Undernutrition is identified as the leading risk factor for tuberculosis (TB). Evidence on the impact of routine, programmatic nutritional support delivered as part of national TB programmes is limited. In West Africa, Benin provides in-kind food assistance to all people treated for drug-susceptible TB, while neighboring Togo does not. This policy difference offered a natural experiment to evaluate associations between programmatic nutritional support and clinical and programmatic treatment outcomes.
The investigators conducted a prospective cohort study at 13 clinical sites in Benin and Togo between September 2023 and June 2024. The design compared outcomes in people receiving programmatic nutritional support versus those not receiving it, with Beninese non-recipients included as an internal comparison group to help account for contextual country differences.
A total of 769 participants were enrolled. Of these, 450 participants received programmatic nutritional support and 319 did not. The report describes recipients (those in Benin receiving in-kind food support) and non-recipients (participants in Togo and Beninese participants who did not receive support).
Primary outcomes prespecified in the report were: achieving ≥5% weight gain at month 2 of treatment, change in 6-minute walk test (6MWT) distance as a measure of functional recovery, and pill-count–based adherence. Secondary and additional outcomes included ≥10% weight gain at month 6, sputum conversion, overall treatment success, and mortality.
The authors used multivariable regression models adjusted for prespecified covariates to compare outcomes between recipients and non-recipients. Effect estimates reported in the abstract include adjusted odds ratios (aORs) for categorical outcomes and adjusted β coefficients for continuous outcomes, with 95% confidence intervals provided for key results.
Weight gain: Recipients of programmatic nutritional support had higher odds of achieving ≥5% weight gain at month 2 (aOR 1.57, 95% CI 1.13–2.19) and higher odds of ≥10% weight gain at month 6 (aOR 1.92, 95% CI 1.35–2.74).
Functional recovery: Participants who received nutritional support showed greater improvement in the 6-minute walk test, with an adjusted mean increase of 40.6 m (95% CI 26.5–54.6) compared with non-recipients.
Adherence: Pill-count adherence was higher among recipients (aOR 3.43, 95% CI 1.81–6.51).
Mortality: Mortality during treatment was lower among recipients of nutritional support (aOR 0.32, 95% CI 0.11–0.93).
Sputum conversion and treatment success: No statistically significant differences in sputum conversion or overall treatment success were reported between recipients and non-recipients.
Internal comparison: The authors note that Beninese non-recipients resembled Togolese participants across outcomes, supporting the use of the neighboring country as a comparison in this natural experiment.
In this natural experiment, programmatic nutritional support provided as in-kind food during treatment for drug-susceptible TB was associated with improved early and later weight gain, greater functional recovery measured by 6MWT, substantially higher pill-count adherence, and lower mortality. The authors interpret these associations as supportive evidence for integrating nutritional support into national TB programmes to improve clinical recovery and survival during treatment.
The study authors declared that ethical approvals were obtained from the Boston Medical Center and Boston University Medical Campus Institutional Review Board. The trial is registered (ClinicalTrials.gov identifier NCT06084715). The manuscript is a preprint and has not been peer reviewed. The authors state that data are available upon reasonable request to the corresponding author.
This summary reflects the information reported in the preprint abstract and author declarations. As a preprint, the findings have not been evaluated through peer review. Details beyond the abstract—such as the full list of adjusted covariates, granular baseline characteristics, adherence measurement procedures, and any subgroup analyses—were not reported in the abstract and would need review of the full manuscript or data to assess potential confounding, generalizability, and programmatic feasibility.
Programmatic in-kind nutritional support during treatment for drug-susceptible TB in this West African natural experiment was associated with improved weight and functional recovery, better pill-count adherence, and lower mortality, while sputum conversion and overall treatment success were similar between groups. These findings, as reported in the preprint, support consideration of nutritional support integration into national TB programmes, pending peer review and further implementation-specific evaluation.