Intervention Review

You have free access to this content

Nutritional supplements for people being treated for active tuberculosis

  1. David Sinclair1,*,
  2. Katharine Abba1,
  3. Liesl Grobler2,
  4. Thambu D Sudarsanam3

Editorial Group: Cochrane Infectious Diseases Group

Published Online: 9 NOV 2011

Assessed as up-to-date: 1 JUL 2011

DOI: 10.1002/14651858.CD006086.pub3


How to Cite

Sinclair D, Abba K, Grobler L, Sudarsanam TD. Nutritional supplements for people being treated for active tuberculosis. Cochrane Database of Systematic Reviews 2011, Issue 11. Art. No.: CD006086. DOI: 10.1002/14651858.CD006086.pub3.

Author Information

  1. 1

    Liverpool School of Tropical Medicine, International Health Group, Liverpool, UK

  2. 2

    Cape Town, Western Province, South Africa

  3. 3

    Christian Medical College, Medicine Unit 2, Vellore, Tamil Nadu, India

*David Sinclair, International Health Group, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool, L3 5QA, UK. sinclad@liverpool.ac.uk.

Publication History

  1. Publication Status: New search for studies and content updated (no change to conclusions)
  2. Published Online: 9 NOV 2011

SEARCH

 

Summary of findings    [Explanations]

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

 
Summary of findings for the main comparison.

Food provision compared with nutritional advice or no intervention for patients with active tuberculosis

Patient or population: adults and children with active tuberculosis

Settings: low- and middle-income countries

Intervention: calorie supplementation as food or energy dense supplements

Comparison: nutritional advice or no intervention

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
(studies)
Quality of the evidence
(GRADE)
Comments

Assumed riskCorresponding risk

Standard careIncreased calorie intake

Death

(at 6 months)
49 per 100020 per 1000
(3 to 110)
RR 0.4
(0.07 to 2.25)
202
(2 studies)
⊕⊝⊝⊝
very low1,2,3
We don't know if food supplementation reduces mortality from tuberculosis in food insecure settings

Cured

(at 6 months)
481 per 1000438 per 1000
(284 to 678)
RR 0.91
(0.59 to 1.41)
102
(1 study)
⊕⊝⊝⊝
very low2,3,4
We don't know if food supplementation increases cure in tuberculosis patients

Treatment completion
(at 6 months)
788 per 1000851 per 1000
(693 to 1000)
RR 1.08
(0.88 to 1.33)
365
(2 studies)
⊕⊝⊝⊝
very low1,3,5
We don't know if food supplementation increases treatment completion in tuberculosis patients

Sputum negative

(at 8 weeks)
644 per 1000773 per 1000
(657 to 901)
RR 1.2
(1.02 to 1.4)
149
(2 studies)
⊕⊝⊝⊝
very low3,5
We don't know if food supplementation reduces the duration of sputum positivity in tuberculosis patients

Mean weight gain

(At 8 weeks)
---731
(4 studies
moderate6,7Supplementation probably increases weight gain during treatment

Quality of life
(At 8 weeks)
---134
(2 studies)
low8,9Supplementation may increase quality of life scores during the first 2 months of treatment

*The assumed risk is taken from the risk in the control groups in the included studies. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk Ratio;

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

 1Two trials report deaths during the 6 months of treatment (Sudarsanam 2010 and Jahnavi 2010). These trials were both conducted in India in participants with signs of undernutrition. Sudarsanam gave monthly ration packs and Jahnavi gave daily locally appropriate supplements. One larger trial giving a hot daily meal to significantly undernourished people with tuberculosis reports no deaths occurred in either group after treatment started (Martins 2009).
2 Downgraded by 1 under indirectness as trials are only available from India and not from food-insecure settings where food supplementation may plausibly have its biggest effect.
3 Downgraded by 2 under precision as the trials are significantly underpowered to either detect or exclude an effect if it exists.
4 Data on successful cure at 6 months is only available from Sudarsanam 2010.
5 Downgraded by 1 for inconsistency. Jahnavi 2010 found a statistically significant benefit while the larger trial, Martins 2009 did not.
6 Four studies report measures of weight gain but at different time-points which prevented meta-analysis.
7 Downgraded by 1 for inconsistency. Praygod 2011 included only HIVpositive patients and although the trend was towards a benefit this did not reach statistical significance. The three other trials all demonstrated clinically important benefits.
8 Downgraded by 1 for indirectness. Only 2 small trials, one from Singapore (Paton 2004) and one from India (Jahnavi 2010) report quality of life scores. The results can not be generalised to other populations or settings with any certainty.
9 Downgraded by 1 for imprecision. The presented data appear highly skewed and could not be pooled.

 Summary of findings 2

 

Background

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Description of the condition

Tuberculosis is an infection caused by the bacterium Mycobacterium tuberculosis, which is spread from person to person by inhalation of respiratory droplets (Harries 2006). In 2009, the World Health Organization (WHO) estimated that there were 9.4 million new cases of active tuberculosis worldwide, and 1.3 million deaths (WHO 2010).

The vast majority of people infected with M. tuberculosis develop what is known as latent tuberculosis, where the infection is contained by the persons immune system, and they remain well (Barry 2009). The immune response to infection is complex, initially involving the uptake of the bacterium into macrophage cells, part of the non-specific 'innate' immune response, and with later recruitment of both B- and T-lymphocytes of the cellular immune response (Schluger 1998). These cells isolate the bacterium as a granuloma, typically in the lung (Saunders 2007).

Active tuberculosis occurs when the infection is no longer contained by the immune system, and can occur at any time following infection. The life-time risk of conversion from latent to active tuberculosis is around 5% to 10% in an otherwise healthy population (Harries 2006), but this can rise to around 50% in people with severe impairment of their immune system, such as occurs with human immunodeficiency virus (HIV) infection (Zumla 2000; Aaron 2004).

Tuberculosis most commonly affects the lungs (pulmonary tuberculosis), but can also spread to affect the central nervous system, lymphatic system, circulatory system, genitourinary system, and bones and joints. The symptoms of active pulmonary tuberculosis include cough, chest pain, fever, night sweats, weight loss, and sometimes coughing up blood (Harries 2006).

Tuberculosis is treated with a combination of antibiotic drugs (antituberculous therapy), which must be taken for a period of at least six months to ensure success (WHO 2010b). If left untreated, around half of those with active tuberculosis will die of the disease (Corbett 2003). With adequate treatment the mortality is around 5% globally (WHO 2009), although this may be higher in patients with HIV co-infection (Aaron 2004). The WHO target for successful cure in national tuberculosis control programmes is 85% (WHO 2009).

Throughout the world, poor nutritional status is more common in people with active tuberculosis than in people without tuberculosis (van Lettow 2003), and weight loss, including loss of lean body mass, is a well-recognized symptom of the disease. Cohort and cross-sectional studies have suggested that active tuberculosis is commonly associated with low serum levels of important micronutrients such as zinc (Taneja 1990), and vitamins A, C, D, and E (Davies 1985; Plit 1998; Nnoaham 2008). However, the measurement of serum vitamin levels during an acute infection, such as tuberculosis, is known to be unreliable, as transient abnormalities can occur (Louw 1992).

The daily nutritional requirements for healthy individuals of all age groups are well described (Meyers 2006), and it is unlikely that patients with active tuberculosis would require less than these recommended amounts. The two important questions are therefore whether tuberculosis patients require more, and whether these increased requirements should be provided as part of routine health care.

The effects of supplementation in people with HIV (but without tuberculosis) is covered in two other Cochrane reviews: Irlam 2010 and Mahlungulu 2007.

 

Description of the intervention

Nutritional requirements can be broadly divided into macronutrients (carbohydrate, protein and fat), and micronutrients (essential vitamins and trace elements).

 
Macronutrients

Each day the average 70 kg male requires approximately 2500 kilocalories (kcal) of energy to maintain body weight and composition; ideally consumed as 55% carbohydrate, 15% protein and 30% fat (Meyers 2006).

If it was shown that patients with active tuberculosis required additional macronutrients; these could be purchased and consumed by the patient simply following nutritional advice. In many situations however, especially in low- and middle-income countries, the patient may not be able to acquire this additional food due to economic hardship through illness and loss of work, or due to local food insecurity (Kamolratanakul 1999; Wyss 2001). In these situations healthcare services might provide increased nutrients through free provision of meals, take home rations, or specific high energy supplements. In many crisis or low-income settings this already happens and the World Food Programme (WFP) in particular is involved in many food support programmes for tuberculosis patients (WFP 2007).

 
Micronutrients

The daily micronutrient requirements for an adult male are given in Table 5. These are usually expressed as the 'dietary reference intake' (DRI), and this is different for each individual micronutrient (Meyers 2006).

These requirements can be gained from the consumption of a healthy, and varied diet, or through pharmaceutical supplementation as tablets, capsules or powders. Any additional requirements could be gained through increased consumption following dietary advice, or through pharmaceutical provision via the health service or tuberculosis programme.

In trials of macronutrient and micronutrient interventions two important factors should be noted:

  • the intervention is a supplement and does not represent the total daily intake of that nutrient;
  • any benefit derived from the intervention is likely to be dependant on the initial nutritional status of the patient.

In order to accurately interpret data it is therefore essential to consider both the baseline nutritional status, and the overall nutritional intake of the patients.

 

How the intervention might work

Tuberculosis and undernutrition interact in a two-way process. Tuberculosis can lead to weight loss and micronutrient deficiencies by increasing nutritional requirements, by changing metabolic processes, or by decreasing appetite and causing a reduction in food intake (Macallan 1999). Alternatively, low body mass index (BMI; a measure of weight for height that is indicative of nutritional status) and some micronutrient deficiencies can depress cell-mediated immunity, the key host defence against tuberculosis, increasing the susceptibility to active tuberculosis and delaying recovery (Chandra 1996; Zachariah 2002; Cegielski 2004).

The micronutrients which have received the most attention are:

  • Vitamin A, which is involved in both T- and B-lymphocyte function, macrophage activity and the generation of antibody responses (Semba 1998; Stephenson 2001);
  • Vitamin D, which is involved in the function of macrophages, a key component of the immune response to tuberculosis (Wintergerst 2007);
  • Vitamin E, which has anti-oxidant properties and may protect against T-lympocyte failure due to oxidative stress (Wintergerst 2007);
  • Zinc, which is necessary for adequate functioning of many aspects of human immunity (Shankar 1998);
  • Selenium; which is essential for both cell-mediated and humoral immunity (Arthur 2003).

Nutritional interventions, in people with active tuberculosis, therefore have the potential to:

  1. Improve tuberculosis treatment outcomes; through restoration of cell-mediated immunity, increasing the individuals' ability to fight the infection and hastening recovery from the illness.
  2. Promote nutritional recovery; with improved weight gain, restoration of muscle strength, function, and quality of life. Nutritional recovery is of great importance in tuberculosis treatment, allowing the patient to return to work, and recover economically as well as physically.

Food may also be given to people with tuberculosis for quite different reasons, such as; to promote adherence to treatment, or to mitigate the financial consequences of prolonged illness. The use of food to promote adherence is being addressed in another Cochrane Review (Lutge 2009).

In addition, it is important to note that pathogens such as tuberculosis also require certain micronutrients for their own metabolism, and greater availability of these nutrients through supplementation could encourage their growth. There is some evidence for this in the case of iron (Lounis 2001), and so nutritional interventions cannot be considered entirely benign.

 

Why it is important to do this review

There is currently no evidence-based guidance on food provision or supplementation for adults or children being treated for tuberculosis, with or without concurrent HIV infection. This review seeks to assess the evidence for the effectiveness of different food and nutritional supplements in helping people to gain weight and recover from tuberculosis, and highlight where more research might be needed.

 

Objectives

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

To assess the effects of oral nutritional supplements in people being treated with antituberculous drug therapy for active tuberculosis.

 

Methods

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Criteria for considering studies for this review

 

Types of studies

Randomized controlled trials (RCTs).

 

Types of participants

Children or adults being treated for active tuberculosis with or without concurrent HIV infection, and with or without a diagnosis of being underweight, malnourished, or nutrient deficient.

 

Types of interventions

 

Intervention

Any oral nutritional supplement given for at least four weeks. Trials assessing tube feeding or parenteral nutrition were excluded, as were trials assessing dietary advice alone without the actual provision of supplements.

 

Control

No nutritional intervention, placebo, or dietary advice alone.

 

Types of outcome measures

 

Primary outcomes

  • All-cause death.
  • Cure (completed treatment and sputum-smear or sputum-culture negative) at six and 12 months.

 

Secondary outcomes

  • Completion of treatment.
  • Sputum positive at follow up.
  • Self-reported recovery from illness or resolution of symptoms.
  • Change in weight, skinfold thickness, or other measure of lean or total mass.
  • Any measure of growth in children.
  • Any measures of physical functioning, quality of life, or ability to return to work.

We intended to include cure assessed at six and 12 months, as is customary. For other outcome measures, data presented at any time point was accepted.

 

Search methods for identification of studies

We attempted to identify all relevant trials regardless of language or publication status (published or unpublished, in press, or in progress).

 

Electronic searches

We searched the following databases using the search terms and strategy described in Table 1: Cochrane Infectious Disease Group Specialized Register; Cochrane Central Register of Controlled Trials (CENTRAL); MEDLINE; EMBASE; and LILACS. We also searched the metaRegister of Controlled Trials (mRCT) using 'tuberculosis' and 'supplementation' as search terms.

In addition we searched the Indian Journal of Tuberculosis using the keywords given in the search strategy (Appendix 1).

