Abstract
Background
Undernutrition is the leading risk factor for tuberculosis (TB) globally. Its impact on treatment outcomes is poorly defined.
Methods
We conducted a prospective cohort analysis of adults with drug-sensitive pulmonary TB at 5 sites from 2015–2019. Using multivariable Poisson regression, we assessed associations between unfavorable outcomes and nutritional status based on body mass index (BMI) nutritional status at treatment initiation, BMI prior to TB disease, stunting, and stagnant or declining BMI after 2 months of TB treatment. Unfavorable outcome was defined as a composite of treatment failure, death, or relapse within 6 months of treatment completion.
Results
Severe undernutrition (BMI <16 kg/m2) at treatment initiation and severe undernutrition before the onset of TB disease were both associated with unfavorable outcomes (adjusted incidence rate ratio [aIRR], 2.05; 95% confidence interval [CI], 1.42–2.91 and aIRR, 2.20; 95% CI, 1.16–3.94, respectively). Additionally, lack of BMI increase after treatment initiation was associated with increased unfavorable outcomes (aIRR, 1.81; 95% CI, 1.27–2.61). Severe stunting (height-for-age z score <−3) was associated with unfavorable outcomes (aIRR, 1.52; 95% CI, 1.00–2.24). Severe undernutrition at treatment initiation and lack of BMI increase during treatment were associated with a 4- and 5-fold higher rate of death, respectively.
Conclusions
Premorbid undernutrition, undernutrition at treatment initiation, lack of BMI increase after intensive therapy, and severe stunting are associated with unfavorable TB treatment outcomes. These data highlight the need to address this widely prevalent TB comorbidity. Nutritional assessment should be integrated into standard TB care.
Keywords: malnutrition, undernutrition, tuberculosis, outcomes, India
Undernutrition is the leading risk factor for TB globally. This multicenter prospective cohort analysis shows that severe undernutrition at baseline or before TB disease, severe stunting, and lack of weight gain during intensive therapy are associated with unfavorable outcomes.
In 2021, an estimated 10.6 million individuals developed tuberculosis (TB) disease and 1.6 million died, making it the second most lethal infection globally after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1]. Undernutrition, the most common cause of secondary immunodeficiency worldwide, impairs innate and adaptive immunity against Mycobacterium tuberculosis [2]. In India, the population attributable fraction of TB due to undernutrition in some regions is as high as 61.5% in women and 57.4% in men [3]. The SARS-CoV-2 pandemic has exacerbated food insecurity worldwide, with the greatest burden falling on high TB burden countries, and climate change may further exacerbate undernutrition globally [4, 5]. Understanding the impact of this increasingly prevalent risk factor on TB treatment outcomes is critical.
Undernutrition is associated with higher bacillary grade in sputum and increased cavitation among undernourished persons with TB (PWTB) [6–8]. Pharmacotherapy may be suboptimal in undernourished PWTB due to reduced absorption of rifampin and isoniazid or increased toxicity from ethambutol and aminoglycosides [9, 10]. Baseline undernutrition is associated with delayed sputum conversion and increased mortality [2, 11–13]. Inadequate weight gain during therapy may also be associated with poor outcomes, although this finding has not been validated in India where phenotypic differences and a lower prevalence of human immunodeficiency virus (HIV) infection may skew the relationship between weight gain and treatment response [2, 14, 15].
Available studies on the impact of nutrition on treatment outcomes have had several gaps that we sought to address. First, most extant studies are underpowered, which makes the association between undernutrition and individual outcomes such as death difficult to define [16]. Second, most studies use either body mass index (BMI) as a continuous variable or dichotomize nutritional status based on a single BMI cutoff (typically 18.5 kg/m2). Using continuous BMI is problematic because it assumes a linear relationship between the degree of undernutrition and treatment outcomes. A binary BMI variable cannot differentiate between different degrees of undernutrition, which may have distinct impacts on health and immunity [17]. Third, BMI at treatment initiation cannot distinguish between undernutrition preceding TB disease and weight loss due to delay in TB treatment, and no study has assessed the impact of premorbid BMI on treatment outcomes [18]. Also, no epidemiological study, to our knowledge, has adjusted for symptom duration to guard against reverse causality in their analysis. Fourth, no study has investigated the potential impact of stunting, which may be associated with immune defects [19]. Last, most epidemiological studies of undernutrition in India are single-center studies, which are not easily generalizable to India's diverse population.
To assess the impact of undernutrition on TB treatment outcomes in India and address the lacunae in previous publications, we conducted a large multicenter prospective cohort study through the Regional Prospective Observational Research for Tuberculosis (RePORT) India Consortium.
METHODS
Study Design: Prospective Cohort Analysis
We used data collected at 5 academic medical centers (Byramjee Jeejeebhoy Government Medical College, Pune; Christian Medical College, Vellore; Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry; National Institute of Research in Tuberculosis, Chennai; and Prof. M. Viswanathan Diabetes Research Centre, Chennai) between 2015 and 2019, which are part of the RePORT India Consortium [3, 7, 20, 21]. We included participants aged >18 years with drug-sensitive pulmonary TB who had received less than 7 days of antimycobacterial therapy. TB was diagnosed using sputum culture and consistent findings on chest radiography. The inclusion criteria were amended during the study to require sputum smear positivity, although some PWTB with negative sputum smears were enrolled before protocol harmonization. We used follow-up data up to 24 months after enrollment.
Predictors
Our primary predictor was BMI at treatment initiation that was categorized into severe undernutrition (<16 kg/m2), moderate undernutrition (16–16.99 kg/m2), mild undernutrition (17–18.49 kg/m2), normal (18.5–22.99 kg/m2), and overweight or obese (≥23 kg/m2) [22]. We used cutoffs for the overweight category based on consensus guidelines for Asian Indians [23]. A secondary predictor was premorbid BMI, that is, the participant's BMI prior to onset of TB symptoms, which we calculated using weight loss self-reported by participants. Premorbid BMI was categorized using the same cutoffs as BMI at treatment initiation. Another secondary predictor was unchanged or decreased BMI after 2 months of treatment. Our last predictor was stunting, which we divided into moderate and severe, defined as height-for-age z score (HAZ) −2 to −3 and <−3, respectively. The reference range was HAZ >−2 [24].
