New research shows that infection-related deaths in type 2 diabetes are vastly underestimated, pushing for better prevention and reporting strategies.
Study: Contribution of infection to mortality in people with type 2 diabetes: a population-based cohort study using electronic data. Image credits: Dragana Gordic / Shutterstock
From a recent study published in The Lancet Regional Health – Europeresearchers quantified the burden of infection-related mortality in people with type 2 diabetes (T2D) compared to the general population, reporting all recorded causes of death and sepsis mentions.
Background
People with diabetes are at increased risk of infections and all-cause mortality compared to the general population. However, traditional assessments often underestimate infection-related mortality due to the International Classification of Diseases, 10th Revision (ICD-10) coding structures that divide infections across multiple chapters or group them under broader categories, such as respiratory diseases. Furthermore, sepsis, a critical infection-related complication, is rarely reported as the underlying cause of death, despite its increasing prevalence among people with diabetes. For example, in only 11% of deaths where sepsis was mentioned, it was recorded as the underlying cause, indicating systematic under-reporting. Limited research has examined infection-related mortality patterns by ethnicity or in younger populations with T2D. Further research is essential to identify preventable deaths and address disparities in infection-related mortality.
About the study
The current study used a February 2022 extract from the Clinical Practice Research Datalink (CPRD) Aurum database, which covers approximately 16 million active patients from 1,447 general practices in England. More than 90% of participating practices agreed to link their data to external sources, such as Office for National Statistics (ONS) mortality data and the Index of Multiple Deprivation (IMD), a measure of socio-economic status. Researchers did not have access to geographic identification data.
The study used a matched cohort design comparing individuals with T2D to those without diabetes. Participants aged 41-90 years with a diabetes diagnosis were identified and matched with non-diabetic individuals based on age, gender and ethnicity, resulting in 509,403 individuals with T2D and 976,431 matched comparators. Mortality data from 2015-2019 were categorized by specific causes, including cancer, cardiovascular disease, respiratory disease, dementia, diabetes, digestive disorders and infections, using ICD-10 codes.
Cox proportional hazards models estimate five-year mortality risk, adjusting for matched factors and practice regions. Sensitivity analyzes examined additional variables such as deprivation and smoking. To address underreporting, researchers analyzed infection-related mortality using comprehensive ICD-10 coding across chapters, revealing significant underestimation when relying on traditional classifications.
Study results
Among 509,403 individuals with T2D and 976,431 matched individuals without diabetes, baseline characteristics revealed notable differences. The mean age of the T2D group was 67.3 years, with 56% being male. Obesity (body mass index (BMI) ≥ 30) was more common in the T2D group (50% vs. 22%), and a greater proportion lived in the socio-economically disadvantaged areas (23% vs. 16%). About 34% of people with T2D were diagnosed in the past five years.
During the study period (2015-2019), 16.8% of T2D patients died compared to 10.9% of those without diabetes, giving a hazard ratio (HR) of 1.65. The excess relative risk was especially large among younger individuals aged 41-60 years, with HRs nearly four times higher in this group compared with their non-diabetic counterparts. Women with type 2 diabetes had slightly higher HRs (1.71) than men (1.61), although absolute differences in mortality rates were similar (13.9 vs. 13.1 per 1,000 person-years). Ethnic differences were observed, with the highest overall HR among South Asians (1.73) and the lowest among black individuals (1.48). White individuals consistently showed larger absolute mortality differences in younger age groups.
Cardiovascular disease was the leading cause of death in type 2 diabetes (29.7%), followed by cancer (26.9%) and infections (13.0%), including pneumonia. Compared to non-diabetic individuals, T2D individuals showed higher HRs for cardiovascular mortality (2.00), digestive diseases (1.98), and infections (1.82). Sensitivity analyzes adjusting for deprivation, smoking, or the use of different statistical methods confirmed these results.
When using traditional coding methods, infections were often underestimated as a cause of death. By considering all infection-related codes across chapters, the study showed that infections accounted for 13% of T2D deaths, a marked increase from the 1.2% recorded under conventional ICD-10 categories. The highest HR for infections was observed in bone and joint infections (3.95), while lower respiratory tract infections, especially pneumonia, contributed to the largest absolute differences in mortality rates.
Sepsis was often a contributing cause of death rather than the underlying cause of death. Of T2D deaths where sepsis appeared on the death certificate, this was recorded as the underlying cause in only 11%. Including some mention of sepsis, the HR increased to 2.26. This discrepancy highlights the critical need to recognize sepsis as a major contributor to mortality among T2D individuals. Younger individuals with T2D showed particularly high HRs for rare infections, such as bone and joint infections (HR = 9.71) and skin/cellulitis (HR = 6.95), highlighting the vulnerability of this population to specific infections.
Conclusions
In summary, this study highlights the underestimated burden of infection-related mortality in individuals with type 2 diabetes, with infections contributing to 13% of deaths, compared to 1.2% according to traditional ICD-10 classifications. The study also revealed significant differences, including greater differences in absolute mortality among white populations and increased risks among younger individuals with type 2D. Sepsis, often underreported as an underlying cause, was an important contributor. Public health efforts should prioritize infection prevention, early diagnosis and treatment to reduce premature deaths and alleviate economic and societal burdens.