When Tuberculosis Meets Diabetes, Blood Fats Tell a Surprising Story

Tuberculosis and diabetes are two of the world’s most burdensome diseases, and they increasingly collide in the same patients. A new systematic review and meta-analysis published in BMC Infectious Diseases by Samuel Yahaya and colleagues at the University of Cape Coast in Ghana has now pulled together the scattered evidence on what happens to blood fats when these two conditions coexist. The findings reveal a strikingly asymmetric picture: patients carrying both diseases show higher triglycerides than those with tuberculosis alone, yet lower levels of nearly every lipid parameter than those with diabetes alone. The work offers one of the most quantitative syntheses to date of the metabolic signature of tuberculosis-diabetes comorbidity, a combination that affects millions in low- and middle-income countries where both epidemics overlap.

The research team followed PRISMA guidelines and registered their protocol in PROSPERO before searching PubMed, Scopus, and Google Scholar through April 7, 2026. The searches returned 150 records from PubMed, 679 from Scopus, and 70 from Google Scholar, from which nine observational studies comprising 1,494 participants met the inclusion criteria. Each study reported lipid parameters — high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), and total cholesterol (TC) — in patients stratified by tuberculosis and diabetes status. Methodological quality was appraised with the Newcastle-Ottawa Scale, and the certainty of evidence was graded using GRADE, a framework that rates confidence in findings from high to very low.

Statistically, the team used random-effects models to pool mean differences and pooled means, estimating between-study variance with restricted maximum likelihood (REML) and applying Hartung-Knapp adjustments to confidence intervals — techniques that guard against overly optimistic precision when studies are few and heterogeneous. Heterogeneity was quantified with the I² statistic and τ², and 95% prediction intervals were calculated to show where future study results might plausibly fall. In the primary comparison, patients with tuberculosis-diabetes comorbidity (TB-DM) were set against patients with tuberculosis only (TB-only); in the secondary comparison, TB-DM patients were set against those with diabetes only (DM-only).

The headline result concerns triglycerides. Compared to the TB-only group, patients with TB-DM exhibited higher concentrations of LDL-C (mean difference 14.99 mg/dL; 95% CI 1.28–28.71), TG (mean difference 30.13 mg/dL; 95% CI 17.58–42.67), and TC (mean difference 17.29 mg/dL; 95% CI 5.42–29.16). HDL-C, often called the good cholesterol, showed no significant difference between the two groups (mean difference −1.15 mg/dL; 95% CI −5.07 to 2.77). A mean difference of roughly 30 mg/dL in triglycerides is clinically meaningful, pointing toward a more atherogenic metabolic milieu when diabetes superimposes on active tuberculosis.

Yet the story is more nuanced than the headline numbers suggest. A principal sensitivity analysis showed that the triglyceride elevation remained robust, whereas the LDL-C and total cholesterol findings lost statistical significance once the analysis was stress-tested. This means the most defensible conclusion from the primary comparison is that TB-DM comorbidity is associated with higher triglycerides relative to tuberculosis alone, while the cholesterol findings should be interpreted with caution. The authors rated the overall certainty of evidence as very low across all outcomes, a reminder that observational data, however carefully pooled, cannot establish causation.

The secondary comparison flipped the direction of nearly every effect. Relative to the DM-only group, TB-DM patients exhibited significantly lower concentrations across all four lipid parameters — HDL-C, LDL-C, triglycerides, and total cholesterol. Heterogeneity in this secondary analysis ranged from just 1% to 67%, far more consistent than the substantial heterogeneity (I² greater than 70%) observed in the primary comparison. Biologically, this pattern is coherent: active tuberculosis is known to suppress appetite, impair nutrient absorption, and drive inflammatory catabolism, all of which can pull lipid levels downward even in patients whose underlying diabetes would normally push them up.

Exploratory subgroup analyses added further texture. Statistical differences were detected for LDL-C, total cholesterol, and triglycerides across subgroups, whereas no subgroup difference was found for confounder adjustment, geographic region, or study design. In other words, the lipid signal did not systematically depend on whether studies adjusted for key covariates, where they were conducted, or how they were designed — though with only nine studies, such subgroup tests have limited power. The very low GRADE rating across all outcomes underscores that these are hypothesis-generating findings rather than practice-changing conclusions.

Why do these lipid shifts matter? Triglyceride-rich lipoproteins and low HDL-C are established markers of cardiovascular risk, and diabetes already triples the likelihood of cardiovascular disease. If TB-DM patients carry a doubly dysregulated lipid profile — elevated triglycerides from the diabetic side and catabolic lipid depletion from the tuberculous side — clinicians may need to monitor cardiovascular risk more aggressively during and after tuberculosis treatment. Conversely, the finding that TB-DM patients show lower lipids than DM-only patients suggests that tuberculosis itself reshapes the metabolic landscape, potentially masking the typical diabetic dyslipidaemia and complicating risk stratification.

The study has limitations inherent to its evidence base. All nine included studies were observational, meaning confounding by age, body mass index, antidiabetic medication, antitubercular drug timing, and nutritional status cannot be excluded. The very low certainty rating reflects this, along with the substantial heterogeneity in the primary analysis. The authors also note that their search was restricted to three databases through a single cutoff date, and that publication bias — the tendency for studies with null findings to go unpublished — could not be formally ruled out with so few studies.

Nevertheless, the meta-analysis provides the most comprehensive quantitative picture to date of lipid alterations in tuberculosis-diabetes comorbidity. Its central message — that triglycerides rise when diabetes joins tuberculosis, while nearly all lipid fractions fall relative to diabetes alone — gives researchers a concrete metabolic signature to investigate in prospective cohorts. As both conditions continue to expand in overlapping geographic hotspots, from sub-Saharan Africa to South Asia, understanding how they jointly remodel blood fat metabolism could prove essential for protecting the hearts of patients already fighting on two fronts.

Subject of Research: Lipid profile alterations in pulmonary tuberculosis patients with and without diabetes mellitus: a systematic review and meta-analysis

Article Title: Lipid profile alterations in pulmonary tuberculosis patients with and without diabetes mellitus: a systematic review and meta-analysis

Article References: Yahaya, S., Goumboundi, I. A., Saahene, R. O., Dankwa, K., Acquah, S., & Ephraim, R. K. D. (2026). Lipid profile alterations in pulmonary tuberculosis patients with and without diabetes mellitus: a systematic review and meta-analysis. BMC Infectious Diseases, 26(1), Article 1727. https://doi.org/10.1186/s12879-026-14453-6

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14453-6

Keywords: Lipid, profile, alterations, pulmonary, tuberculosis, patients, without, diabetes, mellitus, systematic, review, meta-analysis