Outcomes
The primary outcome was a composite of failure, relapse, and death. Failure was defined as having positive culture results after 5 months of treatment or completing at least 4 months of treatment but having progression of disease, persistent or recurrent symptoms, and/or signs of TB adjudicated to be due to TB. Relapse denotes culture-confirmed or clinically diagnosed TB within 6 months of treatment completion. Death refers to all-cause mortality during follow-up.
Potential Confounders
We included age categorized by decades, sex, monthly household income discretized as <Rupees (Rs.) 3000 [USD 37.5], Rs. 3001–5000 [USD 37.5–62.5], Rs. 5001–10000 [USD 62.5–125], and > Rs. 10 000 [USD 125] as demographic variables. Additionally, we considered disease-related factors such as months of symptom duration (fever, cough, or weight loss) prior to enrollment, weeks to positivity on liquid culture, grade of sputum on Ziehl-Neelsen smear (negative, 1+, 2+, 3+, or scanty), weight loss prior to treatment, and cavitation on chest radiograph at diagnosis. Last, we accounted for comorbidities such as smoking status (never, former, or current); alcohol use disorder, which was defined as having an Alcohol Use Disorders Identification Test-Concise (AUDIT-C) score ≥4 for males and ≥3 for females; diabetes (known diabetes, random blood glucose ≥200 mg/dL, or hemoglobin A1c ≥6.5%); HIV positivity at treatment initiation; and antiretroviral status [25].
Statistical Methods
We generated descriptive statistics for the predictor variables and the potential confounders using the t test and Wilcoxon rank sum tests for parametric and nonparametric variables, respectively. We assessed normality using the QQ plots and the Kolmogorov-Smirnov test. We used the χ2 test for categorical variables. We plotted Kaplan–Meier curves for our predictors of interest and conducted log-rank tests to assess the differences in time to unfavorable outcomes for participants in different nutritional categories. We then pursued univariate analysis using Poisson regression, in which we used person-time as an offset to calculate incidence rate ratios (IRR). Person-time was calculated from the time of treatment initiation to the occurrence of the first mutually exclusive outcome of interest (failure/recurrence/death), lost to follow-up (LTFU), or until right-censoring at 24 months of follow-up. Finally, we proceeded with multivariable analysis in which we included age, sex, and cough duration in the multivariable model a priori and again used person-time as an offset to calculate adjusted IRR (aIRR). We included other potential confounders based on prior literature that were significant at the P < .2 level. Last, we assessed interaction between the undernutrition variables and sex, cough duration, smoking status, alcohol use disorder, and diabetes by inputting interaction terms into the multivariable regression model and conducting a likelihood ratio test.
Sensitivity Analyses
We conducted the analysis with 3 alternative outcomes: a composite outcome of failure and relapse, death, and LTFU based on India's National Tuberculosis Elimination Programme records or inability to reach the participant prior to treatment completion [26]. Additionally, we stratified the unchanged or decreased BMI analysis by nutritional status at treatment initiation (BMI <18.5 kg/m2 vs ≥18.5 kg/m2). Based on previous observational data, we believed cavitation to be a mediator on the causal pathway [7]. Therefore, we excluded this covariate from our primary analysis but explored it in sensitivity analyses. Last, we explored whether using BMI cutoffs suggested by the World Health Organization (WHO), where 25 kg/m2 is the upper limit of normal BMI, altered our conclusions regarding the impact of BMI at treatment initiation and premorbid BMI [27].
RESULTS
Descriptive Statistics
After excluding participants who had drug-resistant TB, had missing outcome data, were inadvertently enrolled, or were aged <18 years, we had 2931 participants in our analysis (Figure 1). The median follow-up was 8.47 months (interquartile range [IQR], 5.93–13.87). Of 2931 persons with TB, 1610 (54.9%) were undernourished (BMI <18.5) and 1060 (36.2%) were stunted at treatment initiation. Among 1587 persons who reported their weight prior to the onset of TB symptoms, 454 (28.6%) were undernourished.
Figure 1.
Selection of participants.
Of 2931 participants, 2300 (78.5%) completed treatment or had documented cure and 280 (9.6%) had unfavorable outcomes, which comprised 101 (3.4%) deaths, 129 (4.4%) failures, and 50 (1.7%) relapses within 6 months (Table 1). In addition, 351 individuals (12.0%) were LTFU. Most deaths occurred early (median, 2.2 months after enrollment; IQR, 1.6–6.6). After 2 months of treatment, 713 of 1203 (59.2%) participants with repeat weight assessments in the primary analysis had unchanged or decreased BMI (Table 2).
Table 1.
Descriptive Statistics of Participants With Favorable and Unfavorable Outcomes
| Variable | Total | Favorable Outcome (n = 2300) |
Unfavorable Outcomea (n = 280) |
P Value |
|---|---|---|---|---|
| Age, n (%), y | 2580 | .02 | ||
| 18–29 | 596 (25.9) | 51 (18.2) | ||
| 30–39 | 490 (21.3) | 60 (21.4) | ||
| 40–49 | 597 (26.0) | 86 (30.7) | ||
| 50–59 | 424 (18.4) | 58 (20.7) | ||
| 60–69 | 164 (7.1) | 17 (6.1) | ||
| 70–82 | 29 (1.3) | 8 (2.9) | ||
