). 2020;99:e19287 DOI ARTN e1928710.1097/MD.0000000000019287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Costello A, Romanello M, Hartinger S, Gordon-Strachan G, Huq S, Gong P, et al. Climate change threatens our health and survival within decades. Lancet. 2023;401:85–7. 10.1016/S0140-6736(22)02353-4. [DOI] [PubMed] [Google Scholar]
  • 47.Tao J, Zheng H, Ho HC, Wang X, Hossain MZ, Bai Z, et al. Urban-rural disparity in heatwave effects on diabetes mortality in Eastern China: a case-crossover analysis in 2016–2019. Sci Total Environ. 2023;858:160026. 10.1016/j.scitotenv.2022.160026. [DOI] [PubMed] [Google Scholar]
  • 48.Huang Y, Yang J, Chen JW, Shi HJ, Lu XJ. Association between ambient temperature and age-specific mortality from the elderly: Epidemiological evidence from the Chinese prefecture with most serious aging. Environ. Res. 2022;211:113103. 10.1016/j.envres.2022.113103. [DOI] [PubMed] [Google Scholar]
  • 49.Bellary S, Kyrou I, Brown JE, Bailey CJ. Type 2 diabetes mellitus in older adults: clinical considerations and management. Nat Rev Endocrinol. 2021;17:534–48. 10.1038/s41574-021-00512-2. [DOI] [PubMed] [Google Scholar]
  • 50.Yang KJ, Yang X, Jin CY, Ding SN, Liu TT, Ma B, Sun H, Zhang J, Li YZ. Global burden of type 1 diabetes in adults aged 65 years and older, 1990–2019: population based study. Bmj-British Med J. 2024;385:e078432. ARTN e07843210.1136/bmj-2023-078432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Magliano DJ, Boyko EJ, IDF Diabetes AtlasIDF DIABETES ATLAS International Diabetes Federation.© International Diabetes Federation, 2021. Brussels. https://diabetesatlas.org/atlas/tenth-edition/. (2021). 28 Sept 2024.
  • 52.Chen Y, Wang G, Hou Z, Liu X, Ma S, Jiang M. Comparative diabetes mellitus burden trends across global, Chinese, US, and Indian populations using GBD 2021 database. Sci Rep. 2025;15:11955. 10.1038/s41598-025-96175-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Agarwal S, Wade AN, Mbanya JC, Yajnik C, Thomas N, Egede LE, Campbell JA, Walker RJ, Maple-Brown L, Graham S. Global inequity in diabetes 1 the role of structural racism and geographical inequity in diabetes outcomes. Lancet. 2023;402:235–49. 10.1016/S0140-6736(23)00909-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Liu JL, Bai RH, Chai ZL, Cooper ME, Zimmet PZ, Zhang L. Low- and middle-income countries demonstrate rapid growth of type 2 diabetes: an analysis based on global burden of disease 1990–2019 data. Diabetologia. 2022;65:1339–52. 10.1007/s00125-022-05713-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Meo SA, Meo AS. Climate change and diabetes mellitus - emerging global public health crisis: observational analysis. Pak J Med Sci. 2024;40:559–62. 10.12669/pjms.40.4.8844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Chen S, Zhou M, Liu DL, Tong S, Xu Z, Li M, et al. Mortality burden of diabetes attributable to high temperature and heatwave under climate change scenarios in China. NPJ Clim Atmos Sci. 2024;7:289. 10.1038/s41612-024-00839-3. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (DOCX 787 KB) (787.2KB, docx)
Supplementary Material 2 (DOCX 3.39 MB) (3.4MB, docx)
40200_2025_1717_MOESM3_ESM.docx (18.2KB, docx)

Supplementary Material 3 (DOCX 18.1 KB )

Supplementary Material 4 (DOCX 23.5 KB) (23.5KB, docx)
Supplementary Material 5 (DOCX 2.53 MB) (2.5MB, docx)

Data Availability Statement

The datasets analyzed in the current study are available in the [GBD2021, https://ghdx.healthdata.org/gbd-2021], [ILOSTAT, https://ilostat.ilo.org/], [UNPD, https://population.un.org/wpp/downloads? folder=Documentation&group=Documentation], [GHED, https://apps.who.int/nha/database/Home/Index/en/], [IDF Diabetes Atlas, https://diabetesatlas.org/atlas/tenth-edition/]. We will strictly follow the principles of legality and transparency of data use and ensure the correct use and interpretation of research data. We will avoid data misunderstanding and abuse as much as possible and use the data only for research purposes.

Articles from Journal of Diabetes and Metabolic Disorders are provided here courtesy of Springer

RESOURCES