Diabetes Is Quietly Undermining Tuberculosis Treatment in Southeastern China

A sweeping five-year analysis of tuberculosis surveillance data from Fujian Province has revealed a troubling and accelerating collision between two of the world’s most burdensome diseases. Researchers at the Fujian Provincial Center for Disease Control and Prevention examined records of pulmonary tuberculosis patients registered between 2020 and 2024 in the China Disease Prevention and Control Information System, and found that nearly one in ten of them was also living with diabetes mellitus. More striking still, that proportion climbed steadily over the study window, rising from 7.3 percent in 2020 to 12.1 percent by 2024. The findings, published in BMC Infectious Diseases, arrive at a moment when global tuberculosis control strategies are being forced to reckon with the metabolic diseases that increasingly share the same patients.

The scale of the study lends considerable weight to its conclusions. As a retrospective cohort analysis, it drew on the complete registry of pulmonary tuberculosis cases in a province home to tens of millions of people, capturing sociodemographic details, diagnostic classifications, and treatment outcomes recorded through a standardized national surveillance platform. Because the data were anonymized and contained no personal identifiers, the ethics committee waived individual informed consent, allowing the researchers to analyze the full population rather than a sample. That completeness matters: sample-based studies of comorbidity can miss rare outcome events, but a province-wide registry offers a near-census view of who develops tuberculosis, who also carries a diabetes diagnosis, and who ultimately fails or completes treatment.

The epidemiological portrait that emerges is distinctive. Patients with the combined condition were predominantly male, mostly over the age of forty, and more likely to be local residents of Fujian rather than migrants. Occupationally, they clustered among industrial and agricultural workers and among people who were unemployed, a distribution that echoes the known social gradient of type 2 diabetes itself. Type 2 diabetes, which accounts for the overwhelming majority of diabetes cases worldwide, develops over years and is strongly associated with age, diet, physical inactivity, and economic circumstances. Its concentration among middle-aged and older working populations in the region means that the tuberculosis patients most likely to face a metabolic comorbidity are precisely those in the later, more vulnerable stages of working life.

The upward trend in comorbidity is not a statistical fluke. The researchers formally tested the year-by-year increase using the Cochran-Armitage test, a statistical procedure designed specifically to detect ordered trends across categorical groups, in this case the proportion of tuberculosis patients with diabetes in each successive year. The steady climb from 7.3 to 12.1 percent mirrors the broader trajectory of diabetes in China, where rapid economic development, urbanization, and dietary change have produced one of the largest diabetic populations on Earth. As diabetes prevalence rises in the general population, its footprint inside tuberculosis registries expands in parallel, effectively coupling the two epidemics at the level of individual patients.

Why should the coexistence of these two diseases matter so much biologically? The answer lies in the immune system. Diabetes impairs innate and adaptive immune function in ways that are directly relevant to tuberculosis. Hyperglycemia, the elevated blood sugar that defines the disease, disrupts the ability of macrophages, the immune cells that engulf and destroy Mycobacterium tuberculosis, to kill the bacteria they ingest. Neutrophil function is degraded, T-cell responses are blunted, and the inflammatory signaling that normally contains infection within granulomas becomes dysregulated. Decades of clinical research have shown that diabetic patients face roughly a two- to three-fold higher risk of progressing from latent tuberculosis infection to active disease, and the current study adds to evidence that the interaction does not stop at disease onset.

Indeed, the treatment outcomes tell the more consequential part of the story. Among patients with both conditions, the investigators used binary logistic regression, a statistical technique that estimates how each characteristic independently changes the odds of a binary outcome, in this case whether anti-tuberculosis treatment succeeded or ended in an adverse result. Three factors stood out after adjustment: being older than forty, having a pathogen-positive diagnosis, meaning that bacteria were directly detected in sputum or culture, and undergoing retreatment after a previous episode of tuberculosis. Each of these markers flags a patient whose disease is harder to cure, whose treatment is longer or more toxic, or whose adherence and drug tolerance may be compromised.

Pathogen-positive status deserves particular attention because it sits at the intersection of biology and public health. Patients whose sputum contains detectable bacilli carry higher bacterial burdens, take longer to convert to negative cultures, and remain infectious for longer, sustaining chains of transmission in the community. If diabetes delays bacterial clearance, as mechanistic studies suggest it can, then diabetic tuberculosis patients who are pathogen-positive represent a double hazard: worse prospects for their own recovery and a prolonged window in which they can infect others. The finding that retreatment cases fared worse is consistent with the broader tuberculosis literature, since previously treated patients are more likely to harbor drug-resistant organisms or to have residual lung damage that limits the effectiveness of standard regimens.

Age over forty, the third risk factor, is a reminder that comorbidity management is fundamentally a geriatric and middle-age medicine problem within infectious disease control. Older patients metabolize drugs differently, accumulate comorbidities beyond diabetes, and often juggle multiple medications whose interactions and side effects complicate the six-month standard tuberculosis regimen. For a patient taking insulin or oral hypoglycemic agents alongside rifampicin, isoniazid, pyrazinamide, and ethambutol, the pharmacological and practical burdens are considerable. Rifampicin, a cornerstone of tuberculosis therapy, is a potent inducer of liver enzymes that can destabilize glycemic control and alter the metabolism of diabetes drugs, forcing clinicians to navigate a moving target on both fronts simultaneously.

The authors conclude that patients with the combined condition who are older than forty, pathogen-positive, or undergoing retreatment warrant more attention and enhanced management. In practical terms, that recommendation points toward bidirectional screening, in which every tuberculosis patient is tested for diabetes and diabetic patients with suggestive symptoms are evaluated for tuberculosis, alongside integrated care models that address glycemic control as part of the tuberculosis treatment package. The World Health Organization has already identified the diabetes-tuberculosis nexus as a priority for collaborative activities, and this study provides the kind of granular, regional evidence that national programs need to calibrate such interventions. Fujian’s experience suggests that as diabetes prevalence continues to climb, tuberculosis programs that ignore metabolic comorbidity will see their treatment success rates erode.

There are limits worth acknowledging. As a retrospective analysis of surveillance records, the study captures diabetes diagnoses that were documented in the system, and undiagnosed or unrecorded diabetes among tuberculosis patients would mean the true comorbidity burden is likely higher than reported. Surveillance data also cannot fully disentangle whether adverse outcomes stem from diabetes itself or from the socioeconomic and clinical characteristics that travel with it. Even so, the message is difficult to ignore. Two ancient scourges of human health, one infectious and one metabolic, are converging in the same patients, in the same clinics, and increasingly in the same treatment plans. The five-year Fujian dataset shows that this convergence is not a future threat but a present, growing reality, and that the patients caught between the two diseases are identifiable, measurable, and, with targeted management, treatable.

Subject of Research: Prevalence and treatment outcomes of pulmonary tuberculosis-diabetes mellitus comorbidity in southeastern China

Article Title: Pulmonary tuberculosis-diabetes mellitus comorbidity: prevalence and risk factors associated with treatment outcomes in Southeastern China, 2020–2024

Article References: Chen, K., Lin, S., Dai, Z., Chen, D., Chen, J., Lin, Y., & Zhou, Y. (2026). Pulmonary tuberculosis-diabetes mellitus comorbidity: prevalence and risk factors associated with treatment outcomes in Southeastern China, 2020–2024. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14626-3

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14626-3

Keywords: tuberculosis, diabetes mellitus, comorbidity, treatment outcomes, risk factors, Fujian Province, China, epidemiology, retrospective cohort study, public health, BMC Infectious Diseases, surveillance data