| Male sex, n (%) | 2580 | 1629 (70.8) | 234 (83.6) | <0.001 |
| Monthly household income (₹), n (%) | 2336 | .34 | ||
| <3000 (<USD 37.5) | 167 (8.0) | 21 (8.5) | ||
| 3001–5000 (USD 37.5–62.5) | 486 (23.3) | 45 (18.1) | ||
| 5001–10 000 (USD 62.5–125) | 768 (36.8) | 99 (39.9) | ||
| >10 000 (>USD 125) | 667 (31.9) | 83 (33.5) | ||
| Symptom duration, mean (SD), mo | 2467 | 1.97 (3.58) | 2.80 (5.53) | <.01 |
| Pretreatment weight lost, mean (SD), kg | 1431 | 5.64 (5.64) | 7.21 (5.89) | <.01 |
| Time to positive liquid culture, mean (SD), wk | 1743 | 1.33 (0.68) | 1.36 (0.72) | .53 |
| Person-time, mean (SD), mo | 2563 | 12.07 (6.92) | 9.18 (7.05) | <.001 |
| Sputum smear at baseline, n (%) | 2346 | .53 | ||
| Negative | 231 (11.0) | 29 (11.4) | ||
| 1+ | 739 (35.3) | 85 (33.3) | ||
| 2+ | 634 (30.3) | 87 (34.1) | ||
| 3+ | 433 (20.7) | 45 (17.6) | ||
| Scanty | 54 (2.6) | 9 (3.5) | ||
| Cavitation, n (%) | 1691 | 779 (51.9) | 105 (55.3) | .42 |
| Smoking, n (%) | 2576 | <.001 | ||
| Never | 1411 (61.5) | 137 (48.9) | ||
| Former | 422 (18.4) | 69 (24.6) | ||
| Current | 463 (20.2) | 74 (26.4) | ||
| Alcohol use disorder,b n (%) | 2580 | 263 (11.4) | 67 (23.9) | <.001 |
| Diabetes,c n (%) | 2569 | 764 (33.3) | 75 (27.0) | .04 |
| Human immunodeficiency virus, n (%) | 2580 | 47 (2.0) | 12 (4.3) | .03 |
| Antiretroviral therapy status, n (%) | 55 | 22 (55.0) | 4 (30.8) | .23 |
| Treatment initiation BMI, n (%), kg/m2 | 2565 | <.001 | ||
| <16 | 521 (22.8) | 111 (40.1) | ||
| 16–16.99 | 319 (13.9) | 43 (15.5) | ||
| 17–18.49 | 387 (16.9) | 36 (13.0) | ||
| 18.5–22.99 | 749 (32.7) | 62 (22.4) | ||
| >23 | 312 (13.6) | 25 (9.0) | ||
| Premorbid BMI, n (%), kg/m2 | 1423 | .06 | ||
| <16 | 81 (6.3) | 17 (12.0) | ||
| 16–16.99 | 95 (7.4) | 15 (10.6) | ||
| 17–18.49 | 177 (13.8) | 17 (12.0) | ||
| 18.5–22.99 | 561 (43.8) | 58 (40.8) | ||
| >23 | 367 (28.6) | 35 (24.6) | ||
| Stunting | 2567 | <.01 | ||
| Not stunted (HAZ >−2) | 1451 (63.4%) | 152 (54.9%) | ||
| Moderately stunted ( −3> HAZ ≤−2) | 641 (28.0%) | 86 (31.0%) | ||
| Severely stunted (HAZ ≤3) | 198 (8.6%) | 39 (14.1%) |
Abbreviations: BMI, body mass index; HAZ, height-for-age z score; SD, standard deviation; USD, United States dollars.
Unfavorable outcome is a composite of death, treatment failure, and relapse within 6 months of treatment completion.
Alcohol use disorder defined as having an Alcohol Use Disorders Identification Test-Concise score ≥4 for males and ≥3 for females.
Diabetes defined as known history of diabetes, random blood glucose ≥200 mg/dL, or hemoglobin A1c ≥ 6.5%.
Table 2.
Descriptive Statistics of Weight Change After 2 Months of Therapy
| Group | Total | Favorable Outcome | Unfavorable Outcomea | P Value |
|---|---|---|---|---|
| Weight change, kg | ||||
| Overall, mean (SD) | 1203 | 1.16 (2.55) | 0.92 (2.63) | .27 |
| Among those without weight gain, mean (SD) | 628 | −0.40 (1.13) | −0.61 (1.33) | <.01 |
| Among those who gained weight, mean (SD) | 575 | 3.18 (2.45) | 3.32 (2.43) | .67 |
| Unchanged or decreased body mass index, n (%) | 1203 | 626 (51.5) | 87 (55.1) | .44 |
Abbreviation: SD, standard deviation.
Unfavorable outcome is a composite of death, treatment failure, and relapse within 6 months of treatment completion.
Survival curves demonstrate different rates of unfavorable outcomes among individuals with different BMIs at treatment initiation (P < .001; Figure 2A ) and premorbid BMIs (P < .001; Figure 2B ). Those with unchanged or decreased BMI had an increased rate of unfavorable outcomes (P = < .001; Figure 2C ). Severely stunted individuals had an increased rate of unfavorable outcomes compared with those with moderate stunting or no stunting (P = .03; Figure 2D ).
Figure 2.
Survival curves. Survival curves for participants with different BMI distributions at treatment initiation (A), participants with different premorbid BMI distributions (B), participants with increased BMI compared with unchanged or decreased BMI after 2 months of therapy (C), and participants without stunting compared with those with moderate or severe stunting (D). The number of individuals at risk for unfavorable outcomes at different time points is displayed below the survival curves. Abbreviation: BMI, body mass index.
Univariable Analysis
Age, sex, baseline sputum smear, smoking, and alcohol use disorder were significantly associated with unfavorable outcomes (P < .05; Table 3). The association between BMI at treatment initiation and unfavorable outcomes strengthened as the severity of undernutrition increased: moderate (IRR, 1.70; 95% confidence interval [CI], 1.13–2.51) and severe undernutrition (IRR, 2.46; 95% CI, 1.79–3.40) were significantly associated with increased incidence of unfavorable outcomes compared with individuals with a normal BMI. Having mild undernutrition or being overweight at treatment initiation were not associated with unfavorable outcomes. Premorbid nutritional status was not associated with unfavorable outcomes in univariable analysis. Unchanged or decreased BMI 2 months after treatment initiation was associated with an increased incidence of unfavorable outcomes (IRR, 1.81; 95% CI, 1.33–2.48). Severe (IRR, 1.65; 95% CI, 1.13–2.34) but not moderate (IRR, 1.13; 95% CI, .85–1.48) stunting was significantly associated with unfavorable outcomes compared with those without stunting.
Table 3.
