Association of undernutrition and tuberculosis outcomes: a retrospective multicentre cohort study in Italy

Undernutrition remains one of the most powerful yet neglected determinants of tuberculosis (TB) incidence and progression. Although its impact on TB outcomes has been extensively described in low-income countries, limited evidence is available from high-income settings. This study aimed to assess the association between body mass index (BMI) strata at treatment initiation and TB-related outcomes in a large multicentre cohort from Italy.

We conducted a retrospective, multicentre observational study including all adults diagnosed with TB between January 2021 and August 2025 in 16 Infectious and Tropical Diseases Units across seven Italian regions. Patients were stratified by BMI at treatment initiation (<16, 16-16.99, 17-18.49, 18.5-24.99, and ≥25 kg/m²). Outcomes included time to sputum conversion, length of hospital stay, occurrence and severity of adverse events, incomplete treatment (including failure, death, and loss to follow-up), and loss to follow-up. Multilevel regression models adjusting for study centre and relevant confounders were used to assess associations between BMI strata and each outcome.

A total of 709 people with TB were included (median age, 38 years [interquartile range 28-53]; 488/709, 75.5% male). Overall, 299 of 709 (42%) were underweight (BMI <18.5 kg/m²). Compared with people with TB with normal BMI (18.5-24.99 kg/m²), those with severe undernutrition (BMI <16 kg/m²) showed higher rates of adverse events (54/74 [73.0%] vs. 101/410 [24.6%]; adjusted OR [aOR] 12.17, 95% CI 6.41-23.08), incomplete treatment (53/74 [71.6%] vs. 87/410 [21.2%]; aOR 18.28, 95% CI 6.70-49.80), and loss to follow-up (41/74 [55.4%] vs. 81/410 [19.7%]; aOR 5.99, 95% CI 2.58-13.90). Median hospital stay was also longer in this group (50 days [interquartile range 24-66] vs. 30 days [interquartile range 18-60]), corresponding to a 23% adjusted increase. Weight gain during the first 6 months of treatment was not significantly associated with major outcomes.

Severe underweight at treatment initiation is a major independent predictor of poor outcomes in TB. These findings highlight the need for systematic nutritional assessment and targeted interventions within TB programmes, even in high-income, low-incidence settings.