 

Searching other resources

We also checked the reference lists of all studies identified by the above methods.

 

Data collection and analysis

 

Selection of studies

Two authors (Katharine Abba, Thambu Sudarsanam or David Sinclair) independently screened all citations and abstracts identified in the search for potentially eligible studies. Full reports of potentially eligible studies were obtained and assessed for inclusion in the review by the same authors using a pre-designed eligibility form based on the inclusion criteria. Where it was unclear whether a study was eligible for the review, we attempted to contact the authors for clarification. We resolved differences in opinion by discussion and, where necessary, by discussion with a fourth author (Jimmy Volmink). We screened all papers for multiple publications. We excluded studies that did not meet the criteria and documented the reasons for their exclusion.

 

Data extraction and management

Two authors (Liesl Grobler, KA, TS or DS ) independently extracted data using a tailored data extraction form. We extracted data on study design, participant characteristics, interventions, and outcomes.

For dichotomous data, we extracted the number of participants with the outcome and the total number analysed. For continuous data, we extracted the arithmetic mean and standard deviation for each group. If medians were used, we also extracted the ranges where possible. If there was skewed continuous data, we planned to extract geometric means where presented by the author. We resolved any discrepancies between the extracted data by discussion.

 

Assessment of risk of bias in included studies

Two authors independently assessed components of risk of bias of the included trials using the Cochrane Collaboration's tool for assessing the risk of bias (Review Manager 5), and discussed any differences of opinion. We followed the guidance to assess whether adequate steps were taken to reduce the risk of bias across six domains: sequence generation, allocation concealment, blinding (of participants, personnel and outcome assessors), incomplete outcome data, selective outcome reporting and other sources of bias. We have categorise our judgements as 'yes' (low risk of bias), 'no' (high risk of bias) or 'unclear'. Where our judgment was unclear we attempted to contact the authors for clarification.

The risk of bias judgements are displayed in a table and summarised in Figure 1.

 FigureFigure 1. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

 

Measures of treatment effect

Interventions were compared using risk ratios (RR) for dichotomous data, and mean difference (MD) for continuous data. All results are presented with 95% confidence intervals (CIs).

 

Unit of analysis issues

Trials including more than two comparison groups have been split and analysed as individual pair-wise comparisons. When conducting meta-analysis we have ensured that participants and cases in the placebo group were not counted more than once, by dividing the placebo cases and participants evenly between the intervention groups..

 

Dealing with missing data

If data from the trial reports were insufficient, unclear, or missing, we attempted to contact the authors for additional information.

 

Assessment of heterogeneity

Heterogeneity amongst trials was assessed by inspecting the forest plots (to detect overlapping CIs), the I² statistic with a level of 50% to denote moderate levels of heterogeneity, and applying the Chi² test with a P value of 0.10 to indicate statistical significance.

 

Assessment of reporting biases

We planned to assess the likelihood of publication bias by examining the funnel plots for asymmetry, however there were too few trials to make this assessment meaningful.

 

Data synthesis

The data were analysed using Review Manager 5. Treatments are compared directly using pair-wise meta-analyses. Meta-analyses have been stratified by time-point where appropriate.

When there was no statistically significant heterogeneity we used the fixed-effect meta-analysis model. When moderate statistically significant heterogeneity was observed within groups that could not be explained by subgroup or sensitivity analyses a random-effects meta-analysis model was used to synthesize the data. When a pooled meta-analysis result was considered to be meaningless because of clinical or substantial statistical heterogeneity the results are presented in a forest plot without a pooled estimate of effect.

Highly skewed continuous data (where the standard deviations were larger than the means) are only presented in tables.

 

Subgroup analysis and investigation of heterogeneity

Due to the small number of trials available for each comparison, the investigation of heterogeneity was not necessary or possible.

 

Sensitivity analysis

A sensitivity analysis was planned to investigate the robustness of the results to the risk of bias components, however there are too few trials for each comparison for this to be possible.

 

Results

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.

 

Results of the search

The original search (June 2008) identified forty-seven articles, and the update search (July 2011) identified a further 17 articles. Full text articles of 32 articles were retrieved and underwent screening for inclusion. In total 24 reports, covering 23 individually randomised controlled trials, met the inclusion criteria (see Figure 2).

 FigureFigure 2. Study flowchart (PRISMA template)

 

Included studies

 
Participants

The 23 trials included 6842 participants. Seventeen trials included only adults being treated for pulmonary tuberculosis ( Karyadi 2002; Schön 2003; Paton 2004; Pérez-Guzmán 2005; Range 2005; Nursyam 2006; Seyedrezazadeh 2006; Semba 2007; Villamor 2008; Martins 2009; Armijos 2010; Lawson 2010; Pakasi 2010; Praygod 2011; Praygod 2011a; Martineau 2011; Visser 2011), and three trials also included adults with extrapulmonary tuberculosis (Wejse 2008; Sudarsanam 2010; Jahnavi 2010). Likewise of three studies in children, one included only pulmonary tuberculosis (Hanekom 1997), and two also included extrapulmonary tuberculosis (Morcos 1998; Mehta 2010).

 
HIV status

Seven trials specifically included people with HIV, and presented some results separately for HIV positive and HIV negative participants (Schön 2003; Range 2005; Semba 2007; Wejse 2008; Villamor 2008; Lawson 2010; Sudarsanam 2010). Three of the trials presenting results for HIV positive and HIV negative participants separately used stratified randomization (Villamor 2008; Wejse 2008,Sudarsanam 2010), but four tested participants for HIV following randomization (Schön 2003; Range 2005; Semba 2007; Lawson 2010) and therefore the subgroup analyses by HIV status cannot be said to be truly randomised. Three trials included both HIV-positive and HIV-negative individuals and reported on numbers but did not present results separately (Martineau 2011; Praygod 2011a; Visser 2011). One trial included only HIV-positive participants (Praygod 2011). None of the HIV-positive participants were receiving antiretroviral treatment and one trial was stopped early when antiretrovirals became locally available (Semba 2007). Of the remaining twelve trials, six excluded people with HIV infection (Hanekom 1997; Paton 2004; Pérez-Guzmán 2005; Nursyam 2006; Armijos 2010; Jahnavi 2010), and six did not mention HIV infection (Morcos 1998; Karyadi 2002; Seyedrezazadeh 2006; Martins 2009; Mehta 2010; Pakasi 2010).

 
Study site

Trials were undertaken in:

 
Interventions

Six trials assessed macronutrient supplementation (Paton 2004; Pérez-Guzmán 2005; Martins 2009; Sudarsanam 2010; Jahnavi 2010; Praygod 2011), five trials assessed multi-micronutrient supplementation (Range 2005; Semba 2007; Villamor 2008; Mehta 2010; Praygod 2011a) and the remaining 12 trials assessed single or dual micronutrient supplementation.

 
Sample size

Ten of the 23 trials included less than 100 participants in their final analysis. To aid interpretation and inform future research we have calculated the optimal sample size to reliably demonstrate some suggested clinically important results (Appendix 2; Appendix 3).

As micronutrients are a cheap and easily administered intervention, even a small effect on tuberculosis treatment outcomes might be considered clinically important. For example: to demonstrate a reduction in death from the worldwide average of 5% to just 4% (a relative risk reduction of 20%); a sample size of over 13,000 participants would be necessary. This is far above the data included in this review. Similarly an increase in successful cure rate from 80% to 84% would require almost 3000 participants.

For full details of the included trials see the 'Characteristics of included studies' table.

 

Excluded studies

Seven studies thought to be eligible after initial screening were later excluded for reasons given in the Characteristics of excluded studies table. One trial is currently awaiting classification pending further information from the author (Chandra 2004). In addition we are aware of seven relevant ongoing or unpublished trials; see Characteristics of ongoing studies table.

 

Risk of bias in included studies

The risk of bias assessments are summarized in Figure 1.

 

Allocation

Sixteen trials described an adequate method of generating a truly random allocation sequence (Karyadi 2002; Schön 2003; Paton 2004; Range 2005; Semba 2007; Villamor 2008; Wejse 2008; Martins 2009; Lawson 2010; Mehta 2010; Pakasi 2010; Sudarsanam 2010; Martineau 2011; Praygod 2011; Praygod 2011a; Visser 2011). The other trials did not report how the random sequences were generated although all were described as "randomised".

Seventeen trials described an adequate method of ensuring allocation concealment (Armijos 2010; Karyadi 2002; Schön 2003; Paton 2004; Range 2005; Semba 2007; Villamor 2008; Wejse 2008; Martins 2009; Jahnavi 2010; Lawson 2010; Mehta 2010; Sudarsanam 2010; Martineau 2011; Praygod 2011; Praygod 2011a; Visser 2011). The other trials did not provide sufficient information to determine if the allocation sequence was truly concealed from the person allocating participants to the treatment groups.

 

Blinding

It is generally not possible to blind patients to macronutrient supplementation. However, it is possible to blind the outcome assessors but only one of the six studies reports an attempt to do this (Martins 2009).

Sixteen of the 17 trials assessing micronutrients used placebos and adequately blinded participants and study staff to be considered at low risk of bias.

 

Incomplete outcome data

Eight studies were considered at high risk of bias for some outcomes due to high losses to follow-up (Hanekom 1997; Karyadi 2002; Martins 2009; Mehta 2010; Pakasi 2010; Paton 2004; Range 2005; Semba 2007; Sudarsanam 2010; Villamor 2008).

 

Selective reporting

Although treatment outcomes in tuberculosis such as cure, and treatment completion are well established in tuberculosis programmes only one trial reported cure (Sudarsanam 2010), and four reported treatment completion (Jahnavi 2010; Pakasi 2010; Martins 2009; Sudarsanam 2010). Trial protocols were not retrieved.

 

Other potential sources of bias

One trial did not adequately describe baseline nutritional status (Morcos 1998), and one trial had a large imbalance in HIV status at baseline (Mehta 2010).

 

Effects of interventions

See:  Summary of findings for the main comparison;  Summary of findings 2

 

1. Macronutrients

 

1.1 Increased energy supplementation (average daily requirement for a male adult: 2500 Kcal)

Five trials have examined the effects of providing macronutrient supplements through the health service (Jahnavi 2010; Martins 2009; Paton 2004; Praygod 2011; Sudarsanam 2010). Two of these trials confirmed that supplementation had increased nutritional intake compared to dietary advice alone, and not simply substituted food that patients might have obtained elsewhere (Paton 2004; Sudarsanam 2010). In neither of these trials was the total energy intake with supplementation higher than the average daily requirement for the non-infected population (2500 kcal/day).

 
Tuberculosis treatment outcomes

The number of deaths reported from these trials was very low, and trials were too small to reliably detect or exclude important differences in mortality (two trials, 202 participants,  Analysis 1.1), or cure (one trial, 102 participants,  Analysis 1.2). Jahnavi 2010 did find a statistically significant difference in treatment completion in favour of supplementation, but this was not seen in the larger trial conducted in Timor Leste (two trials, 365 participants,  Analysis 1.3).

Jahnavi 2010 also found that more patients given supplements were smear negative at 8 weeks, while Martins 2009 found no statistically significant difference (two trials, 149 participants, RR 1.2, 95% CI 1.02 to 1.40,  Analysis 1.4).

 
Nutritional recovery and quality of life

Supplementation does seem to improve weight gain during treatment, although one large trial exclusively in HIV co-infected patients found no difference at any time-point (four trials, 731 participants,  Analysis 1.5).

Because the interventions are so different in nature, we will now present their results separately.

 
A daily cooked meal

In Timor Leste, adults with pulmonary tuberculosis and a mean weight of 43.3 kg (+/- 7.5 kg) were randomised to nutritional advice plus a daily cooked meal or nutritional advice alone (Martins 2009). Of note; 31% of participants had a BMI of less than 16 kg/m². The daily midday meal (administered for 2 months during the intensive phase) consisted of a bowl of meat, kidney beans and vegetable stew with rice. During the continuation phase participants in the supplement group also received a weekly food parcel containing unprepared red kidney beans, rice and oil adequate for one meal per day.

 
Tuberculosis treatment outcomes

No participants in the study died after enrolment and no statistically significant difference was found between the comparison groups for treatment completion, cough clearance, or sputum clearance (one trial, 365 participants,  Analysis 2.1;  Analysis 2.2;  Analysis 2.3).

 
Nutritional recovery and quality of life

The participants in the supplement group did however gain significantly more weight when measured at 8 weeks (Mean difference in mean weight gain 1.70 kg, 95% CI 0.19 to 3.21,  Analysis 2.4) and 32 weeks (MD 2.60 kg, 95% CI 0.52 to 4.68,  Analysis 2.4).

 
High energy dietary supplements

Three trials have compared daily high energy supplements with nutritional advice alone (Jahnavi 2010; Paton 2004 & Praygod 2011).

Jahnavi 2010 randomised 100 HIV-negative adults in India to receive food supplements with a target intake of 35 kcal/day/kg bodyweight with specific dietary advice or general dietary advice only. The supplements were provided as sweet balls (containing 600 kcal, and 6 g protein, and made from wheat flour, caramel, ground nuts and vegetable ghee) and 100g of sprouted grains and nuts to provide vitamins and minerals (amounts not stated). The mean BMI at baseline was 17.9 kg/m² in the supplement group and 17.1 kg/m² in the control group.