Incidence Rate Ratios for Unfavorable Outcomes From Univariable Poisson Regression
| Variable | Incidence Rate Ratio (95% CI)a | P Value |
|---|---|---|
| Treatment initiation BMI, kg/m2 | <.01 | |
| <16 | 2.46 (1.79–3.40)b | |
| 16–16.99 | 1.70 (1.13–2.51)b | |
| 17–18.49 | 1.08 (.69–1.64) | |
| 18.5–22.99 | Ref | |
| >23 | .91 (.56–1.44) | |
| Premorbid BMI, kg/m2 | .07 | |
| <16 | 1.68 (.93–2.86) | |
| 16–16.99 | 1.41 (.77–2.42) | |
| 17–18.49 | .93 (.53–1.57) | |
| 18.5–22.99 | Ref | |
| >23 | .96 (.62–1.45) | |
| Unchanged or decreased BMI | 1.81 (1.33–2.48)b | <.001 |
| Stunting | <.01 | |
| Not stunted (HAZ >−2) | Ref | |
| Moderately stunted ( −3> HAZ ≤−2) | 1.13 (0.85–1.48) | |
| Severely stunted (HAZ ≤3) | 1.65 (1.13–2.34)b | |
| Age, y | <.001 | |
| 18–29 | Ref | |
| 30–39 | 1.49 (1.02–2.18)b | |
| 40–49 | 1.87 (1.33–2.67)b | |
| 50–59 | 1.86 (1.27–2.74)b | |
| 60–69 | 1.36 (.72–2.39) | |
| 70–82 | 3.73 (1.43–8.01)b | |
| Male sex | 2.32 (1.70–3.23)b | <.001 |
| Monthly household income (₹) | .99 | |
| <3000 (<USD 37.5) | Ref | |
| 3001–5000 (USD 37.5–62.5) | .94 (.55–1.66) | |
| 5001–10 000 (USD 62.5–125) | 1.00 (.63–1.69) | |
| >10 000 (>USD 125) | .78 (.48–1.32) | |
| Symptom duration, m | 1.01 (.98–1.03) | .41 |
| Liquid culture time to positivity, wk | .96 (.77–1.18) | .73 |
| Sputum smear at baseline | <.01 | |
| Negative | Ref | |
| 1+ | 1.51 (1.00–2.35) b | |
| 2+ | 1.91 (1.27–2.96) b | |
| 3+ | 1.48 (0.91–2.43) | |
| Scanty | 1.73 (0.77–3.52) | |
| Cavitation | 1.36 (1.01–1.82)b | .04 |
| Pretreatment weight loss, kg | 1.06 (1.03–1.08)b | <.001 |
| Smoking | <.001 | |
| Never | Ref | |
| Former | 1.85 (1.37–2.48)b | |
| Current | 1.82 (1.35–2.43)b | |
| Alcohol use disorderc | 1.62 (1.22–2.13)b | <.001 |
| Diabetesd | .83 (.63–1.08) | .18 |
| Human immunodeficiency virus | 1.45 (.77–2.47) | .21 |
| Antiretroviral therapy status | .31 (.07–1.04) | .08 |
Abbreviations: BMI, body mass index; HAZ, height-for-age z score; USD, United States dollars.
Unfavorable outcome is a composite of death, treatment failure, and relapse within 6 months of treatment completion.
Denotes statistical significance.
Alcohol use disorder defined as having an Alcohol Use Disorders Identification Test-Concise score ≥4 for males and ≥3 for females.
Diabetes defined as known history of diabetes, random blood glucose ≥200 mg/dL, or hemoglobin A1c ≥ 6.5%.
Multivariable Analysis
After adjusting for age, sex, symptom duration, sputum smear grade at baseline, smoking status, alcohol use disorder, diabetes, and HIV infection, severe (aIRR, 2.05; 95% CI, 1.42–2.98) undernutrition at treatment initiation was associated with increased incidence of unfavorable outcomes compared with those with normal BMI (Table 4). Moderate and mild undernutrition and being overweight were not significantly associated with unfavorable outcomes.
Table 4.
Adjusted Incidence Rate Ratios for Unfavorable Outcomes From Multivariable Poisson Regression
| Variable | Adjusted Incidence Rate Ratios for Unfavorable Outcomea (95% Confidence Interval) | ||||
|---|---|---|---|---|---|
| Predictor of interest | Baseline BMI | Premorbid BMI | Unchanged or Decreased BMI | Stunting | |
| Treatment initiation BMI, kg/m2 | <16 | 2.05 (1.42–2.98)b | … | … | … |
| 16–16.99 | 1.50 (0.97–2.31) | … | … | … | |
| 17–18.49 | .99 (0.62–1.54) | … | … | … | |
| 18.5–22.9 | Ref | … | … | … | |
| >23 | .76 (0.42–1.29) | … | … | … | |
| Premorbid BMI, kg/m2 | <16 | … | 2.20 (1.16–3.94)b | … | … |
| 16–16.99 | … | 1.67 (.89–2.95) | … | … | |
| 17–18.49 | … | 1.01 (.56–1.73) | … | … | |
| 18.5–22.9 | … | Ref | … | … | |
| >23 | … | 1.16 (.71–1.85) | … | … | |
| Unchanged or decreased BMI | … | … | … | 1.81 (1.27–2.61)b | … |
| Stunted | No | … | … | … | Ref |
| Moderate | … | … | … | 1.11 (0.82–1.49) | |
| Severe | … | … | … | 1.52 (1.00–2.24)b | |
| Age, y | 18–29 | Ref | Ref | Ref | Ref |
| 30–39 | 1.49 (.96–2.32) | 1.30 (.76–2.24) | 1.22 (.69–2.15) | 1.42 (0.91–2.22) | |
| 40–49 | 1.65 (1.08–2.56)b | 1.35 (.78–2.36) | 1.61 (.92–2.80) | 1.58 (1.03–2.45)b | |
| 50–59 | 1.82 (1.14–2.91)b | 1.64 (.90–2.98) | 2.41 (1.36–4.26)b | 1.73 (1.08–2.79)b | |
| 60–69 | 1.30 (.65–2.44 | .89 (.32–2.10) | .98 (.33–2.91) | 1.20 (0.60–2.28) | |
| 70–82 | 3.56 (1.20–8.55)b | 5.44 (1.25–16.53)b | .00 (0–∞) | 3.13 (1.05–7.50)b | |
| Male sex | … | 1.58 (1.06–2.40)b | 1.56 (.92–2.70)b | 1.56 (.94–2.61) | 1.58 (1.05–2.39)b |
| Symptom duration, m | … | 1.02 (.99–1.05) | 1.03 (.99–1.06) | 1.04 (1.01–1.08)b | 1.02 (0.99–1.05) |
| Sputum at baseline | Negative | Ref | Ref | Ref | Ref |
| 1+ | 1.24 (.76–2.09) | 1.30 (.73–2.39) | 1.19 (.67–2.09) | 1.29 (0.79–2.17) | |
| 2+ | 1.42 (.87–2.40) | 1.70 (.95–3.15) | 1.54 (.87–2.71) | 1.44 (0.88–2.44) | |
| 3+ | 1.00 (.56–1.78) | 1.06 (.53–2.11) | 1.54 (.75–3.18) | 1.02 (0.58–1.82) | |
| Scanty | 1.46 (.61–3.12) | 1.55 (.52–3.81) | 1.69 (.68–4.23) | 1.49 (0.62–3.18) | |
| Smoking | Never | Ref | Ref | Ref | Ref |
| Former | 1.21 (.84–1.74) | 1.73 (1.07–2.78)b | 1.52 (.93–2.50) | 1.24 (0.86–1.78) | |
| Current | 1.16 (.80–1.65) | 1.51 (.90–2.51) | 1.52 (.94–2.46) | 1.26 (0.88–1.81) | |
| Alcohol use disorderc | … | 1.13 (.81− 1.55) | 1.23 (.90–2.51) | 1.11 (.74–1.66) | 1.18 (0.82–1.69) |
| Diabetesd | … | .84 (0.60–1.17) | .76 (.47–1.21) | .59 (.40–.89)b | 0.87 (0.62–1.21) |
| Human immunodeficiency virus | … | 2.02 (.93–3.97) | 2.58 (1.15–5.36)b | 1.02 (.42–2.48) | 1.83 (0.85–3.62) |
Abbreviation: BMI, body mass index.
