Climate Driven Interactions Between Malaria Transmission and Diabetes Prevalence

TL;DR

The paper addresses the dual burden of climate-driven malaria transmission and rising diabetes by developing a climate-informed, compartmental framework that stratifies humans by diabetes status and models vector dynamics. Using synthetic Indian data (2019–2021) and next-generation matrix methods, it quantifies how diabetes amplifies malaria risk and alters seasonality, producing an approximate baseline R0 of 1.53 and strong seasonal oscillations. Key findings show diabetics have 1.8–4.0 times higher odds of infection, with peak prevalence around 35–36% versus 20–21% in non-diabetics, and longer infectious periods that sustain transmission. The results underscore the need for integrated climate-adaptive health strategies that jointly address malaria and diabetes, especially in resource-limited settings where diabetic prevalence is rising.

Abstract

Climate change is intensifying infectious and chronic diseases like malaria and diabetes, respectively, especially among the vulnerable populations. Global temperatures have risen by approximately $0.6^\circ$C since 1950, extending the window of transmission for mosquito-borne infections and worsening outcomes in diabetes due to metabolic stress caused by heat. People living with diabetes have already weakened immune defenses and, therefore, are at an alarmingly increased risk of contraction of malaria. However, most models rarely include both ways of interaction in changing climate conditions. In the paper, we introduce a new compartmental epidemiological model based on synthetic data fitted to disease patterns of India from 2019 to 2021. The framework captures temperature-dependent transmission parameters, seasonal variability, and different disease dynamics between diabetic and non-diabetic groups within the three-compartment system. Model calibration using Multi-Start optimization combined with Sequential Quadratic Programming allows us to find outstanding differences between populations. The odds of malaria infection in diabetic individuals were found to be 1.8--4.0 times higher, with peak infection levels in 35--36\%, as compared to 20--21\% in the non-diabetic ones. The fitted model was able to capture well the epidemiological patterns observed, while the basic reproduction number averaged around 2.3, ranging from 0.31 to 2.75 in different seasons. Given that India's diabetic population is set to rise to about 157 million people by 2050, these findings point to a pressing need for concerted efforts toward climate-informed health strategies and monitoring systems that address both malaria and diabetes jointly.

Figures (6)

• Figure 1: Graph gives a frightening projection of diabetes in India. The population with the condition is likely to rise to a staggering 156.7 million as compared to 32.7 million in 2000, this is almost half the world, showing that a gigantic public health challenge is on the offing IDF2023.
• Figure 2: The diagram shows how malaria spreads differently among people with and without diabetes. A general susceptible population $S(t)$ splits into diabetic ( $\alpha$) and non-diabetic ( $1-\alpha$) subpopulations with different transmissions. Diabetic individuals have transmission rate $\lambda_h \cdot b_D$ and recovery rate $\gamma_{MD}$, while non-diabetics follow similar flows with rate $\lambda_h \cdot b_{-D}$ and recovery $\gamma_{-D}$. Mosquito populations transition from susceptible $S_v(t)$ to infected $I_v(t)$ at rate $\lambda_v$, with mortality $\mu_v$ in both compartments. Cross-transmission occurs bidirectionally: infected vectors transmit to humans via force of infection $\lambda_h$ represented by dashed arrows, While infected humans transmit back to vectors through rates $\lambda_{v,D}$ for diabetic individuals and $\lambda_{v,-D}$ for non-diabetic individuals. The model demonstrates how diabetes creates differential transmission and recovery dynamics within a vector-borne disease system while maintaining homogeneous mixing between populations.
• Figure 3: The figure shows the difference in the spread of malaria between diabetic and non-diabetic populations in the 36-month period with a Susceptible-Infected-Susceptible (SIS) model. The trends of the two groups are clearly different. Strong seasonal swings are seen among the diabetic population of 80,000 people. The weakened group is reduced to an average of between 50,000 to 52,000 during the peak transmission as the population of infected individuals increases to between 28,000 to 29,000, implying that at these times there is a 35--36% infection rate among diabetics. Compared to this, the non-diabetic population of 920,000 people exhibits synchronized seasonal variations with less dramatic changes in comparison. The vulnerable population is between 750,000 and 850,000 people, with the infected cases ranging from 180,000 to 190,000, which represents lower infection rates of 20--21%. The model has three full cycles of epidemics per year. The relative infection peaks, oscillations, and epidemic volatility are higher in the diabetic population compared to the non-diabetics. These variations are a factor of higher transmission probability ($0.65$ in diabetics versus $0.50$ in non-diabetics) and slower recovery, with an average of 120 days of infection compared with 60 days of infection in non-diabetics. Even with these distinctions, both populations exhibit synchronized seasonal patterns, which demonstrate that both populations are influenced by common environmental factors. .
• Figure 4: The figure shows how the model adjusted using different recovery rates for each group. The left panel tracks infected diabetics ($I_{MD}$) who recover more slowly ($\gamma_{MD} = \tfrac{1}{120}$ per day), while the right panel shows infected non-diabetics ($I_{M}$) recovering faster ($\gamma_{-D} = \tfrac{1}{60}$ per day). The red and blue dots are our synthetic data with added Gaussian noise, based on real clinical statistics. Solid lines show how well the model fits this data, with shaded regions marking 95% confidence intervals. The tight fit confirms the model successfully captures seasonal patterns, elevated baseline infections, and sustained epidemic dynamics that align with clinical observations. Data points plotted are month wise.
• Figure 5: The heatmap displays the relationships between different disease compartments across diabetic and non-diabetic individuals and mosquito populations over a 3-year study period. The model parameters set included the mean biting rate of $0.099$, seasonal amplitude of $0.810$, and recovery rates between groups: for diabetics at $0.0082$ per day (122.2-day infection duration) and for non-diabetics at $0.0163$ per day (61.2-day duration). Several key patterns arise within the correlation structure. Within each population group, susceptible and infected compartments display perfect negative correlations ($-1.00$), which intuitively makes sense given the closed population assumption of SIS models where individuals simply move back and forth between these two states. Strong positive correlations ($0.90$–$0.97$) across all infected populations imply that epidemics among diabetic and non-diabetic groups rise and fall in tandem. Both populations respond to the same seasonal transmission drivers, explaining the synchronized pattern, even though non-diabetics clear infections roughly twice as fast as diabetics. Simulated predictions from the model show high agreement with observed data ($0.94$–$0.97$), confirming that real underlying relationships are well captured even in the presence of measurement noise. The color gradient ranges from dark blue ($-1.00$), through white (no correlation), to dark red ($+1.00$).