Type 2 diabetes is a significant driver of multiple long-term conditions (MLTCs) due to the systemic effects of chronic hyperglycemia, insulin resistance, and related pathophysiological dysfunctions. While progress has been made in reducing classic diabetes complications like cardiovascular disease, the burden of MLTCs is increasing. This diversification of complications may be due to increased life expectancy, reduced cardiovascular disease risk in older adults, and increasing obesity and early-onset diabetes in younger adults. MLTCs create substantial burdens on healthcare systems, clinical management, and individual quality of life. However, existing metrics are inadequate for characterizing the complete burden of MLTCs and their impact on life expectancy across the life course. This study aimed to quantify the burden of diabetes-related MLTCs in England using a comprehensive dataset and novel metrics to assess years of life spent with and lost due to these conditions. This would provide valuable information for population monitoring, resource allocation, and the development of effective prevention and management strategies.

Previous research has established a strong association between diabetes and various complications, both classic microvascular and macrovascular diseases, and other conditions such as cancer, infections, respiratory diseases, liver disease, and dementia. Studies have shown an increased likelihood of developing two, three, or more conditions in individuals with diabetes. Clustering analyses have suggested that comorbid conditions may aggregate around mental health outcomes or aging-related conditions. However, existing studies have limitations in quantifying the burden of MLTCs in terms of years spent with and years of life lost, crucial metrics for understanding the impact on both individuals and health systems. This study addresses this gap by employing novel methods to quantify these metrics.

The study utilized the National Bridges to Health Segmentation Dataset, which comprises data from over 15 sources including all adults registered with a general practitioner (GP) in England. The data, highly representative of the English population, included longitudinal information on the onset of 35 conditions. The researchers developed new metrics to assess the MLTC burden associated with diabetes, including traditional prevalence metrics and estimates of age of onset, years spent with, and years of life lost due to MLTCs, from both individual and community perspectives. To estimate years spent and lost, a three-state illness-death Markov model was constructed. This model incorporated age-dependent transition probabilities between three states: healthy, illness (presence of at least two specified MLTCs), and death. The model calculated lifetime risk of MLTCs, median age of onset, years of life lived with MLTCs, age at death, and years of life lost. Community-level metrics were calculated by considering the prevalence of conditions in the population and scaling the average metrics to 1,000 individuals. The analysis focused on combinations of diabetes with other conditions where at least 1000 observations were available for each transition type in the model. Specific conditions were excluded due to insufficient data or incompatibility with the model.

Among 3,663,429 adults (7.8%) with diagnosed diabetes in England, the prevalence of MLTCs was significantly higher than in the general population across all age ranges. By age 50, approximately one-third of adults with diabetes had at least three conditions, compared to around 17% in the general population (the same prevalence not being reached in the general population until after age 70 for women and 65 for men). The most prevalent conditions varied by age group. In older adults (≥70 years), hypertension, coronary heart disease (CHD), osteoarthritis, and atrial fibrillation were most common. Middle-aged adults showed similar patterns but with lower prevalence. In younger adults (20-49 years), hypertension and CHD were present, but depression, asthma, and serious mental illness were more prevalent. Analysis of bivariate combinations showed that most conditions exceeded expected prevalence, indicating a non-random association. The median age of onset for at least two MLTCs was 67 years in women and 66 years in men. Individuals with more conditions had fewer years living with MLTCs and died earlier. Analyzing years spent with and years of life lost, both from individual and community perspectives, revealed that hypertension, depression, cancer, and CHD accounted for the largest community-level burden in terms of years spent and lost. In contrast, serious mental illness, alcohol dependence, and learning disabilities had a large individual impact on years spent and lost but a smaller community-level impact due to their lower prevalence. Asthma had a large impact on years spent with but a smaller impact on years lost.

This study demonstrates the extensive and diverse burden of diabetes-associated MLTCs across the lifespan. The findings extend beyond the well-established classic complications, showing a substantial burden even by middle age. The variation in prevalent conditions by age highlights the need for age-specific prevention and management approaches. The results underscore the importance of integrating the management of diabetes with other co-occurring conditions, particularly mental health conditions, where strong bidirectional associations exist. The large individual burden of certain conditions, such as those involving serious mental illness, warrants targeted interventions that coordinate diabetes management with the management of the co-occurring condition. The study highlights three key areas for future research: understanding the interplay of risk factors for different types of comorbid conditions and progression to MLTCs; optimizing models of care to address the complexities of MLTCs; and determining the most effective prevention strategies for MLTCs.

This study provides a comprehensive assessment of the burden of diabetes-associated MLTCs in England, quantifying the impact in terms of years spent with and years of life lost. The findings emphasize the significant individual and community-level burdens of these conditions, highlighting the urgency for innovative prevention and management approaches. Future research should focus on identifying modifiable risk factors, refining models of care, and evaluating the effectiveness of targeted interventions to reduce the burden of MLTCs and improve the lives of individuals with diabetes.

The study's data sources were predominantly hospital and community-coded datasets, potentially leading to under-ascertainment of conditions typically diagnosed in primary care. Eye diseases, a significant source of disability in diabetes, were not included. The age of median onset was determined by the second condition, potentially obscuring variations in the onset of the first condition. The analysis did not adjust for additional conditions that may be differentially associated with the combinations under study. The study was limited to 35 prioritized conditions due to the computational demands of modeling a larger number of combinations and the need to focus on conditions of higher public health priority. Finally, the data did not differentiate between type 1 and type 2 diabetes.