Unfavorable outcome is a composite of death, treatment failure, and relapse within 6 months of treatment completion.
Denotes statistical significance.
Alcohol use disorder defined as having an Alcohol Use Disorders Identification Test-Concise score ≥4 for males and ≥3 for females.
Diabetes defined as known history of diabetes, random blood glucose ≥200 mg/dL, or hemoglobin A1c ≥ 6.5%.
After adjusting for confounders, severe undernutrition prior to TB disease symptoms was associated with unfavorable outcomes (aIRR, 2.20; 95% CI, 1.16–3.94). Unchanged or decreased BMI after 2 months of treatment was associated with an increased incidence of unfavorable outcomes (aIRR, 1.81; 95% CI, 1.27–2.61) compared with those who had an increase in BMI. Severe stunting was also associated with unfavorable outcomes compared with those without stunting (aIRR, 1.52; 95% CI, 1.00–2.24; Table 4).
We did not find evidence of interactions between cough duration, smoking status, alcohol use disorder, or diabetes and the impact of any nutritional predictor on unfavorable outcomes (Supplementary Table 1). However, there was interaction between BMI at treatment initiation and sex (P = .02). Upon stratifying the analysis by sex, severe undernutrition remained a significant predictor of unfavorable outcomes (aIRR, 2.34; 95% CI, 1.57–3.54) among males but lost significance (aIRR, 0.95; 95% CI, .32–2.78) among females (Supplementary Table 2).
Sensitivity Analysis
Alternate Outcome Variables
Table 5 presents results of the sensitivity analysis of alternate outcome variables. Severe undernutrition at treatment initiation was associated with death (aIRR, 4.17; 95% CI, 1.97–9.53) but was not significantly associated with failure and recurrence (aIRR, 1.51; 95% CI, .96–2.40). Both severe and moderate undernutrition at treatment initiation were associated with increased LTFU. Severe premorbid undernutrition was not significantly associated with increased failure or relapse, death, or LTFU. Unchanged or decreased BMI after 2 months of treatment was associated with increased failure or recurrence (aIRR, 1.72; 95% CI, 1.09–2.72), death (aIRR, 5.16; 95% CI, 1.51–17.65), and LTFU (aIRR, 6.24; 95% CI, 3.78–10.87). Severe stunting was not associated with any alternative outcome. Moderate stunting was not associated with increased rates of failure or relapse or death but appeared to be associated with a lower rate of LTFU (aIRR, 0.56; 95% CI, .41–.77).
Table 5.
Adjusted Incidence Rate Ratios for Alternative Treatment Outcome Variables
| Variable | Unfavorable Outcome,a,b aIRR (95%CI) | Failure or Relapse aIRR(95%CI)c | Death aIRR (95%CI)d | Loss to Follow-up aIRR (95%CI)e |
|---|---|---|---|---|
| Treatment initiation BMI, kg/m2 | ||||
| <16 | 2.05 (1.42–2.98)f | 1.51 (.96–2.40) | 4.17 (1.97–9.53)f | 1.81 (1.46–2.25)f |
| 16–16.99 | 1.50 (.97–2.31) | 1.38 (.82–2.32) | 1.23 (.43–3.36) | 1.37 (1.05–1.78)f |
| 17–18.49 | .99 (.62–1.54) | 1.09 (.64–1.86) | .53 (.12–1.78) | 1.14 (.86–1.48) |
| 18.5–22.9 | Ref | Ref | Ref | Ref |
| >23 | .76 (.42–1.29) | .71 (.35–1.43) | .82 (.18–2.77) | 1.02 (0.76–1.35) |
| Premorbid BMI | ||||
| <16 | 2.20 (1.16–3.94)f | 2.03 (.96–4.31) | 1.18 (.06–6.41) | 0.76 (0.28–1.69) |
| 16–16.99 | 1.67 (.89–2.95) | 1.15 (.52–2.53) | 2.62 (.81–7.49) | 1.19 (0.54–2.35) |
| 17–18.49 | 1.01 (.56–1.73) | .71 (.32–1.54) | 2.14 (.77–5.58) | 1.36 (.82–2.20) |
| 18.5–22.9 | Ref | Ref | Ref | Ref |
| >23 | 1.16 (.71–1.85) | .91 (.49–1.68) | 1.70 (.70–4.12) | 1.65 (1.09–2.50)f |
| Unchanged or decreased BMI | 1.81 (1.27–2.61)f | 1.72 (1.09–2.72)f | 5.16 (1.51–17.65)f | 6.24 (3.78–10.87)f |
| Stunted | ||||
| Not stunted | Ref | Ref | Ref | Ref |
| Moderate stunting | 1.11 (.82–1.49) | 1.28 (.88–1.87) | 1.08 (.56–1.98) | .56 (.41–0.77)f |
| Severe stunting | 1.52 (1.00–2.24)f | 1.62 (.95–2.80) | 1.30 (.48–2.97) | .82 (.52–1.25) |
Abbreviations: aIRR, adjusted incidence rate ratio; BMI, body mass index; CI, confidence interval.
Unfavorable outcome is a composite of death, treatment failure, and relapse within 6 months of treatment completion;
Adjusted for age, sex, symptom duration, sputum smear grade at baseline smoking status, alcohol use disorder, diabetes, and human immunodeficiency virus (HIV).
Adjusted for age, sex, income, symptom duration, sputum smear grade at baseline, smoking status, alcohol use disorder.
Adjusted for age, sex, income, symptom duration, time to positivity of mycobacteria growth indicator tube, smoking status, diabetes, HIV.