In Singapore, Paton 2004 randomised 36 HIV-negative adults with pulmonary tuberculosis and a BMI < 20 kg/m² to nutritional advice plus high energy oral supplements, or nutritional advice alone. A target energy intake was calculated for each participant and advice was provided on how to reach this target based on a 24-hour food diary. The intervention group were additionally advised to consume two or three packets of an oral supplement each day between meals until they reached a BMI of 20 kg/m² or their normal body weight. Each 200 ml supplement contained 6.25 g protein, 20.2 g carbohydrate, and 4.29 g fat (approximately 20% protein, 66% carbohydrates, and 14% fat), providing 300 kcal per packet. The energy intake, as assessed by 24-hour food diaries, was confirmed as being significantly higher in the supplement group during the first six weeks of the trial (one trial, 36 participants; Mean energy intake: 2562 +/- 460 kcal /day supplement group vs 1640 +/- 440 kcal/day control group, P < 0.001, authors own figures).

In a much larger trial of tuberculosis /HIV co-infected patients in Tanzania (Praygod 2011), adults were randomised to receive six or one energy-protein bars each day for two months. Each bar contained 4.5 g protein, 146 kcal energy and a variety of micronutrients (120 mg phosphate, 120 mg calcium, 36 mg magnesium, 70 mg sodium, 150 mg potassium and traces of iron and zinc). In addition one of the biscuits given to each group contained the following; 1.5 mg vitamin A, 20 mg vitamin B1, 20 mg vitamin B2, 25 mg vitamin B6, 50 g vitamin B12, 0.8 mg folic acid, 40 mg niacin, 200 mg vitamin C, 60 mg vitamin E, 5 µg vitamin D, 0.2 mg selenium, 5 mg copper, and 30 mg zinc. The mean BMI at baseline was 18.7 kg/m² in the supplement group and 18.5 kg/m² in the control group.

 
Tuberculosis treatment outcomes

Only Jahnavi 2010 reports on deaths, treatment completion and sputum clearance. Only two deaths occurred, both in the control group, but the trial is significantly underpowered to be able to detect an effect (one trial, 100 participants,  Analysis 3.1). A statistically significant benefit with supplementation was seen on treatment completion (one trial, 100 participants: RR 1.20, 95% CI 1.04 to 1.37,  Analysis 3.2) and sputum conversion at eight weeks (one trial, 72 participants: RR 1.21, 95% CI 1.02 to 1.43,  Analysis 3.3).

 
Nutritional recovery and quality of life

The two smaller trials in HIV-negative patients found that supplementation resulted in significantly more weight gain than advice alone at six weeks (one trial, 34 participants; MD 1.73 kg, 95% CI 0.81 to 2.65,  Analysis 3.4) and 12 weeks (one trial, 100 participants; MD 2.6 kg, 95% CI 1.74 to 3.46,  Analysis 3.4) respectively, and Paton 2004 further quantified this as an increase in lean body mass (one trial, 34 participants, MD 1.13 kg, 95% CI 0.37 to 1.89,  Analysis 3.5), with no significant difference in total fat mass ( Analysis 3.6). The large trial from Tanzania found no significant difference in weight gain with supplementation in HIV-positive patients at eight or 20 weeks (one trial, 332 participants,  Analysis 3.4).

All three trials report changes in maximum grip strength, and again a statistically significant benefit was seen in the small trials of HIV-negative patients but not the larger trial of HIV-positive patients although the data appear skewed (three trials, 466 participants,  Analysis 3.7).

Jahnavi 2010 and Paton 2004 also report that the benefits on weight gain and grip strength were accompanied by improvements in some quality of life scores. It was not possible to assess whether these difference were statistically significant because the data appeared highly skewed (the standard deviations were larger than the means for most outcomes); see Appendix 4.

 
Monthly rations

In India, adults with pulmonary tuberculosis and a BMI <19 kg/m² were randomised to receive a macro and micronutrient supplement plus standard care versus standard care alone (Sudarsanam 2010). The supplement consisted of three daily servings of a cereal and lentil mixture (providing 930 kcal and 31.5 g protein) and a once a day multivitamin tablet. Patients were given a months supply of supplement at a time.

 
Tuberculosis treatment outcomes

No effect was shown on the risk of death, cure, or treatment failure but the trial is underpowered to detect clinically important effects even if they existed (one trial, 103 patients,  Analysis 4.1;  Analysis 4.2)

 
Nutritional recovery and quality of life

Data for mean weight are only presented graphically but there do not seem to be important differences between groups. Change in lean body mass, and percentage body fat were not significantly different between groups (P = 0.479 and P = 0.573 respectively, authors own figures).

 

1.2 Altered dietary composition

One very small trial of 21 patients compared a high cholesterol diet (2500 kcal per day with 800 mg cholesterol per day) with a diet with a similar nutritional profile but lower in cholesterol (2500 kcal per day with 250 mg cholesterol per day) for eight weeks in adults being treated for sputum-culture positive pulmonary tuberculosis (Pérez-Guzmán 2005).

 
Tuberculosis treatment outcomes

This trial did not report on death, cure or treatment completion.

Fewer participants in the high cholesterol group were still sputum-culture positive at two weeks compared with those in the normal cholesterol group (one trial, 20 participants, RR of remaining sputum positive at two weeks 0.22, 95% CI 0.06 to 0.77), the difference was not significant at four weeks, and at eight weeks all participants in both groups were sputum-culture negative. Self reported cough and dyspnoea are reported to have decreased at the same rate in both groups (figures not given).

 
Nutritional recovery and quality of life

Not reported.

 

2. Micronutrients

 

2.1 Multivitamins and trace elements

Four trials in adults being treated for sputum test positive or negative pulmonary tuberculosis (Range 2005; Semba 2007; Villamor 2008, Praygod 2011a), and one trial in children with all forms of tuberculosis (Mehta 2010), have compared daily multiple micronutrient supplements (containing a range of vitamins and trace elements), with placebo. The exact doses of the individual constituents ranged from one to 10 times the DRI, and are given in Appendix 5. In summary; vitamin A (2 to 3 x DRI), B vitamins (1 to 10 x DRI), vitamin C (1 to 5 x DRI), vitamin D (≈ 1 DRI), vitamin E (1 to 10 x DRI), zinc (1 to 5 x DRI), selenium (1 to 4 x DRI).

In two trials participants received daily supplements for two months (Mehta 2010, Praygod 2011a), one trial for eight months (Range 2005), and in two trials they received daily supplements for 24 months (Semba 2007; Villamor 2008).

Tuberculosis treatment outcomes

No statistically significant effect on mortality from tuberculosis has been shown, regardless of HIV status (four trials, 2601 participants;  Analysis 6.1). Much larger trials are necessary before a clinically important effect can be excluded.

There was no statistically significant difference between the supplement and control groups in the numbers of participants who were sputum-culture or sputum-smear positive at one month (two trials, 1020 participants,  Analysis 6.2) or two months (one trial, 439 participants,  Analysis 6.3).

Nutritional recovery and quality of life

Four trials reported changes in weight or body mass using a variety of parameters.

Range 2005 found that participants receiving multiple micronutrients had gained significantly more weight at seven months than those in the placebo group. This was a 2 x 2 factorial study and the difference in weight was statistically significant in both treatment groups compared to placebo. However, in the treatment arm which received both high dose multivitamins and high dose zinc the weight gain appeared clinically important (MD 2.37 kg, 95% CI 2.21 to 2.53; 192 participants,  Analysis 6.4), while in those who received high-dose multivitamins alone it did not (one trial, 198 participants; MD 0.30 kg, 95% CI 0.17 to 0.43,  Analysis 6.4).

The supplement administered by Praygod 2011a included a similar dose of zinc to that used by Range 2005 but found no statistically significant difference in weight gain at two or five months (one trial, 775 participants,  Analysis 6.4) and no significant difference in arm fat area, or arm muscle area. However, a subgroup analysis by HIV status found a statistically significant difference in weight gain in favour of supplements in HIV-negative patients, and in favour of placebo in HIV-positive patients.

In addition, Villamor 2008 reports that the high dose multi-micronutrient supplement had no statistically significant effect on BMI, mid-upper arm circumference, fat mass, or fat free mass at baseline, eight months or 24 months, regardless of HIV status (one trial, 887 participants, figures not reported). Mehta 2010 also reports no effect of multi-micronutrients on median weight, height, or mid-upper arm circumference at two months after supplementation in children with tuberculosis (one trial, 237 participants, see Table 6).

One study (Praygod 2011a) found a statistically significant improvement in mean handgrip strength at two months but not five months (one trial, 771 participants, mean difference (change in handgrip strength) 1.22 kg, 95% CI 0.49 to 1.95,  Analysis 6.5). The clinical importance of this difference is unclear. Consistent with the change in weight, this increase was only present in HIV-negative patients.

 

2.2 Individual micronutrients

 
Vitamin A (DRI: 900µg/3000 IU per day)

Two studies have directly compared vitamin A given alone versus placebo (Hanekom 1997: vitamin A 200,000 IU on Day 0 and Day 1, Pakasi 2010: vitamin A 5000 IU daily). In addition five studies combined vitamin A with zinc (Karyadi 2002: vitamin A 5000 IU daily, Armijos 2010: vitamin A 5000 IU daily, Pakasi 2010: vitamin A 5000 IU daily, Lawson 2010: vitamin A 5000 IU weekly, Visser 2011: vitamin A 200,000 IU on day 1), and four studies gave vitamin A as part of a multi-micronutrient supplement (Range 2005; vitamin 5000 IU daily, Semba 2007; vitamin A 8000 IU daily, Villamor 2008: Vitamin A 5000 IU daily, Praygod 2011a: vitamin A 5000 IU daily).

Six studies report on measures of vitamin A status at baseline and during follow-up (Hanekom 1997; Pakasi 2010; Karyadi 2002; Armijos 2010; Visser 2011; Semba 2007). Hanekom 1997 reports that 62% of participants had vitamin A deficiency at baseline (< 20 µg/dL), but the mean vitamin A level increased in both groups regardless of supplementation, and was not significantly different between groups at six weeks or three months (one trial, 85 participants,  Analysis 7.1). Karyadi 2002 reports that 32% of participants had vitamin A deficiency at baseline (< 70 µmol/l), and Armijos 2010 reports a mean retinol level within the normal reference range in both groups. In both trials the mean vitamin A level rose in both groups regardless of supplementation and was not significantly different at any time point (two trials, 113 participants,  Analysis 9.1). Pakasi 2010 and Visser 2011 reported data as 'median' plasma levels so could not be included in the meta-analysis, and Semba 2007 only presented data graphically. In all three trials vitamin A levels rose in both groups regardless of supplementation. Only Semba 2007 reports a difference that was statistically significant in favour of supplements, but this difference (at eight months) is unlikely to be of clinical significance.

 
Tuberculosis treatment outcomes

No effect on mortality has been seen with vitamin A alone or in combination with other micronutrients (seven trials, 3006 participants,  Analysis 7.2,  Analysis 9.3,  Analysis 6.1). These trials, even the larger ones, are significantly underpowered to rule out a clinically important effect.

Pakasi 2010 reports on treatment completion and no significant difference was seen (one trial, 158 participants,  Analysis 7.3)

Hanekom 1997 reports that more children in the supplement group remained symptomatic after six weeks of tuberculosis treatment than in the control group, but this was not statistically significant (one trial, 76 participants,  Analysis 7.4). The trial authors also report no statistically significant differences in respiratory symptoms at three months, or in chest x-ray resolution; but specific data were not presented.

In Pakasi 2010 there were no statistically significant differences between trial arms in smear negativity at two weeks or one month. At two months, all participants in both arms were smear negative.

 
Nutritional recovery and quality of life

Hanekom 1997 reports that the mean weight z-score at baseline was -1.41 (SD 1.41) in the supplement group, and -1.44 (SD 1.34) in the placebo group (it is not clear whether this is weight for age or weight for height). No statistically significant differences in change-in-weight z-scores were recorded at any time-point, but the data were not presented.

Pakasi 2010 reports that the mean BMI at baseline was 16.5 kg/m² (SD 2.2) in the supplement group and 16.4 kg/m² (SD 2.5) in the placebo group. No statistically significant difference between groups was seen at two or six months (one trial, 158 participants,  Analysis 7.6). Pakasi 2010 also reports that there were no statistically significant differences in mid upper arm circumference (MUAC) or percentage body fat ( Analysis 7.7).

 
Zinc (DRI: 11 mg/day)

Three studies have directly compared daily zinc given alone versus placebo (Range 2005: zinc 45 mg daily, Lawson 2010: 90 mg elemental zinc weekly, Pakasi 2010: 15 mg zinc sulphate daily). In addition, five studies have given zinc in combination with vitamin A (Armijos 2010; zinc 50 mg daily, Karyadi 2002; zinc 15 mg daily, Lawson 2010: 90 mg elemental zinc weekly, Pakasi 2010: 15 mg zinc sulphate daily, Visser 2011: zinc 15 mg for five days per week) and three have given zinc as part of a multi-micronutrient supplement (Range 2005; zinc 45 mg daily, Semba 2007; zinc 10 mg daily, Praygod 2011a: zinc 30 mg daily).

Pakasi 2010, Armijos 2010, and Karyadi 2002 report mean plasma zinc levels at baseline and during follow-up. Pakasi 2010 and Armijos 2010 report mean zinc levels within the normal range at baseline, and Karyadi 2002 reports that 30% had low zinc levels (< 10.7 µmol/l). In the trials giving zinc with vitamin A, the mean zinc level fell during the first two months in the placebo group and was significantly lower than the supplement group at this time-point (three trials, 265 participants, MD 1.07 µmol/l, 95% CI 0.48 to 1.66,  Analysis 9.2). At six months the difference was no longer statistically significant (three trials, 246 participants,  Analysis 9.2). However this effect was not seen when zinc was given alone (one trial, 162 participants,  Analysis 8.1). In addition Visser 2011 reports median zinc levels with no statistically significant difference between groups at baseline, two or eight months.

