Underweight Linked to Slower Extrapulmonary Tuberculosis Response
Nutritional vulnerability accompanied slower early clinical response in one prospective tuberculosis study, but its contribution cannot be separated confidently from the broader clinical pattern or interpreted as causal.
Evidence
Underweight was more common among patients with a slow eight-week clinical response to treatment for extrapulmonary tuberculosis—tuberculosis outside the lungs—than among responders: 47.8% versus 27.9%. This comparison involved patients whose disease was sensitive to rifampicin, an antibiotic used to treat tuberculosis.
The prospective Central India tuberculosis study included 532 patients at a specialist teaching hospital in Ujjain, Madhya Pradesh. Tuberculosis was confirmed using the Xpert MTB/RIF laboratory test. Published November 2, 2026, the study collected data between 2018 and 2024. Researchers recorded baseline demographic, clinical and laboratory information and assessed clinical response at eight weeks using standardized criteria.
Of 496 patients with rifampicin-sensitive disease, 113 (22.8%) had slow clinical response. Underweight was recorded in 54 of these 113 patients, compared with 107 of 383 responders. Underweight meant a body mass index below 18.5 kg/m²; body mass index expresses weight relative to height. The reported statistical probability value was below 0.001: under the analysis model assuming no group difference, results at least this extreme would occur less than 0.1% of the time.
A multivariable logistic regression—a statistical model examining several factors together—also identified malnutrition as independently associated with slow response. Other reported associations included coexisting illnesses, longer symptom duration and disease at sites outside lymph nodes.
Analysis
Interpretation: The underweight comparison and the adjusted malnutrition finding point in the same direction: nutritional vulnerability accompanied slower early response. These are findings from the same study, not independent confirmation.
Nutrition was part of a broader clinical pattern, rather than an isolated determinant. Importantly, the percentages describe how common underweight was within each response group; they are not the percentages of underweight patients who responded slowly.
Limitations
The observational design cannot establish that underweight or malnutrition caused slow response, or that changing nutrition would improve it. Statistical significance alone does not establish clinical importance.
The available results do not provide an adjusted effect size or an uncertainty range for malnutrition, limiting assessment of the association’s magnitude and precision. They also do not establish whether “malnutrition” used the same definition as body-mass-index-defined underweight.
This is a single-study report from one hospital, so applicability to other settings remains uncertain. The eight-week endpoint does not show whether slower responders eventually recovered.
What to watch
Longer follow-up would show whether early response differences persist. A specific confidence-changing result would be a multi-hospital prospective estimate of the association between clearly defined underweight and slow eight-week response, adjusted for coexisting illnesses, symptom duration and disease site, with an uncertainty range. A precise estimate close to no association would weaken confidence in underweight as a marker of slower response.