Adjusted for age, sex, income, symptom duration, sputum smear grade at baseline, smoking status, alcohol use disorder, and HIV.
Denotes statistical significance.
Accounting for Nutritional Status at Treatment Initiation
Unchanged or decreased BMI after 2 months of therapy was a significant predictor of unfavorable outcomes among both undernourished participants (aIRR, 1.67; 95% CI, 1.05–2.65) and participants without undernutrition (aIRR, 2.64; 95% CI, 1.32–5.28).
Impact of Cavitation on Chest Radiograph
When we included cavitation in our multivariable model, the effect of severe undernutrition at treatment initiation on unfavorable outcomes remained significant (aIRR, 1.69; 95% CI, 1.09–2.63). After including cavitation in the multivariable model, severe premorbid undernutrition was not significantly associated with unfavorable outcomes (aIRR, 1.77; 95% CI, .83–3.79). Of note, only 322 participants had data for both cavitation and BMI prior to symptoms. Adjusting for cavitation did not affect the incidence of unfavorable outcomes associated with unchanged or decreased BMI (aIRR, 1.78; 95% CI, 1.22–2.63). The risk associated with moderate (aIRR, 1.19; 95% CI, .84–1.69) and severe (aIRR, 1.35; 95% CI, .82–2.24) stunting also did not change substantially.
Use of WHO Cutoffs
When we used WHO cutoffs for normal BMI, both severe (aIRR, 2.13; 95% CI, 1.48–3.06) and moderate (aIRR, 1.56; 95% CI, 1.01–2.37) undernutrition at treatment initiation were associated with unfavorable outcomes. Severe premorbid undernutrition (aIRR, 2.18; 95% CI, 1.16–3.88) remained significantly associated with unfavorable outcomes (Supplementary Table 3).
DISCUSSION
In this multicenter prospective cohort analysis, we show that undernutrition and stunting are highly prevalent among PWTB in India. We found that undernutrition at treatment initiation, undernutrition prior to TB symptom onset, unchanged or decreased BMI during the intensive phase of therapy, and severe stunting were associated with unfavorable outcomes. By analyzing the impact of varying degrees of undernutrition, this study provides greater resolution on the subgroups of undernourished individuals who are at a particularly high risk. Notably, severely undernourished participants had a 4-fold increased incidence rate of death and those without BMI increase after 2 months of therapy had a 5-fold increase in death. To our knowledge, this is the largest prospective study of the effect of nutritional status on treatment outcomes.
Based on our findings, clinicians should systematically assess the nutritional status of PWTB using anthropometry at treatment initiation to gauge risk of poor outcomes and consider screening for reversible causes of undernutrition such as intestinal parasitic infections. The increased unfavorable outcomes among those without BMI increase at 2 months may have been due to suboptimal TB therapy or poor nutritional intake. Given its prognostic value, anthropometry at 2 months should also be standard practice for TB follow-up. PWTB with stagnant or reduced BMI should be assessed for causes of treatment failure such as nonadherence or drug resistance. Those who are severely undernourished at treatment initiation or fail to gain weight during therapy should also be offered nutritional counseling and support.
The WHO recommended the integration of nutritional assessment and care into standard TB treatment in 2013, but implementation has been variable [28]. In April 2018, the Nikshay Poshan Yojana began providing $6.50 monthly cash transfer to PWTB to improve nutritional intake [29]. Our study had recruited more than 87% of its participants prior to this. We did not have data on the receipt of this subsidy by our participants, so we could not assess its effectiveness in our cohort. A contemporaneous study reported that less than one-quarter of PWTB received cash transfers [30]. Further study is needed to determine the value of enhanced subsidies for severely undernourished PWTB and whether cash subsidies or provision of food results in improved outcomes. Notably, 2 Indian studies have shown that providing inexpensive food baskets during therapy can improve treatment outcomes [31, 32].
Importantly, those with severe undernutrition prior to TB disease or severe stunting, a consequence of early childhood malnutrition, were also at increased risk of unfavorable outcomes. Since weight loss may correlate with severity of TB and a longer duration of untreated TB, the direction of causality has been frequently questioned in observational studies that showed that undernutrition at treatment initiation is associated with increased risk of unfavorable outcomes. By showing the impact of undernutrition prior to TB disease on treatment outcomes, our study helps answer this longstanding “chicken-and-egg” question. Further, it supports the view that reducing undernutrition at a population scale in high TB burden countries may help decrease unfavorable TB outcomes. Mathematical models have suggested that doing so would reduce TB incidence and mortality while being highly cost-effective [33, 34].
This study has numerous strengths. Our multicenter design makes our sample more representative of India's diverse population. Given the large study size and high rate of poor treatment outcomes, the study was well powered. The use of finer BMI cutoffs allowed us to perform better risk stratification of PWTB. We adjusted for symptom duration, a confounder that has not always been included in prior studies of undernutrition's impact on TB. Last, our analysis of BMI prior to the onset of symptoms provides fresh insights on the effect of premorbid nutritional status on TB outcomes.
Our analysis does have limitations. First, there may have been unmeasured confounders. We did not adjust for adherence as we did not have access to objective adherence metrics (eg, urine or blood drug levels). Further, self-reported income may not sufficiently account for socioeconomic determinants. Also, as we did not quantitate smoking status, we may not have fully captured the impact of tobacco use on treatment outcomes. We calculated premorbid BMI using self-reported weight loss, which may have been subject to recall bias, and only a subset of individuals had data on premorbid weight. Additionally, prior to standardization in the consortium, the definition of diabetes was variable across sites, some using hemoglobin A1c and others used blood glucose. This may have limited our ability to examine interactions between undernutrition and diabetes that have been previously reported in India [35].
In conclusion, undernutrition at treatment initiation, undernutrition prior to onset of TB disease, severe stunting and weight gain during therapy are independent predictors of treatment outcomes with a sizable effect. Our study highlights the need for recognizing and treating undernutrition during TB care. More research is needed on the best methods of providing nutritional support to improve treatment outcomes for undernourished PWTB.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Supplementary Material
Contributor Information
Pranay Sinha, Section of Infectious Diseases, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
Chinnaiyan Ponnuraja, Indian Council of Medical Research, National Institute for Research in Tuberculosis, Chennai, Tamil Nadu, India.