 
Tuberculosis treatment outcomes

No statistically significant effect on mortality by six to eight months was seen with zinc alone or in combination with other micronutrients (five trials, 614 participants,  Analysis 8.2;  Analysis 8.3;  Analysis 9.3;  Analysis 6.1). Only Pakasi 2010 reports on treatment completion and no difference was found (one trial, 152 participants,  Analysis 8.4).

There were no statistically significant differences between the groups in the numbers who were sputum-culture positive at four weeks (three trials, 783 participants,  Analysis 8.5) or eight weeks (three trials, 808 participants,  Analysis 8.5).

 
Nutritional recovery and quality of life

Range 2005 reports that the mean BMI at baseline was 17.8 kg/m² (SD 2.5) in the supplement group and 18.7 kg/m² (SD 2.7) in the placebo group.There was a statistically significant but very small difference between the zinc only and placebo only groups in mean weight at seven months, favouring placebo (MD -0.21 kg, 95% CI -0.36 to -0.06; 193 participants,  Analysis 8.6).

Pakasi 2010 reports a mean BMI at baseline of 16.5 kg/m² (SD 2.2) and 16.4 kg/m² (SD 2.5) in the zinc an placebo groups respectively. There were no statistically significant differences in BMI at two or six months (one trial, 162 participants,  Analysis 8.7).

In Lawson 2010 data on changes in BMI are only presented graphically, and BMI appears to improve at the same rate in all groups.

 
Vitamin A plus zinc (DRI: vitamin A 900 µg/3000 IU, zinc 11 mg per day)

Five studies, in adults with sputum positive pulmonary tuberculosis have compared the combination of vitamin A and zinc versus placebo (Karyadi 2002: vitamin A 5000IU and zinc 15 mg daily for six months, Armijos 2010: Vitamin A 5000 IU plus zinc 50 mg daily for four months, Lawson 2010: vitamin A 5000 IU/day plus 90 mg elemental zinc/week for six months, Pakasi 2010: vitamin A 5000 IU plus 15 mg zinc sulphate daily for six months, Visser 2011: vitamin A 100,000 IU at baseline plus zinc 15 mg for five days per week for two months).

The effect of supplementation on vitamin A and zinc levels has been reported above.

 
Tuberculosis treatment outcomes

Mortality was not the primary outcome for any of these five trials, and consequently they are not adequately powered to detect an effect. However, three trials report on deaths occurring during the trial, and the risk of death was significantly higher in HIV-positive participants receiving micronutrients (two trials, 119 participants: RR 6.40, 95% CI 1.16 to 35.20,  Analysis 9.3), although there was no statistically significant difference in HIV-negative individuals (three trials, 313 participants). Only Pakasi 2010 reports on treatment completion and found no statistically significant difference between the groups (one trial, 152 participants,  Analysis 9.4).

No statistically significant difference has been shown between supplementation and placebo in the number of participants who were sputum-smear positive at one, or two months (five trials, 639 participants,  Analysis 9.5). One study (Armijos 2010) did find a statistically significant difference in sputum positivity at three months in favour of supplementation (one trial, 33 participants: RR 0.12, 95% CI 0.02 to 0.84,  Analysis 9.5), but the difference was not significant at two or four months. Visser 2011 found no statistically significant difference in time to smear or culture conversion (one trial, 154 participants: P = 0.15 and P = 0.38 respectively, authors own figures).

 
Nutritional recovery and quality of life

In Karyadi 2002 the mean BMI was 17.6 kg/m² (SD 1.9) in the supplement group, and 18.1 kg/m² (SD 3.16) in the placebo group suggesting significant undernutrition. In Pakasi 2010 the mean BMI was 16.6 kg/m² (SD 2.1) in the supplement group, and 16.4 kg/m² (SD 2.5) in the placebo group. There were no statistically significant differences in BMI at two or six months (two trials, 232 participants'  Analysis 9.7).

In Karyadi 2002 multiple other anthropometric measures were used. Of these only mean body weight was significantly higher in the group treated with vitamin A plus zinc compared to placebo at six months (one trial, 80 participants, MD 3.10 kg, 95% CI 0.74 to 5.46; 80 participants,  Analysis 9.8); the point estimate of which was 2 kg greater than the difference at baseline (MD 1.1 kg, 95% CI -1.26 to 3.46,  Analysis 9.8). There were no statistically significant differences in mid-upper arm circumference, biceps skinfold thickness, triceps skinfold thickness, subscapular skinfold thickness, supra-iliac skinfold thickness, body fat (%) or fat mass (kg) at any time point (one trial, 80 participants,  Analysis 9.9;  Analysis 9.10;  Analysis 9.11;  Analysis 9.12;  Analysis 9.13;  Analysis 9.14;  Analysis 9.15).

In Visser 2011 the mean BMI at baseline in the supplement group was 18.9 kg/m² (SD 2.7) in males and 23.0 kg/m² (SD 4.3) in females and similar in the placebo group. There was no statistically significant difference in weight gain during the first two months of treatment (one trial, 154 participants: mean weight gain 2.3 kg supplement group vs 2.2 kg placebo group, P = 0.68, authors own figures). Lawson 2010 only reports BMI data graphically but there do not seem to be important differences between groups at two or six months.

Two trials report on changes in Karnofsy score, a rating scale of a persons ability to carry out activities of daily living ranging from 0 (dead) to 100 (normal). Karyadi 2002 reports that supplementation resulted in a small but statistically significant difference in Karnofsky score at six months (one trial, 80 participants, MD 2.5%, 95% CI 0.91 to 4.09; 80 participants,  Analysis 9.6), and Lawson 2010 found no difference at two or six months but only presented data graphically (one trial, 233 participants). A difference in Karnofsky score of 2.5% is unlikely to be of clinical significance.

 
Vitamin D (DRI: 5 to 15 µg/200 to 600 IU per day)

Two trials have compared daily vitamin D supplements versus placebo (Morcos 1998: 1000 IU per day for eight weeks, Nursyam 2006: 250 µg/day for six weeks), one trial compared three high doses of vitamin D versus placebo (Wejse 2008: 100,000 IU at 0, 5 and 8 months), and one trial compared four high doses of vitamin D versus placebo (Martineau 2011: 2.5 mg vitamin D3 on days 0, 14, 28 and 42). In addition, two trials of multi-micronutrients versus placebo included vitamin D in standard daily doses (Range 2005; 5 µg per day, Semba 2007; 10 µg per day).

Only Morcos 1998, Wejse 2008 and Martineau 2011 recorded the vitamin D status at baseline and during follow-up. Morcos 1998 reports a mean vitamin D level of 17.91 pg/ml at baseline which is below the normal reference range (20 to 42 pg/ml). The vitamin D level rose in both groups and was not statistically different after four months of supplementation (P > 0.05, authors own figures). At baseline Wejse 2008 reports that approximately 10% of participants were defined as being vitamin D deficient (serum 25 (OH)D3 < 50 nmol/L), and 45% as vitamin D insufficient (serum 25(OH)D3 < 75 nmol/L), there were no statistically significant differences at 0, 2 or 8 months (one trial, 203 participants, P > 0.05, authors own figures). Martineau 2011 reports that almost all participants had serum vitamin D levels at baseline below the definition of insufficiency used by Wejse 2008 (95% supplement group vs 98% control group). At day 56 the mean serum vitamin D was significantly higher in those who received the supplement (101.4 nmol/L supplement vs 22.8 nmol/L placebo, P < 0.0001, authors own figures).

 
Tuberculosistreatment outcomes

Wejse 2008 found no statistically significant difference between the supplement and placebo groups in the number of deaths in adults after 12 months follow-up, regardless of vitamin D or HIV status at baseline (one trial, 359 participants,  Analysis 10.1). An additional 828 patients received standard dose vitamin D as part of a multi-micronutrient supplement and no effect on mortality was shown (three trials, 1459 participants,  Analysis 6.1). Morcos 1998 and Nursyam 2006 do not report on death, cure or treatment completion, and Martineau 2011 only reports one death in each treatment arm.

Wejse 2008 also found no statistically significant difference in recovery, as defined by a newly developed tuberculosis scoring system (one trial, 359 participants,  Analysis 10.2). This system rates the patients condition on a scale of 0 to 13, based on signs and symptoms and anthropometric measurements (Wesje 2008b).

During intermittent supplementation (as used by Wejse 2008), there was no statistically significant difference in the proportion of people with positive sputum at baseline, 2, 5 or 8 months (one trial, 359 participants,  Analysis 10.3). Similarly Martineau 2011 found no statistically significant difference in median time to culture clearance (one trial, 126 participants, P = 0.41, authors own figures). Daily supplementation with vitamin D (Nursyam 2006) did show a statistically significant difference in the proportion of participants who remained sputum positive at six weeks (RR 0.06, 95% CI 0.00 to 0.95; one trial 77 participants,  Analysis 10.3); but the difference was not significant two weeks later.

 
Nutritional recovery and quality of life

Nursyam 2006 and Martineau 2011 found no statistically significant difference in mean BMI after six and eight weeks of supplementation respectively (two trials, 193 participants,  Analysis 10.4). Morcos 1998 found no statistically significant difference in body weight between the vitamin D and usual care groups after two months of daily supplementation (one trial, 24 participants,  Analysis 10.5). Wejse 2008 found no statistically significant difference in weight gain at eight months (one trial, 359 participants, P = 0.9, authors own figures).

Adverse events

Wejse 2008 also reports that of the 365 included participants; one in the vitamin D group and two in the placebo group developed calcium levels above the upper limit of the reference range. No cases of symptomatic hypercalcaemia were observed. Martineau 2011 reported that supplementation was discontinued in three patients in the vitamin D group and no patients in the placebo group due to adverse events judged by physicians to be potentially due to vitamin D. These consisted of mild hypercalcaemia and worsening of tuberculous abscesses. Rates of other serious and non-serious adverse events were similar between the two groups.

 
Vitamin E and selenium capsules (DRI: vitamin E 15 mg, Selenium 55 µg per day)

One trial compared a daily vitamin E (140mg) and selenium (200µg) supplement with placebo in adults being treated for sputum-smear positive pulmonary tuberculosis (Seyedrezazadeh 2006).

The authors report the median plasma vitamin E and selenium levels at baseline and at eight weeks. They report that the median level of both micronutrients rose in the supplement group and decreased in the placebo group. We are unable to assess whether these differences between groups were statistically significant (Appendix 6). In addition one study which gave multi-micronutrients including vitamin E (133 mg) and selenium (65 µg) measured the vitamin E and selenium levels at baseline and during follow-up (Semba 2007). The authors report that both vitamin E and selenium levels were 'significantly higher' in the supplement group after eight months but the data is only presented graphically.

 
Tuberculosis treatment outcomes

No deaths are reported and this paper does not report cure or treatment completion.

There was no statistically significant difference between the supplement and placebo groups in the numbers of participants who were sputum-smear positive at 15, 30, 45 and 60 days after the start of antituberculous treatment (one trial, 35 participants,  Analysis 11.1).

 
Nutritional recovery and quality of life

The authors report 'constant increment' in BMI for the two months of treatment with no statistically significant differences between the groups, but do not present this data.

 
Arginine (not currently considered an essential amino acid)

One trial compared daily supplementation with arginine (1 g) for four weeks compared with placebo in adults being treated for smear positive pulmonary tuberculosis (Schön 2003). Over half (52%) of the participants were HIV-positive.

 
Tuberculosis treatment outcomes

Three deaths were reported, all in HIV-positive patients and with no statistically significant difference between the groups (one trial, 120 participants,  Analysis 12.1). This paper does not report cure or treatment completion.

The authors reported a statistically significant increase in sputum-smear conversion in HIV-negative participants receiving arginine; however, our analysis of the data showed a non-significant difference between groups in the numbers of participants still sputum-smear positive at eight weeks (one trial, 56 participants,  Analysis 12.2).

Fewer HIV-negative participants in the arginine group reported cough at two weeks and at eight weeks (RR 0.38, 95% CI 0.18 to 0.80; 56 participants,  Analysis 12.3); this difference should be viewed with caution as no baseline data were presented for this outcome.

The data for HIV-positive participants is only presented graphically but the difference is reported as non-significant for cough clearance and sputum-smear conversion.

 
Nutritional recovery and quality of life

The authors reported no statistically significant differences in weight gain between the arginine and placebo groups but data were only presented graphically.

 

Discussion

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Summary of main results

 
Macronutrient supplementation

We found no trials where the intervention group achieved a daily kcal intake above the daily recommended intake for non-infected individuals.

The available trials were too small to reliably prove or exclude clinically important benefits on mortality, cure, or treatment completion. One small trial from India did find a statistically significant benefit on treatment completion, and clearance of the bacteria from the sputum, but these findings have not been confirmed in larger trials elsewhere (VERY LOW quality evidence).

The provision of free food or high-energy nutritional products probably does produce a modest increase in weight gain during treatment for active tuberculosis (MODERATE quality evidence). Two small studies found some evidence that physical function and quality of life may also be improved but unfortunately the data were highly skewed and the trials too small to have much confidence in the result (LOW quality evidence). These effects were not seen in the one trial which included exclusively HIV-positive patients.

 
Micronutrient supplementation

There is insufficient evidence to judge whether multi-micronutrients have a beneficial effect on mortality in HIV-negative patients with tuberculosis (VERY LOW quality evidence), but the available studies show that multi-micronutrients probably do not have any effect on mortality in HIV-positive patients with tuberculosis (MODERATE quality evidence). No studies have assessed the effect of multi-micronutrients on cure, or treatment completion.

Multi-micronutrient supplementation may not have any effect on the proportion of tuberculosis patients remaining sputum positive during the first 8 weeks (LOW quality evidence).

Multi-micronutrient supplementation probably has no clinically important effect on weight gain during treatment for active tuberculosis (MODERATE quality evidence).

No studies of multi-micronutrient supplements have assessed quality of life.

 
Individual micronutrients

Although low vitamin A levels are common in tuberculosis, plasma levels probably increase following initiation of tuberculosis treatment regardless of supplementation. There is no evidence that supplementation in doses up to three times the DRI has a beneficial effect on mortality from tuberculosis, or nutritional recovery.

B vitamins have been given in doses up to ten times the DRI as part of multi-micronutrient supplementation. There is no evidence of an effect on mortality from tuberculosis. No conclusions about other effects can be made.

Vitamin C has only been evaluated as part of multi-micronutrient supplements. In doses up to five times the DRI no effect of tuberculosis mortality has been shown. No conclusions about other effects can be made.

Vitamin D deficiency may be seen in patients with tuberculosis but it is unclear whether supplementation improves plasma levels compared to placebo as trials have found inconsistent results. Vitamin D supplements may have a beneficial effect on early sputum conversion but larger trials are needed to confirm this.

Vitamin E has only been assessed in combination with other vitamins or selenium. The dose used was up to 10 times the DRI. None of these trials has shown any significant benefit or harm with supplementation, but supplementation probably does improve blood levels of both vitamin E and selenium.

There is some evidence that plasma zinc levels may fall during the first two months of treatment without supplementation. There is no convincing evidence of other benefits.

 

Overall completeness and applicability of evidence

The included studies are generally too small and too limited to make broad conclusions on the presence or absence of clinically important benefits of nutritional supplementation in tuberculosis.

Although the included studies are from low- and middle-income countries, they may not reflect the food insecure settings were most supplementation programmes take place, and where the benefit may plausibly be greatest.

Where a supplement has so far not shown any benefit, this may also be an issue relating to the dose used as people recovering from tuberculosis may have micronutrient requirements which are higher than healthy people, however there is currently no evidence to support this.

In addition, it should be noted that most of the HIV-positive participants in these trials were not taking antiretroviral therapy.

 

Quality of the evidence

The quality of evidence has been assessed using the GRADE methodology and is displayed in two summary of findings tables: Summary of findings for the main comparison;  Summary of findings 2. 'Moderate' quality evidence implies we can have some confidence in the result, but further research evidence would still be helpful. 'Low' and 'Very low' quality reflect decreasing levels of confidence in the result.

The quality of evidence was mainly downgraded under 'directness' and 'precision'. Directness is an assessment of how well the evidence matches the PICO question being asked (population, intervention, control, outcome). Because nutritional deficiencies are likely to differ widely among populations, it is difficult to generalise the results of one or two trials to other settings or subgroups. Precision involves an assessment of the statistical and clinical significance of the result, and also whether the sample size of the trials is adequate to reliably detect an effect. Most of these trials were small and well below the optimal information size for the outcomes that were being measured.

 

Agreements and disagreements with other studies or reviews

Nutritional supplementation in HIV-positive patients without tuberculosis has been assessed by two further Cochrane reviews (Mahlungulu 2007 & Irlam 2010).

Mahlungulu 2007 found 8 small trials comparing macronutrient supplementation with no supplementation and concluded that although there was evidence that supplementation increased both energy and protein intake, there was no evidence of a clinical benefit on either nutritional or HIV related outcomes.

Irlam 2010 assessed 30 trials comparing micronutrients with placebo from both developed and developing countries. The authors conclude that multi-micronutrient supplements have significant benefits in HIV-infected pregnant and children. In children there was also evidence of a reduction in mortality with vitamin A, and a reduction in diarrhoeal morbidity with zinc.

 

Authors' conclusions

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

 

Implications for practice

There is insufficient research to know whether routinely providing free food or energy supplements results in better tuberculosis treatment outcomes, or improved quality of life.

Although blood levels of some vitamins may be low in patients starting treatment for active tuberculosis, there is currently no reliable evidence that routinely supplementing at or above recommended daily amounts has clinical benefits.

 
Implications for research

High quality studies of food provision to tuberculosis patients in food insecure settings are urgently needed to support the continued expenditure on food support to tuberculosis programmes. In designing these studies, researchers and programme managers need to be clear about the aim of food provision. Whilst an effect on mortality would provide strong advocacy for continued financial support, if the primary aim is to promote adherence, or to mitigate the catastrophic financial consequences of the illness, then these are the outcomes that should be measured, and appropriate comparison interventions should be selected. If the primary aim is to reduce mortality, then future trials must be large enough to reliably detect or exclude an effect.

The failure to demonstrate a beneficial effect with micronutrient supplementation does not imply that one does not exist. Further studies, perhaps using higher doses, would still be beneficial but should have adequate sample sizes to reliably detect or exclude clinically important benefits. It would also be useful if some standardization of outcome measurements could be made.

For nutritional recovery it seems important to assess changes in weight and lean body mass during the first two months of treatment rather than as a single measure at the end of treatment. It is also essential that measures of how this translates into improved quality of life and physical functioning are made, as weight gain on its own is of little interest.

 

Acknowledgements

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

The authors are grateful to Professor Jimmy Volmink who was an author on the first version of this review, who has now stepped down from the author team for this update.

This document is an output from a project funded by the UK Department for International Development (DFID) for the benefit of low- and middle-income countries. The views expressed are not necessarily those of DFID. Additional support for authors was provided by Stellenbosch University, Stellenbosch, South Africa, and Christian Medical College, Vellore, India.

The contact editor for the original version of this review was Dr Mical Paul; the contact editor for this update was Professor Paul Garner.

 

Data and analyses

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
Download statistical data

 
Comparison 1. Macronutrient supplementation

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death (1 year of follow-up)2202Risk Ratio (M-H, Random, 95% CI)0.40 [0.07, 2.25]

    1.1 HIV-positive
122Risk Ratio (M-H, Random, 95% CI)2.14 [0.10, 47.38]

    1.2 HIV-negative
2180Risk Ratio (M-H, Random, 95% CI)0.18 [0.02, 1.48]

 2 Cured (at 6 months)1102Risk Ratio (M-H, Random, 95% CI)0.91 [0.59, 1.41]

    2.1 HIV-positive
122Risk Ratio (M-H, Random, 95% CI)1.38 [0.46, 4.14]

    2.2 HIV-negative
180Risk Ratio (M-H, Random, 95% CI)0.85 [0.53, 1.35]

 3 Treatment completion2Risk Ratio (M-H, Random, 95% CI)Subtotals only

    3.1 All participants
2365Risk Ratio (M-H, Random, 95% CI)1.08 [0.88, 1.33]

 4 Sputum negative at 8 weeks2149Risk Ratio (M-H, Random, 95% CI)1.20 [1.02, 1.40]

 5 Mean weight gain4Mean Difference (IV, Random, 95% CI)Subtotals only

    5.1 After 6 weeks
134Mean Difference (IV, Random, 95% CI)1.73 [0.81, 2.65]

    5.2 After 8 weeks
2597Mean Difference (IV, Random, 95% CI)0.89 [-0.34, 2.13]

    5.3 After 12 weeks
1100Mean Difference (IV, Random, 95% CI)2.6 [1.74, 3.46]

    5.4 After 20 weeks
1306Mean Difference (IV, Random, 95% CI)-0.20 [-1.34, 0.94]

    5.5 After 24 weeks
126Mean Difference (IV, Random, 95% CI)1.78 [-0.25, 3.81]

    5.6 After 32 weeks
1265Mean Difference (IV, Random, 95% CI)2.60 [0.52, 4.68]

 6 Change in maximum grip strength (kg)3Mean Difference (IV, Random, 95% CI)Totals not selected

    6.1 at 6 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

    6.2 at 8 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

    6.3 at 12 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

    6.4 at 20 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

    6.5 at 24 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

 7 Change in quality of life score1Mean Difference (IV, Random, 95% CI)Totals not selected

    7.1 at 6 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

    7.2 at 24 weeks
1Mean Difference (IV, Random, 95% CI)0.0 [0.0, 0.0]

 
Comparison 2. Daily meal versus routine care

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Treatment completion1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    1.1 All participants
1265Risk Ratio (M-H, Fixed, 95% CI)0.98 [0.86, 1.12]

 2 Cough clearance1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    2.1 At 4 weeks
1265Risk Ratio (M-H, Fixed, 95% CI)0.91 [0.68, 1.23]

    2.2 At 8 weeks
1265Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.69, 1.18]

    2.3 At 32 weeks
1265Risk Ratio (M-H, Fixed, 95% CI)1.00 [0.82, 1.22]

 3 Sputum clearance1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    3.1 At 4 weeks
177Risk Ratio (M-H, Fixed, 95% CI)1.13 [0.73, 1.74]

 4 Mean weight gain1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    4.1 After 8 weeks
1265Mean Difference (IV, Fixed, 95% CI)1.70 [0.19, 3.21]

    4.2 After 32 weeks
1265Mean Difference (IV, Fixed, 95% CI)2.60 [0.52, 4.68]

 
Comparison 3. High energy oral supplements and dietary advice vs dietary advice alone

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death1100Risk Ratio (M-H, Fixed, 95% CI)0.2 [0.01, 4.06]

 2 Treatment completion rate1100Risk Ratio (M-H, Fixed, 95% CI)1.20 [1.04, 1.37]

 3 Sputum conversion rate at 8 weeks172Risk Ratio (M-H, Fixed, 95% CI)1.21 [1.02, 1.43]

 4 Change in body weight (kg)3Mean Difference (IV, Fixed, 95% CI)Totals not selected

    4.1 at 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.2 at 8 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.3 at 12 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.4 at 20 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.5 at 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 5 Change in total lean mass (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    5.1 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    5.2 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 6 Change in total fat mass (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    6.1 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    6.2 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 7 Change in maximum grip strength (kg)3Mean Difference (IV, Fixed, 95% CI)Totals not selected

    7.1 at 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    7.2 at 8 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    7.3 at 12 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    7.4 at 20 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    7.5 at 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 8 Change in overall health score2Mean Difference (IV, Fixed, 95% CI)Totals not selected

    8.1 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    8.2 at 12 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    8.3 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 9 Change in quality of life score1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    9.1 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    9.2 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 10 Change in physical function score2Mean Difference (IV, Fixed, 95% CI)Totals not selected

    10.1 6 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    10.2 at 12 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    10.3 24 weeks
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 4. Monthly ration

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death (at 6 months)1102Risk Ratio (M-H, Fixed, 95% CI)0.47 [0.08, 2.80]

    1.1 HIV-positive
122Risk Ratio (M-H, Fixed, 95% CI)2.14 [0.10, 47.38]

    1.2 HIV-negative
180Risk Ratio (M-H, Fixed, 95% CI)0.17 [0.01, 3.10]

 2 Cured (at 6 months)1102Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.60, 1.42]

    2.1 HIV-positive
122Risk Ratio (M-H, Fixed, 95% CI)1.38 [0.46, 4.14]

    2.2 HIV-negative
180Risk Ratio (M-H, Fixed, 95% CI)0.85 [0.53, 1.35]

 3 Treatment failure (before 6 months)1102Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.25, 3.27]

    3.1 HIV-positive
122Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.12, 4.05]

    3.2 HIV-negative
180Risk Ratio (M-H, Fixed, 95% CI)1.16 [0.17, 7.85]

 
Comparison 5. High cholesterol (850 mg/day) vs low cholesterol (250 mg/day) diet

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Sputum-culture positive1Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

    1.1 At baseline
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.2 At 2 weeks
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.3 At 4 weeks
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.4 At 8 weeks
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 6. Multivitamin and trace element tablets vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death during follow-up (subgrouped by HIV status)42601Risk Ratio (M-H, Fixed, 95% CI)0.97 [0.84, 1.11]

    1.1 HIV-negative individuals
3917Risk Ratio (M-H, Fixed, 95% CI)0.76 [0.41, 1.42]

    1.2 HIV-positive individuals
31429Risk Ratio (M-H, Fixed, 95% CI)0.97 [0.84, 1.12]

    1.3 HIV-positive and HIV-negative individuals
1255Risk Ratio (M-H, Fixed, 95% CI)1.59 [0.53, 4.72]

 2 Sputum-smear or sputum-culture positive at 1 month2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    2.1 All
21020Risk Ratio (M-H, Fixed, 95% CI)1.01 [0.85, 1.20]

    2.2 HIV-negative individuals only
1306Risk Ratio (M-H, Fixed, 95% CI)0.73 [0.47, 1.13]

    2.3 HIV-positive individuals only
1322Risk Ratio (M-H, Fixed, 95% CI)0.77 [0.45, 1.31]

 3 Sputum-smear or sputum-culture positive at 2 months1Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

 4 Mean weight gain2Mean Difference (IV, Fixed, 95% CI)Totals not selected

    4.1 After 2 months of supplementation
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.2 After 5 months (only recieving supplementation for the first 2 months)
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.3 After 7 months of supplementation
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 5 Mean change in handgrip strength11480Mean Difference (IV, Fixed, 95% CI)1.16 [0.50, 1.81]