Nikhil Gupte, Byramjee Jeejeebhoy Government Medical College, Sassoon General Hospitals–Johns Hopkins University Clinical Research Site, Pune, Maharashtra, India; Center for Infectious Diseases in India, Johns Hopkins India, Pune, Maharashtra, India; Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Senbagavalli Prakash Babu, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India.
Samyra R Cox, Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Sonali Sarkar, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India.
Vidya Mave, Byramjee Jeejeebhoy Government Medical College, Sassoon General Hospitals–Johns Hopkins University Clinical Research Site, Pune, Maharashtra, India; Center for Infectious Diseases in India, Johns Hopkins India, Pune, Maharashtra, India; Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Mandar Paradkar, Byramjee Jeejeebhoy Government Medical College, Sassoon General Hospitals–Johns Hopkins University Clinical Research Site, Pune, Maharashtra, India; Center for Infectious Diseases in India, Johns Hopkins India, Pune, Maharashtra, India.
Chelsie Cintron, Section of Infectious Diseases, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA.
S Govindarajan, Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry, India; National Tuberculosis Elimination Program, Puducherry, India.
Aarti Kinikar, Byramjee Jeejeebhoy Government Medical College and Sassoon General Hospitals, Pune, Maharashtra, India.
Nadesan Priya, Christian Medical College, Vellore, Tamil Nadu, India.
Sanjay Gaikwad, Byramjee Jeejeebhoy Government Medical College and Sassoon General Hospitals, Pune, Maharashtra, India.
Balamugesh Thangakunam, Christian Medical College, Vellore, Tamil Nadu, India.
Arutselvi Devarajan, Prof. M. Viswanathan Diabetes Research Centre, Chennai, Tamil Nadu, India.
Mythili Dhanasekaran, Prof. M. Viswanathan Diabetes Research Centre, Chennai, Tamil Nadu, India.
Jeffrey A Tornheim, Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Amita Gupta, Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Padmini Salgame, Center for Emerging Pathogens, Department of Medicine, New Jersey Medical School, Rutgers Biomedical and Health Sciences, Newark, New Jersey, USA.
Devashyam Jesudas Christopher, Christian Medical College, Vellore, Tamil Nadu, India.
Hardy Kornfeld, Department of Medicine, University of Massachusetts Chan Medical School, Worcester, Massachusetts, USA.
Vijay Viswanathan, Prof. M. Viswanathan Diabetes Research Centre, Chennai, Tamil Nadu, India.
Jerrold J Ellner, Center for Emerging Pathogens, Department of Medicine, New Jersey Medical School, Rutgers Biomedical and Health Sciences, Newark, New Jersey, USA.
C Robert Horsburgh, Jr, Section of Infectious Diseases, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA; Department of Epidemiology, Boston University School of Public Health, Boston, Massachusetts, USA; Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts, USA; Department of Global Health, Boston University School of Public Health, Boston, Massachusetts, USA.
Akshay N Gupte, Division of Infectious Diseases, Johns Hopkins University, School of Medicine, Baltimore, Maryland, USA.
Chandrasekaran Padmapriyadarsini, Indian Council of Medical Research, National Institute for Research in Tuberculosis, Chennai, Tamil Nadu, India.
Natasha S Hochberg, Section of Infectious Diseases, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, USA; Department of Epidemiology, Boston University School of Public Health, Boston, Massachusetts, USA.
Notes
Disclaimer . The contents presented here are solely the responsibility of the authors and do not represent the official views of the Department of Biotechnology (DBT), the Indian Council of Medical Research (ICMR), the US National Institutes of Health (NIH), or CRDF Global. Any mention of trade names, commercial projects, or organizations does not imply endorsement by any of the sponsoring organizations.
Financial support . Data presented here were collected as part of the Regional Prospective Observational Research for Tuberculosis India Consortium. This project was funded in whole or in part with federal funds from the government of India's DBT, the ICMR, the NIH, National Institute of Allergy and Infectious Diseases (NIAID), Office of AIDS Research and distributed in part by CRDF Global. In addition, the following funding sources have supported this work: P. S. is supported by NIAID (award 5T32 AI-052074-13) and the Burrough's Wellcome Fund/American Society of Tropical Medicine and Hygiene fellowship; J. A. T. is supported by NIAID (award K23AI135102); A. G., N. G., V. M., A. N. G., M. P., and S. R. C. are supported by NIAID via CRDF Global (awards DAA3-19-65671-1 and DAA3-18-64774-1) and by NIAID (award R01A1I097494); A. N. G. is supported by NIAID (award K99AI151094); and V. V. and H. K. are supported by the Indian DBT (grant USB1-31149-XX-13). A. K. also reports support for this work in the form of NIH funding for Johns Hopkins Baltimore–Washington–India Clinical Trials Unit from NIAID Networks and CRDF Global. A. G. also reports support for this work in the form of funding paid to institution from NIH, Unitaid, and the Centers for Disease Control and Prevention, outside the submitted work. M. P., S. Ga., and V. M. report support for this work in the form of NIH funding for Johns Hopkins Baltimore–Washington–India Clinical Trials Unit from NIAID Networks and CRDF Global, payment to institution.