    5.1 At 2 months
1771Mean Difference (IV, Fixed, 95% CI)1.22 [0.49, 1.95]

    5.2 At 5 months
1709Mean Difference (IV, Fixed, 95% CI)0.90 [-0.56, 2.36]

 
Comparison 7. Vitamin A vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Mean serum retinol (normal range > 20 µg/l)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    1.1 At baseline
185Mean Difference (IV, Fixed, 95% CI)-1.0 [-5.38, 3.38]

    1.2 At 6 weeks
185Mean Difference (IV, Fixed, 95% CI)1.30 [-3.91, 6.51]

    1.3 At 3 months
185Mean Difference (IV, Fixed, 95% CI)0.70 [-3.85, 5.25]

 2 Death73006Risk Ratio (M-H, Fixed, 95% CI)0.96 [0.83, 1.11]

    2.1 Vitamin A alone
1115Risk Ratio (M-H, Fixed, 95% CI)1.79 [0.19, 16.69]

    2.2 Vitamin A plus zinc
4535Risk Ratio (M-H, Fixed, 95% CI)2.38 [0.87, 6.53]

    2.3 Vitamin A as part of a multi micronutrient supplement
32356Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.81, 1.08]

 3 Treatment completion1158Risk Ratio (M-H, Fixed, 95% CI)0.92 [0.80, 1.04]

 4 Symptomatic at 6 weeks1Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

 5 Sputum smear/culture positive during follow up1Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    5.1 Baseline
1158Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.98, 1.03]

    5.2 1 month
1148Risk Ratio (M-H, Fixed, 95% CI)0.70 [0.33, 1.48]

    5.3 2 months
1136Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    5.4 2 weeks
1158Risk Ratio (M-H, Fixed, 95% CI)0.80 [0.50, 1.28]

 6 BMI (kg/m2)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    6.1 Baseline
1158Mean Difference (IV, Fixed, 95% CI)0.10 [-0.63, 0.83]

    6.2 2 months
1148Mean Difference (IV, Fixed, 95% CI)0.30 [-0.44, 1.04]

    6.3 6 months
1136Mean Difference (IV, Fixed, 95% CI)-0.30 [-1.15, 0.55]

 7 Body fat (%)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    7.1 Baseline
1158Mean Difference (IV, Fixed, 95% CI)-0.90 [-2.80, 1.00]

    7.2 2 months
1148Mean Difference (IV, Fixed, 95% CI)-0.90 [-2.84, 1.04]

    7.3 6 months
1136Mean Difference (IV, Fixed, 95% CI)-1.80 [-3.96, 0.36]

 
Comparison 8. Zinc vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Serum zinc levels (normal range > 10.7 µmol/L)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    1.1 Baseline
1162Mean Difference (IV, Fixed, 95% CI)-0.20 [-0.91, 0.51]

    1.2 2 months
1151Mean Difference (IV, Fixed, 95% CI)0.0 [-0.76, 0.76]

    1.3 6 months
1140Mean Difference (IV, Fixed, 95% CI)0.5 [-0.11, 1.11]

 2 Death by 6/8 months62459Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.81, 1.11]

    2.1 Zinc alone
3513Risk Ratio (M-H, Fixed, 95% CI)1.20 [0.67, 2.16]

    2.2 Zinc plus vitamin A
4477Risk Ratio (M-H, Fixed, 95% CI)2.08 [0.65, 6.64]

    2.3 Zinc as part of a multimicronutrient supplement
21469Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.76, 1.06]

 3 Death by 6/8 months (subgrouped by HIV status)3614Risk Ratio (M-H, Fixed, 95% CI)1.21 [0.72, 2.03]

    3.1 HIV-negative individuals
2250Risk Ratio (M-H, Fixed, 95% CI)1.78 [0.43, 7.32]

    3.2 HIV-positive individuals
2202Risk Ratio (M-H, Fixed, 95% CI)1.05 [0.59, 1.89]

    3.3 HIV status unknown
1162Risk Ratio (M-H, Fixed, 95% CI)1.70 [0.29, 9.89]

 4 Treatment completion at 6 months1152Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.84, 1.07]

 5 Sputum smear/culture positive during follow-up3Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    5.1 At baseline
2395Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.99, 1.01]

    5.2 At 2 weeks
3806Risk Ratio (M-H, Fixed, 95% CI)1.09 [1.02, 1.17]

    5.3 At 4 weeks
3783Risk Ratio (M-H, Fixed, 95% CI)1.05 [0.93, 1.20]

    5.4 At 8 weeks
3808Risk Ratio (M-H, Fixed, 95% CI)0.94 [0.69, 1.29]

 6 Weight change (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    6.1 At baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    6.2 At 7 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 7 BMI (kg/m2)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    7.1 Baseline
1162Mean Difference (IV, Fixed, 95% CI)0.10 [-0.62, 0.82]

    7.2 2 months
1151Mean Difference (IV, Fixed, 95% CI)0.20 [-0.55, 0.95]

    7.3 6 months
1140Mean Difference (IV, Fixed, 95% CI)0.10 [-0.70, 0.90]

 8 Body fat (%)1Mean Difference (IV, Fixed, 95% CI)Subtotals only

    8.1 Baseline
1162Mean Difference (IV, Fixed, 95% CI)-0.90 [-2.51, 0.71]

    8.2 2 months
1151Mean Difference (IV, Fixed, 95% CI)-1.30 [-3.07, 0.47]

    8.3 6 months
1140Mean Difference (IV, Fixed, 95% CI)-1.5 [-3.51, 0.51]

 
Comparison 9. Zinc + Vitamin A vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Serum retinol (normal range > 70 µmol/l)2Mean Difference (IV, Fixed, 95% CI)Subtotals only

    1.1 At baseline
2113Mean Difference (IV, Fixed, 95% CI)-0.07 [-0.18, 0.03]

    1.2 At 2 months
2113Mean Difference (IV, Fixed, 95% CI)0.07 [-0.05, 0.18]

    1.3 At 6 months
2113Mean Difference (IV, Fixed, 95% CI)0.10 [-0.04, 0.24]

 2 Serum zinc level (normal range > 10.7 µmol/l)3Mean Difference (IV, Fixed, 95% CI)Subtotals only

    2.1 At baseline
3265Mean Difference (IV, Fixed, 95% CI)0.26 [-0.28, 0.79]

    2.2 At 2 months
3252Mean Difference (IV, Fixed, 95% CI)1.07 [0.48, 1.66]

    2.3 At 6 months
3246Mean Difference (IV, Fixed, 95% CI)0.05 [-0.51, 0.61]

 3 Death by 6 months4578Risk Ratio (M-H, Fixed, 95% CI)2.39 [0.93, 6.15]

    3.1 HIV-negative individuals
4447Risk Ratio (M-H, Fixed, 95% CI)1.08 [0.30, 3.91]

    3.2 HIV-positive individuals
2119Risk Ratio (M-H, Fixed, 95% CI)6.40 [1.16, 35.20]

    3.3 Unknown HIV status
112Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 4 Treatment completion at 6 months1152Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.84, 1.07]

 5 Sputum smear/culture positive during follow-up5Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    5.1 At baseline
5652Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.99, 1.01]

    5.2 1 month
4485Risk Ratio (M-H, Fixed, 95% CI)1.01 [0.86, 1.17]

    5.3 2 months
5639Risk Ratio (M-H, Fixed, 95% CI)0.93 [0.73, 1.19]

    5.4 3 months
2266Risk Ratio (M-H, Fixed, 95% CI)0.95 [0.61, 1.49]

    5.5 4 months
2266Risk Ratio (M-H, Fixed, 95% CI)1.51 [0.86, 2.65]

    5.6 5 months
133Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    5.7 6 months
2113Risk Ratio (M-H, Fixed, 95% CI)0.2 [0.01, 4.04]

 6 Karnofsky score1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    6.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    6.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    6.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 7 BMI (kg/m²)2664Mean Difference (IV, Fixed, 95% CI)-0.09 [-0.45, 0.28]

    7.1 Baseline
2232Mean Difference (IV, Fixed, 95% CI)-0.00 [-0.62, 0.61]

    7.2 2 months
2219Mean Difference (IV, Fixed, 95% CI)-0.03 [-0.65, 0.58]

    7.3 6 months
2213Mean Difference (IV, Fixed, 95% CI)-0.24 [-0.91, 0.43]

 8 Body weight (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    8.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    8.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    8.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 9 Mid upper arm circumference (cm)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    9.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    9.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    9.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 10 Biceps skinfold thickness (mm)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    10.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    10.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    10.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 11 Triceps skinfold thickness (mm)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    11.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    11.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    11.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 12 Subscapular skinfold thickness (mm)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    12.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    12.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    12.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 13 Suprailiac skinfold thickness (mm)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    13.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    13.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    13.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 14 Body fat (%)2Mean Difference (IV, Fixed, 95% CI)Totals not selected

    14.1 Baseline
2Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    14.2 2 months
2Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    14.3 6 months
2Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 15 Fat mass (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    15.1 Baseline
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    15.2 2 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    15.3 6 months
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 10. Vitamin D vs placebo or no supplements

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death by 12 months1365Risk Ratio (M-H, Fixed, 95% CI)1.07 [0.67, 1.72]

    1.1 HIV-positive individuals
1131Risk Ratio (M-H, Fixed, 95% CI)1.15 [0.65, 2.02]

    1.2 HIV-negative individuals
1228Risk Ratio (M-H, Fixed, 95% CI)0.92 [0.35, 2.46]

    1.3 HIV status unknown
16Risk Ratio (M-H, Fixed, 95% CI)1.0 [0.18, 5.46]

 2 TB score11142Mean Difference (IV, Fixed, 95% CI)-0.13 [-0.32, 0.06]

    2.1 At baseline
1348Mean Difference (IV, Fixed, 95% CI)-0.10 [-0.52, 0.31]

    2.2 2 months
1297Mean Difference (IV, Fixed, 95% CI)-0.02 [-0.47, 0.42]

    2.3 5 months
1271Mean Difference (IV, Fixed, 95% CI)-0.16 [-0.51, 0.19]

    2.4 8 months
1226Mean Difference (IV, Fixed, 95% CI)-0.19 [-0.52, 0.14]

 3 Sputum-smear positive2Risk Ratio (M-H, Fixed, 95% CI)Subtotals only

    3.1 At baseline
2432Risk Ratio (M-H, Fixed, 95% CI)0.96 [0.85, 1.07]

    3.2 4 weeks
1213Risk Ratio (M-H, Fixed, 95% CI)0.78 [0.49, 1.23]

    3.3 6 weeks
2217Risk Ratio (M-H, Fixed, 95% CI)0.35 [0.15, 0.82]

    3.4 8 weeks
2267Risk Ratio (M-H, Fixed, 95% CI)1.00 [0.40, 2.49]

    3.5 5 months
1148Risk Ratio (M-H, Fixed, 95% CI)3.0 [0.32, 28.18]

    3.6 8 months
1147Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 4 Body Mass Index2Mean Difference (IV, Fixed, 95% CI)Subtotals only

    4.1 At baseline
2193Mean Difference (IV, Fixed, 95% CI)-0.42 [-1.17, 0.33]

    4.2 At 6 to 8 weeks
2193Mean Difference (IV, Fixed, 95% CI)-0.35 [-1.05, 0.35]

 5 Body weight (kg)1Mean Difference (IV, Fixed, 95% CI)Totals not selected

    5.1 Before treatment
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    5.2 At end of supplementation, 2 months after start of antituberculosis treatment
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

    5.3 Difference in means
1Mean Difference (IV, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 11. Vitamin E plus selenium vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Sputum-smear positive at follow up1Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

    1.1 15 days
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.2 30 days
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.3 45 days
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.4 60 days
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 
Comparison 12. Arginine vs placebo

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Death during treatment1115Risk Ratio (M-H, Fixed, 95% CI)1.58 [0.15, 16.44]

    1.1 HIV-negative
156Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    1.2 HIV-positive
159Risk Ratio (M-H, Fixed, 95% CI)1.58 [0.15, 16.44]

 2 Sputum-smear positive at week 81115Risk Ratio (M-H, Random, 95% CI)0.54 [0.05, 5.98]

    2.1 HIV-negative
156Risk Ratio (M-H, Random, 95% CI)0.12 [0.01, 2.07]

    2.2 HIV-positive
159Risk Ratio (M-H, Random, 95% CI)1.31 [0.35, 4.99]

 3 Cough at week 2 (HIV-negative participants only)1Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

    3.1 HIV-negative
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 4 Cough at week 81Risk Ratio (M-H, Fixed, 95% CI)Totals not selected

    4.1 HIV-negative
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

    4.2 HIV-positive
1Risk Ratio (M-H, Fixed, 95% CI)0.0 [0.0, 0.0]

 

Appendices

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Appendix 1. Search strategies for databases


Search setCIDG SRaCENTRALMEDLINEbEMBASEbLILACSb

1tuberculosistuberculosistuberculosistuberculosistuberculosis

2dietary supplementsDIETARY SUPPLEMENTSDIETARY SUPPLEMENTSdietary supplement$dietary supplements

3macronutrientsfood supplement*food supplement*DIET-SUPPLEMENTATIONmacronutrients

4micronutrientsFOOD, FORTIFIEDFOOD, FORTIFIEDMACRONUTRIENTmicronutrients

5zincmacronutrientsmacronutrientsmicronutrient$zinc

62 or 3 or 4 or 5MICRONUTRIENTSMICRONUTRIENTSVITAMIN-SUPPLEMENTATION2 or 3 or 4 or 5

71 and 6TRACE ELEMENTSTRACE ELEMENTSIRON1 and 6

8VITAMINSVITAMINSZINC

9vitamin*vitamin*TRACE-ELEMENT

10zinczinc2-9

11ironiron1 and 10

122-11/OR2-11/ORLimit 11 to human

131 and 121 and 12

14Limit 13 to human



aCochrane Infectious Diseases Group Specialized Register.
bSearch terms used in combination with the search strategy for retrieving trials developed by The Cochrane Collaboration (Higgins 2005); upper case: MeSH or EMTREE heading, lower case: free text term.

 

Appendix 2. Optimal Information Size calculations (tuberculosis treatment outcomes)


OutcomePowerTwo-sided significance levelRatio of group 1:group2Risk in Control groupRR deemed clinically significant
(examples)
Risk in Intervention groupSample size (Total)

Death80%95%15%10.84%213,890

Death80%95%15%10.52.5%31,968

Death80%95%140%40.832%1,028

Death80%95%140%40.520%162

Cure80%95%180%51.188%6658

Cure80%95%180%51.0584%62,894

Sputum positive at 1 month80%95%130%70.7522.5%1,078

Sputum Positive at 2 months80%95%110%70.757.5%4,010



Footnotes:

1 Globally the risk of death in patients receiving treatment for tuberculosis is around 5%.
2 Vitamins are relatively cheap and safe interventions, therefore even very modest reductions in the risk of death might be considered important.
3 A sample size of 2000 participants (higher than any of the included studies) would be necessary to reliably detect even a very large relative reduction in death (50%).
4 A very high risk of death was seen in HIV-positive patients in some trials due to antiretrovirals being unavailable at the study site at the time. Death rates this high should not be seen in patients taking antiretrovirals.
5 The target cure rate for DOTs programs is 80%. The current global average is 86%.
6 Vitamins are relatively cheap and safe interventions, therefore even very modest increases in successful cure might be considered important.
7 Sputum positivity rates in these trials were very variable. These examples are for illustrative purposes only.

Calculations were performed using nMaster 1.0

 

Appendix 3. Optimal Information Size calculations (nutritional and quality of life outcomes)


OutcomePowerTwo-sided significance levelRatio of group 1:group2Mean in control groupSDMean in supplement groupSDMean differenceSample size (Total)

Mean vitamin level180%95%135 µg/l1040 µg/l105 µg/l126

Mean increase in weight at 8 weeks280%95%13.5 kg6.35.5 kg6.32 kg312

Mean increase in weight at 8 weeks280%95%13.5 kg6.38.5 kg6.35 kg50

Mean BMI at 8 weeks380%95%118.5 kg/m²2.619.5 kg/m²2.61 kg/m²214

Mean change in BMI at 8 weeks380%95%118.5 kg/m²2.620.5 kg/m²2.62 kg/m²54

Mean Karnofsky score480%95%1804.0854.05 points20



Footnotes:

1 This example is taken from Hanekom 1997
2 This example is taken from Martins 2009. A 1.7 kg mean difference was shown in this trial.
3 This example uses the SD from Martins 2009 but uses a 5kg mean difference. This is for illustrative purposes.
4 This example uses the SD taken from Karyadi 2002.

 

Appendix 4. High energy oral supplements vs dietary advice (additional data from Jahnavi 2010 and Paton 2004)


OutcomeTimepointSupplementsDietary adviceP value


MeanSDnMeanSDn

Change in physical function score (Jahnavi 2010)At 3 months23.3433.87506.7031.2750

Change in physical function score (Paton 2004)At 6 weeks11.8430.21191.6717.59150.48

At 12 weeks24.4426.446.4116.370.052

At 24 weeks22.7832.801512.0023.66110.500

Change in emotional well-being score (Jahnavi 2010)At 3 months22.3222.69502.5621.4550

Change in mental health score (Paton 2004)At 6 weeks9.0516.481910.0722.20150.781

At 12 weeks10.1326.748.9220.080.581

At 24 weeks11.2024.941511.0023.01110.810

Change in general health score (Jahnavi 2010)At 3 months32.522.17506.5019.4250

Change in overall health score (Paton 2004)At 6 weeks27.6321.88196.6724.02150.053

At 12 weeks30.0019.3625.0027.000.544

At 24 weeks36.6718.581529.1733.43110.465



In total Jahnavi 2010 reports 8 and Paton 2004 reports 12 quality of life/physical function scores. See the original paper for full results.

 

Appendix 5. Composition of multimicronutrient supplements


NutrientAdultsChildren


DRI for male aged 19-70 yrsSemba 2007Range 2005Villamor 2008Praygod 2011aDRI for child aged 1-3 yrsMehta 2010

< 6 months7 to 36 months> 36 months

Vit A900 µg
(3000IU)
2400 µg
(8000 IU)
1500 µg
(5000 IU)
1500 µg
(5000 IU)
1500 µg
(5000 IU)
300 µg---

B1 (Thiamine)1.2 mg1.5 mg20 mg20 mg20 mg0.5 mg0.5 mg1 mg1.5 mg

B2 (Riboflavin)1.3 mg1.7 mg20 mg20 mg20 mg0.5 mg0.6 mg1.2 mg1.8 mg

B3 (Niacin)16 mg20 mg40 mg100 mg40 mg6 mg4 mg8 mg12 mg

B6 (Pyridoxine)1.3 - 1.7 mg2 mg25 mg25 mg25 mg0.5 mg0.6 mg1.2 mg1.8 mg

B9

(Folic Acid)
400 µg400 µg800 µg800 µg800 µg150 µg130 µg260 µg390 µg

B122.4 µg6 µg50 µg50 µg50 µg0.9 µg1 µg1 µg3 µg

Vit C90 mg500 mg200 mg500 mg200 mg15 mg60 mg120 mg180 mg

Vit D5 - 15 µg10 µg
(400 IU)
5 µg-5 µg5 µg---

Vit E15 mg133mg
(200 IU)
60 mg200 mg60 mg6 mg8 mg16 mg24 mg

Selenium55 µg65 µg200 µg100 µg200 µg---

Copper0.9 mg-5 mg-5 mg---

Zinc11 mg10 mg+/- 45 mg (elementary zinc)-30 mg

(as acetate)
---

Iodine150 µg175 µg------

Calcium1000 mg-------

Manganese2.3 mg-------

Magnesium410 - 420 mg-------

D-panthenol-------



Footnotes:

DRI = Dietary Reference Intake

Standards taken from the US Department of Agriculture Dietary Guidance available at http://fnic.nal.usda.gov/nal_display/index

 

Appendix 6. Vitamin E and Selenium levels (Seyedrezazadeh 2006)


At baselineAt 8 weeks


Median (µmol/l)RangeMedian (µmol/l)Range





Plasma Vitamin ESupplement
(n = 17)
24.70.0 to 87.028.210.5 to 86.5

Placebo
(n = 18)
20.25.1 to 4919.35.1 to 48.6

Serum seleniumSupplement
(n = 17)
1.00.34 to 2.5Not reported

Placebo
(n = 18)
0.930.1 to 1.9Not reported



 

What's new

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

Last assessed as up-to-date: 1 July 2011.


DateEventDescription

30 September 2011New citation required but conclusions have not changedEight new trials have been added and we have considered more carefully the nutritional status at baseline(both weight and biochemical status of individual micronutrients). The author team has also been updated.

20 September 2011New search has been performedA new search was performed, and eight new trials added. Summary of findings tables have also been added. These summarise the quality of the evidence, and a calculation of the optimal information size to reliably detect clinically important effects if they exist.

Risk of bias assessments has been updated to the new format.



 

History

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

Protocol first published: Issue 3, 2006
Review first published: Issue 4, 2008


DateEventDescription

10 November 2009AmendedIt was noted by an observant reader that there was an error in referencing in the risk of bias tables. This error has now been corrected.

10 November 2008AmendedMinor errors corrected. No change to conclusions.



 

Contributions of authors

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

This protocol was conceived by Thambu David Sudarsanam and designed in collaboration with all authors. The selection of trials for inclusion, assessment of risk of bias and data extraction was undertaken as indicated in the review methods. The analyses were undertaken mainly by Katharine Abba and David Sinclair in consultation with the other authors.

 

Declarations of interest

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

None known.

 

Sources of support

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms
 

Internal sources

  • Liverpool School of Tropical Medicine, UK.
  • Stellenbosch University, South Africa.
  • Christian Medical College Vellore, India.

 

External sources

  • Department for International Development (DFID), UK.

 

Differences between protocol and review

  1. Top of page
  2. Summary of findings    [Explanations]
  3. Background
  4. Objectives
  5. Methods
  6. Results
  7. Discussion
  8. Authors' conclusions
  9. Acknowledgements
  10. Data and analyses
  11. Appendices
  12. What's new
  13. History
  14. Contributions of authors
  15. Declarations of interest
  16. Sources of support
  17. Differences between protocol and review
  18. Index terms

The objective of the review was changed from "To assess the provision of oral nutritional supplements to promote the recovery of people being treated with anti-tuberculous drug therapy for active tuberculosis" to the current objective to simplify the wording and reflect the need to assess any differences in response depending on HIV status.

Some of the outcomes were also amended, as follows: "change in weight or skinfold thickness" was expanded to "change in weight, skinfold thickness, or other measure of lean or total mass' because we became aware that other measures, beside skinfold thickness, are equally valid indicators of overall nutritional status; "any measure of growth in children" was added because it is a useful indicator of health and nutritional status in children that we had overlooked at the protocol stage; and "sputum positive at follow up" was added because it was a primary outcome of many of the included trials, and it became apparent that it is a meaningful outcome, as, depending on the period of follow up, it may be used as a proxy for cure or as an indicator of the time taken to become sputum-smear negative. Early sputum conversion is a desirable outcome because on becoming smear negative, patients become less ill and less infectious to those around them.

Between the publication of the original review and the update, the method of assessing and reporting risk of bias has changed slightly. The methods have been adapted to reflect this.

For this update, additional detail has been included on the nutritional and micronutrient status at baseline. Where reported we have also included plasma micronutrient levels during follow-up as an outcome. This outcome is not a patient important outcome, and is of little interest on its own, but does contribute to understanding of the results.