References
- 1. World Health Organization . Global tuberculosis report 2022. Geneva, Switzerland: WHO, 2022. [Google Scholar]
- 2. Sinha P, Davis J, Saag L, et al. Undernutrition and tuberculosis: public health implications. J Infect Dis 2019; 219:1356–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hochberg NS, Sarkar S, HorsburghCR, Jr., et al. Comorbidities in pulmonary tuberculosis cases in Puducherry and Tamil Nadu, India: opportunities for intervention. PLoS One 2017; 12: e0183195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. FAO, IFAD, UNICEF, WFP, and WHO . The state of food security and nutrition in the world 2021. Transforming food systems for food security, improved nutrition and affordable healthy diets for all. Rome, Italy: FAO, 2021. [Google Scholar]
- 5. Sinha P, Carwile M, Cintron C, de Perez EC, Hochberg N. Climate change and TB: the soil and seed conceptual framework. Public Health Action 2021; 11:108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Van Lettow M, Kumwenda J, Harries A, et al. Malnutrition and the severity of lung disease in adults with pulmonary tuberculosis in Malawi. Int J Tuberc Lung Dis 2004; 8:211–7. [PubMed] [Google Scholar]
- 7. Hoyt KJ, Sarkar S, White L, et al. Effect of malnutrition on radiographic findings and mycobacterial burden in pulmonary tuberculosis. PLoS One 2019; 14:e0214011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Podewils L, Holtz T, Riekstina V, et al. Impact of malnutrition on clinical presentation, clinical course, and mortality in MDR-TB patients. Epidemiol Infect 2011; 139:113–20. [DOI] [PubMed] [Google Scholar]
- 9. Beek LT, Alffenaar J-WC, Bolhuis MS, van der Werf TS, Akkerman OW. Tuberculosis-related malnutrition: public health implications. J Infect Dis 2019; 220:340–1. [DOI] [PubMed] [Google Scholar]
- 10. Ramachandran G, Hemanth Kumar A, Bhavani P, et al. Age, nutritional status and INH acetylator status affect pharmacokinetics of anti-tuberculosis drugs in children. Int J Tuberc Lung Dis 2013; 17:800–6. [DOI] [PubMed] [Google Scholar]
- 11. Putri FA, Burhan E, Nawas A, et al. Body mass index predictive of sputum culture conversion among MDR-TB patients in Indonesia. Int J Tuberc Lung Dis 2014; 18:564–70. [DOI] [PubMed] [Google Scholar]
- 12. Birlie A, Tesfaw G, Dejene T, Woldemichael K. Time to death and associated factors among tuberculosis patients in Dangila Woreda, northwest Ethiopia. PLoS One 2015; 10:e0144244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Bhargava A, Chatterjee M, Jain Y, et al. Nutritional status of adult patients with pulmonary tuberculosis in rural central India and its association with mortality. PLoS One 2013; 8:e77979. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Khan A, Sterling TR, Reves R, Vernon A, Horsburgh CR, Consortium TT. Lack of weight gain and relapse risk in a large tuberculosis treatment trial. Am J Respir Crit Care Med 2006; 174:344–8. [DOI] [PubMed] [Google Scholar]
- 15. Peetluk LS, Rebeiro PF, Cordeiro-Santos M, et al. Lack of weight gain during the first 2 months of treatment and human immunodeficiency virus independently predict unsuccessful treatment outcomes in tuberculosis. J Infect Dis 2020; 221:1416–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Santha T, Garg R, Frieden T, et al. Risk factors associated with default, failure and death among tuberculosis patients treated in a DOTS programme in Tiruvallur District, south India, 2000. Int J Tuberc Lung Dis 2002; 6:780–8. [PubMed] [Google Scholar]
- 17. Sinha P, White L, Hochberg N, Cegielski J. Avoiding pitfalls in calculating the population attributable fraction of undernutrition for TB. Int J Tuberc Lung Dis 2022; 26:80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Maciel ELN, Golub JE, Peres RL, et al. Delay in diagnosis of pulmonary tuberculosis at a primary health clinic in Vitoria, Brazil. Int J Tuberc Lung Dis 2010; 14:1403–10. [PMC free article] [PubMed] [Google Scholar]
- 19. Rytter MJH, Kolte L, Briend A, Friis H, Christensen VB. The immune system in children with malnutrition—a systematic review. PLoS One 2014; 9:e105017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Gupte A, Padmapriyadarsini C, Mave V, et al. Cohort for tuberculosis research by the Indo-US Medical Partnership (CTRIUMPH): protocol for a multicentric prospective observational study. BMJ Open 2016; 6:e010542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Hamilton CD, Swaminathan S, Christopher DJ, et al. RePORT International: advancing tuberculosis biomarker research through global collaboration. Clin Infect Dis 2015; 61(Suppl 3):S155–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. James W, Ferro-Luzzi A, Waterlow JC. Definition of chronic energy deficiency in adults. Report of a working party of the International Dietary Energy Consultative Group. Eur J Clin Nutr 1988; 42:969–81. [PubMed] [Google Scholar]
- 23. Misra A, Chowbey P, Makkar B, et al. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian Indians and recommendations for physical activity, medical and surgical management. J Assoc Physicians India 2009; 57:163–70. [PubMed] [Google Scholar]
- 24. World Health Organization . WHO child growth standards: training course on child growth assessment. Geneva, Switzerland: WHO, 2008.
- 25. Bradley KA, DeBenedetti AF, Volk RJ, Williams EC, Frank D, Kivlahan DR. AUDIT-C as a brief screen for alcohol use disorder in primary care. Alcohol Clin Exp Res 2007; 31:1208–17. [DOI] [PubMed] [Google Scholar]
- 26. Zhou TJ, Lakshminarayanan S, Sarkar S, et al. Predictors of loss to follow-up among men with tuberculosis in Puducherry and Tamil Nadu, India. Am J Trop Med Hyg 2020; 103:1050–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. World Health Organization . Cut-off for BMI according to WHO standards. Available at: https://gateway.euro.who.int/en/indicators/mn_survey_19-cut-off-for-bmi-according-to-who-standards/. Accessed 25 October 2022.
- 28. World Health Organization . Guideline: nutritional care and support for patients with tuberculosis. Geneva, Switzerland: WHO, 2013.
- 29. Patel BH, Jeyashree K, Chinnakali P, et al. Cash transfer scheme for people with tuberculosis treated by the National TB Programme in western India: a mixed methods study. BMJ Open 2019; 9:e033158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Nirgude AS, Kumar AMV, Collins T, et al. ‘" am on treatment since 5 months but I have not received any money": coverage, delays and implementation challenges of "Direct Benefit Transfer" for tuberculosis patients—a mixed-methods study from South India. Global Health Action 2019; 12:1633725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Samuel B, Volkmann T, Cornelius S, et al. Relationship between nutritional support and tuberculosis treatment outcomes in West Bengal, India. J Tuberc Res 2016; 4:213–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Singh AK, Siddhanta A, Goswami L. Improving tuberculosis treatment success rate through nutrition supplements and counselling: findings from a pilot intervention in India. Clin Epidemiol Global Health 2021; 11:100782. [Google Scholar]
- 33. Oxlade O, Huang C-C, Murray M. Estimating the impact of reducing under-nutrition on the tuberculosis epidemic in the central eastern states of India: a dynamic modeling study. PLoS One 2015; 10:e0128187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Sinha P, Lakshminarayanan SL, Cintron C, et al. Nutritional supplementation would be cost-effective for reducing tuberculosis incidence and mortality in India: the Ration Optimization to Impede Tuberculosis (ROTI-TB) model. Clin Infect Dis 2021; 75:577–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Kornfeld H, Sahukar SB, Procter-Gray E, et al. Impact of diabetes and low body mass index on tuberculosis treatment outcomes. Clin Infect Dis 2020; 71:e392–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.