* Indicates the major publication for the study

References

References to studies included in this review

  1. Top of page
  2. AbstractRésumé
  3. Summary of findings
  4. Background
  5. Objectives
  6. Methods
  7. Results
  8. Discussion
  9. Authors' conclusions
  10. Acknowledgements
  11. Data and analyses
  12. Appendices
  13. What's new
  14. History
  15. Contributions of authors
  16. Declarations of interest
  17. Sources of support
  18. Differences between protocol and review
  19. Characteristics of studies
  20. References to studies included in this review
  21. References to studies excluded from this review
  22. References to studies awaiting assessment
  23. References to ongoing studies
  24. Additional references
Armijos 2010 {published data only}
Hanekom 1997 {published data only}
Jahnavi 2010 {published data only}
  • Jahnavi G, Sudha CH. Randomised controlled trial of food supplements in patients with newly diagnosed tuberculosis and wasting. Singapore Medical Journal 2010;51(12):957-62.
Karyadi 2002 {published data only}
  • Karyadi E, West CE, Schultink W, Nelwan RH, Gross R, Amin Z, et al. A double-blind, placebo-controlled study of vitamin A and zinc supplementation in persons with tuberculosis in Indonesia: effects on clinical response and nutritional status. American Journal of Clinical Nutrition 2002;75(4):720-7.
Lawson 2010 {published data only}
Martineau 2011 {published data only}
  • Martineau AR, Timms PM, Bothamley GH, Hanifa Y, Islam K, Claxton AP, et al. High-dose vitamin D3 during intensive-phase antimicrobial treatment for pulmonary tuberculosis: a double-blind randomised controlled trial. Lancet 2011;377(9761):242-250.
Martins 2009 {published data only}
  • Martins N, Morris P, Kelly PM. Food incentives to improve completion of tuberculosis treatment: randomised controlled trial in Dili, Timor-Leste. BMJ 2009;229:b4248.
Mehta 2010 {unpublished data only}
  • Mehta S, Mugusi FM, Bosch RJ, Aboud A, Chatterjee A, Finkelstein JL, et al. A Randomized Trial of Multivitamin Supplementation in Children withTuberculosis in Tanzania. clinicaltrials.gov/ct2/show/NCT00145184 unpublished.
Morcos 1998 {published data only}
  • Morcos MM, Gabr AA, Samual S, Kamel M, Baz ME, Beshry ME, et al. Vitamin D administration to tuberculous children and its value. Bollettino Chimico Farmaceutico 1998;137(5):157-64.
Nursyam 2006 {published data only}
  • Nursyam EW, Amin Z, Rumende CA. The effects of vitamin D as supplementary treatment in patients with moderately advanced pulmonary tuberculosis lesion. Acta Medica Indonesiana 2006;38(1):3-5.
Pakasi 2010 {published data only}
  • Pakasi TA, Karyadi E, Suratih NMD, Salean M, Darmawidjaja N, Bor H, et al. Zinc and vitamin A supplementation fails to reduce sputum conversion time in severely malnourished pulmonary tuberculosis patients in Indonesia. Nutrition Journal 2010;9(1):41.
Paton 2004 {published data only}
  • Paton NI, Chua YK, Earnest A, Chee CB. Randomized controlled trial of nutritional supplementation in patients with newly diagnosed tuberculosis and wasting. American Journal of Clinical Nutrition 2004;80(2):460-5.
Praygod 2011 {published data only}
  • Praygod G, Range N, Faurholt-Jepsen D, Jeremiah K, Faurholt-Jepsen M, Aabye MG, et al. The effect of energy-protein supplementation on weight, body composition and handgrip strength among pulmonary tuberculosis HIV-co-infected patients: randomised controlled trial in Mwanza, Tanzania. The British Journal of Nutrition 2011;1(1):1-9 ePub.
Praygod 2011a {published data only}
  • Praygod G, Range N, Faurholt-Jepsen D, Jeremiah K, Faurholt-Jepsen M, Aabye MG, et al. Daily multi-micronutrient supplementation during tuberculosis treatment increases weight and grip strength among HIV-uninfected but not HIV-infected patients in Mwanza, Tanzania. The Journal of Nutrition 2011;141(4):685-91.
Pérez-Guzmán 2005 {published data only}
  • Pérez-Guzmán C, Vargas MH, Quiñonez F, Bazavilvazo N, Aguilar A. A cholesterol-rich diet accelerates bacteriologic sterilization in pulmonary tuberculosis. Chest 2005;127(2):643-51.
Range 2005 {published data only}
  • Range N, Andersen AB, Magnussen P, Mugomela A, Friis H. The effect of micronutrient supplementation on treatment outcome in patients with pulmonary tuberculosis: a randomized controlled trial in Mwanza, Tanzania. Tropical Medicine and International Health 2005;10(9):826-32.
  • Range N, Changalucha J, Krarup H, Magnussen P, Andersen AB, Friis H. The effect of multi-vitamin/mineral supplementation on mortality during treatment of pulmonary tuberculosis: a randomised two-by-two factorial trial in Mwanza, Tanzania. British Journal of Nutrition 2006;95(4):762-70.
Schön 2003 {published data only}
  • Schön T, Elias D, Moges F, Melese E, Tessema T, Stendahl O, et al. Arginine as an adjuvant to chemotherapy improves clinical outcomes in active tuberculosis. European Respiratory Journal 2003;21(3):483-8.
Semba 2007 {published data only}
  • Semba RD, Kumwenda J, Zijlstra E, Ricks MO, van Lettow M, Whalen C, et al. Micronutrient supplements and mortality of HIV-infected adults with pulmonary TB: a controlled clinical trial. International Journal of Tuberculosis and Lung Disease 2007;11(8):854-9.
Seyedrezazadeh 2006 {published data only}
  • Seyedrezazadeh E, Ostadrahimi A, Mohboob S, Asadi Y, Ghaemmagami J, Pourmogaddam M. Effect of vitamin E and selenium supplementation on oxidative stress status in pulmonary tuberculosis patients. Respirology 2008;13(2):294-8.
  • Seyedrezazadeh E, Ostradrahimi AR, Mahboob SA, Assadi Y, Ansarin K, Sakoori P, Pourmoghaddam M. Vitamin E-Selenium supplement and clinical responses of active pulmonary tuberculosis. Tanaffos 2006;5(2):49-55.
Sudarsanam 2010 {published data only}
  • Sudarsanam TD, John J, Kang G, Mahendri V, Gerrior J, Franciosa M, et al. Pilot randomized trial of nutritional supplementation in patients with tuberculosis and HIV-tuberculosis coinfection receiving directly observed short-course chemotherapy for tuberculosis. Tropical Medicine and International Health 2011;16(6):699-706.
Villamor 2008 {published and unpublished data}
  • Villamor E, Mugusi F, Urassa W, Bosch R J, Saathoff E, Matsumoto K, et al. A trial of the effect of micronutrient supplementation on treatment outcome, T cell counts, morbidity, and mortality in adults with pulmonary tuberculosis. Journal of Infectious Diseases 2008;197(11):1499-505.
Visser 2011 {published data only}
  • Visser ME, Grewal HMS, Swart EC, Dhansay MA, Walzl G, Swanevelder S, et al. The effect of vitamin A and zinc supplementation on treatment outcomes in pulmonary tuberculosis: a randomized controlled trial. Am J Clin Nutr 2011;93:93-100.
Wejse 2008 {published and unpublished data}
  • Wejse C, Gomes VF, Rabna P, Gustafson P, Aaby P, Lisse IM, et al. Vitamin D as supplementary treatment for tuberculosis - a double-blind randomised placebo-controlled trial. American Journal of Respiratory and Critical Care Medicine 2009;179:843-50.

References to studies excluded from this review

  1. Top of page
  2. AbstractRésumé
  3. Summary of findings
  4. Background
  5. Objectives
  6. Methods
  7. Results
  8. Discussion
  9. Authors' conclusions
  10. Acknowledgements
  11. Data and analyses
  12. Appendices
  13. What's new
  14. History
  15. Contributions of authors
  16. Declarations of interest
  17. Sources of support
  18. Differences between protocol and review
  19. Characteristics of studies
  20. References to studies included in this review
  21. References to studies excluded from this review
  22. References to studies awaiting assessment
  23. References to ongoing studies
  24. Additional references
Gwinup 1981 {published data only}
Martineau 2009 {published data only}
  • Martineau AR, Nanzer AM, Satkunam KR, Packe GE, Rainbow SJ, Maunsell ZJ, et al. Influence of a single oral dose of vitamin D(2) on serum 25-hydroxyvitamin D concentrations in tuberculosis patients. The International Journal of Tuberculosis and Lung Disease 2009;13(1):119-25.
Mbala 1998 {published data only}
Narang 1984 {published data only}
  • Narang NK, Gupta RC, Jain MK. Role of vitamin D in pulmonary tuberculosis. Journal of the Association of Physicians of India 1984;32(2):185-8.
Oluboyede 1978 {published data only}
  • Oluboyede OA, Onadeko BO. Observation on haematological patterns in pulmonary tuberculosis in Nigerians. Journal of Tropical Medicine and Hygiene 1978;81(5):91-5.
Ramakrishnan 1961 {published data only}
  • Ramakrishnan CV, Rajendran K, Jacob PG, Fox W, Radhakrishnan S. The role of diet in the treatment of pulmonary tuberculosis. An evaluation in a controlled chemotherapy study in home and sanatorium patients in South India. Bulletin of the World Health Organization 1961;25:339-59.
Shi 2001 {published data only}
  • Shi XY, Xiao JQ, YI JY, Zhu YQ, Deng CS. Influence of partial parenteral nutrition with fat emulsion on nutritional status in patients with abdominal tuberculosis. Journal of Clinical Internal Medicine 2001;18(2):120-1.

References to ongoing studies

  1. Top of page
  2. AbstractRésumé
  3. Summary of findings
  4. Background
  5. Objectives
  6. Methods
  7. Results
  8. Discussion
  9. Authors' conclusions
  10. Acknowledgements
  11. Data and analyses
  12. Appendices
  13. What's new
  14. History
  15. Contributions of authors
  16. Declarations of interest
  17. Sources of support
  18. Differences between protocol and review
  19. Characteristics of studies
  20. References to studies included in this review
  21. References to studies excluded from this review
  22. References to studies awaiting assessment
  23. References to ongoing studies
  24. Additional references
ISRCTN16469166 {unpublished data only}
  • ISRCTN16469166. Nutrition and wasting in tuberculosis (TB): Can nutritional supplementation in TB patients improve body weight gain, body composition and treatment outcome?. www.controlled-trials.com/ISRCTN16469166/ISRCTN16469166 (accessed 10 October 2007).
NCT00366470 {unpublished data only}
  • NCT00366470. A clinical trial to study the effect of the addition of vitamin D to conventional treatment in new pulmonary tuberculosis patients. clinicaltrials.gov/ct2/show/NCT00366470 (accessed 10 October 2007).
NCT00507000 {unpublished data only}
  • NCT00507000. Role of oral vitamin D as an adjunct therapy in category I pulmonary tuberculosis along with assessment of immunological parameters. clinicaltrials.gov/ct2/show/NCT00507000 (accessed 10 October 2007).
NCT00698386 {unpublished data only}
  • NCT00698386. Efficacy of Oral Zinc Administration as an Adjunct Therapy in New Pulmonary Tuberculosis (Category I) Patients. ClinicalTrials.gov Accessed on July 21, 2011.
NCT00788320 {unpublished data only}
  • NCT00788320. Antimicrobial Peptide LL-37 (Cathelicidin) Production in Active Tuberculosis Disease: Role of Vitamin D Supplementation. ClinicalTrials.gov Accessed on July 21, 2011.
NCT00801606 {unpublished data only}
  • NCT00801606. Micronutrient Supplementation in Conjunction With Standard Anti-TuberculosisTherapy in Paediatric (6 Months-15 Years) New Pulmonary Tuberculosis Patients. ClinicalTrials.gov Accessed on July 21, 2011.
NCT00918086 {unpublished data only}
  • NCT00918086. Impact of Vitamin D Supplementationon Host Immunity to Mycobacterium Tuberculosisand Response to Treatment: Building Translational Research Capacity in Nutrition and Infectious Diseases in the Republic of Georgia. ClinicalTrials.gov Accessed on July 21, 2011.

Additional references

  1. Top of page
  2. AbstractRésumé
  3. Summary of findings
  4. Background
  5. Objectives
  6. Methods
  7. Results
  8. Discussion
  9. Authors' conclusions
  10. Acknowledgements
  11. Data and analyses
  12. Appendices
  13. What's new
  14. History
  15. Contributions of authors
  16. Declarations of interest
  17. Sources of support
  18. Differences between protocol and review
  19. Characteristics of studies
  20. References to studies included in this review
  21. References to studies excluded from this review
  22. References to studies awaiting assessment
  23. References to ongoing studies
  24. Additional references
Aaron 2004
Arthur 2003
  • Arthur JR, McKenziey RC, Beckett GJ. Selenium in the Immune System. The Journal of Nutrition 2003;133(5):1457S-1459S. (2003).
Barry 2009
  • Barry CE, Boshoff HI, Dartois V, Dick T, Ehrt S, Flynn J, et al. The spectrum of latent tuberculosis:rethinking the biology and intervention strategies. Nature Reviews 2009;7:845-55.
Cegielski 2004
  • Cegielski JP, McMurray DN. The relationship between malnutrition and tuberculosis, evidence from studies in humans and experimental animals. International Journal of Tuberculosis and Lung Disease 2004;8(3):286-98.
Chandra 1996
  • Chandra RK. Nutrition, immunity and infection: from basic knowledge of dietary manipulation of immune responses to practical application of ameliorating suffering and improving survival. Proceedings of the National Academy of Sciences of the United States of America 1996;93(25):14304-7.
Corbett 2003
  • Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Raviglione MC, et al. The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Archives of Internal Medicine 2003;163(9):1009-21.
Davies 1985
Harries 2006
Higgins 2005
  • Higgins J, Green S, editors. Highly sensitive search strategies for identifying reports of randomized controlled trials in MEDLINE. Cochrane Handbook for Systematic Reviews of Interventions 4.2.5 [updated May 2005]; Appendix 5b. www.cochrane.org/resources/handbook/hbook.htm (accessed 1 March 2006).
Irlam 2010
Kamolratanakul 1999
  • Kamolratanakul P, Sawert H, Kongsin S, Lertmaharit S, Sriwongsa J, Na-Songkhla S, et al. Economic impact of tuberculosis at the household level. International Journal of Tuberculosis and Lung Disease 1999;3(7):596-602.
Lounis 2001
Louw 1992
Lutge 2009
Macallan 1999
Mahlungulu 2007
Meyers 2006
  • National Academy of Sciences. Dietary Reference Intakes:The Essential Guide to Nutrient Requirements. 06. National Academies Press, 2006.
Nnoaham 2008
Plit 1998
  • Plit Ml, Theron AJ, Fickl H, Van Rensburg CE, Pendel S, Anderson R. Influence of antimicrobial chemotherapy and smoking status on plasma concentrations of vitamin C, vitamin E, beta-carotene, acute phase reactants, iron and lipid peroxides in patients with pulmonary tuberculosis. International Journal of Tuberculosis and Lung Disease 1998;2(7):590-9.
Review Manager 5
  • The Nordic Cochrane Centre, The Cochrane Collaboration. Review Manager (RevMan). 5.0. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2003.
Saunders 2007
Schluger 1998
Semba 1998
Shankar 1998
  • Shankar AH, Prasad AS. Zinc and immune function: the biological basis of altered resistance to infection. American Journal of Clinical Nutrition 1998;68(2):447S-463S..
Stephenson 2001
Taneja 1990
  • Taneja DP. Observations of serum zinc in patients with pulmonary tuberculosis. Journal of the Indian Medical Association 1990;88(10):280-1.
van Lettow 2003
Wesje 2008b
  • Wesje C, Gustafson P, Nielson J, Gomes VF, Aaby P, Anderson PL, et al. A clinical scoring system for monitoring tuberculosis in a low-resource setting. Scandinavian Journal of Infectious Diseases 2008;40:111-20.
WFP 2007
  • Ba M, Strauss A. Getting Started: WFP food assistance in the context of tuberculosis care and treatment. WFP, 2007.
WHO 2009
  • WHO. Global Tuberculosis Control: A short update on the 2009 report. Geneva: WHO, 2009.
WHO 2010
  • WHO. Global Tuberculosis Control 2010. Geneva: WHO, 2010.
WHO 2010b
  • WHO. Treatment of Tuberculosis Guidelines. 4th Edition. Geneva: WHO, 2010.
Wintergerst 2007
Wyss 2001
Zachariah 2002
  • Zachariah R, Spielmann MP, Harries AD, Salaniponi FM. Moderate to severe malnutrition in patients with tuberculosis is a risk factor associated with early death. Transactions of the Royal Society of Tropical Medicine and Hygiene 2002;96(3):291-4.
Zumla 2000