Single dose vitamin A treatment in acute shigellosis in Bangladeshi children: randomised double blind controlled trial

BMJ 1998; 316 doi: (Published 07 February 1998) Cite this as: BMJ 1998;316:422

This article has a correction. Please see:

  1. Shahadat Hossain, senior medical officer (shossain{at},
  2. Rabi Biswas, research assistanta,
  3. Iqbal Kabir, scientista,
  4. Shafique Sarker, associate scientista,
  5. Michael Dibley, senior lecturerb,
  6. George Fuchs, directora,
  7. Dilip Mahalanabis, former directora
  1. a Clinical Sciences Division, International Centre for Diarrhoeal Disease Research, Bangladesh, GPO Box 128, Dhaka 1000, Bangladesh
  2. b Centre for Clinical Epidemiology and Biostatistics, University of Newcastle, Newcastle, NSW 2300, Australia
  1. Correspondence to: Dr Hossain
  • Accepted 3 October 1997


Objective: To evaluate the efficacy of a single large oral dose of vitamin A in treating acute shigellosis in children in Bangladesh.

Design: Randomised double blind controlled clinical trial.

Setting: Dhaka Hospital, International Centre for Diarrhoeal Disease Research, Bangladesh.

Subjects: 83 children aged 1–7 years with bacteriologically proved shigellosis but no clinical signs of vitamin A deficiency; 42 were randomised to treatment with vitamin A and 41 formed a control group.

Intervention: Children were given a single oral dose of 200 000 IU of vitamin A plus 25 IU vitamin E or a control preparation of 25 IU vitamin E.

Main outcome measures: Clinical cure on study day 5 and bacteriological cure.

Results: Baseline characteristics of the subjects in the two treatment groups were similar. Significantly more children in the vitamin A group than in the control group achieved clinical cure (19/42 (45%) v 8/14 (20%); χ2=5.14, 1 df, P=0.02; risk ratio=0.68 (95% confidence interval: 0.50 to 0.93)). When cure was determined bacteriologically, the groups had similar rates (16/42 (38%) v 16/41 (39%); χ2=0.02, 1 df, P=0.89; risk ratio=0.98 (0.70 to 1.39)).

Conclusions: Vitamin A reduces the severity of acute shigellosis in children living in areas where vitamin A deficiency is a major public health problem.

Key messages

  • A single oral dose of 200 000 IU vitamin A acts as an adjunct in the treatment of acute shigellosis among the children in geographical areas where vitamin A deficiency is a major public health problem

  • Vitamin A supplementation hastens clinical cure in acute shigellosis

  • Vitamin A supplementation during acute shigellosis has no effect on bacteriological clearance

  • Vitamin A may reduce the severity of acute shigellosis by promoting repair of the colonic mucosa and stimulating the immune system


Shigellosis remains one of the most severe enteric infections affecting children in developing countries, including Bangladesh.1 2 It results in the frequent passage of small, bloody mucoid stools; abdominal cramps; and tenesmus caused by ulceration of the colonic epithelium.3 4 5 6 In addition to high mortality from shigellosis, the protein losing enteropathy is a serious complication that probably contributes to the malnutrition and growth stunting associated with the disease.7 8 9 10 11

Reports that vitamin A supplementation reduces childhood mortality from diarrhoeal diseases and measles related diarrhoea have been published.12 13 14 15 16 17 Other studies have found reduced disease severity and lower morbidity in patients with diarrhoea treated with vitamin A.18 19 20 These findings are important for developing countries such as Bangladesh, where shigellosis is one of the most important contributors to childhood morbidity and mortality. Although appropriate antibiotic treatment shortens the course of shigellosis, morbidity and mortality are still high.21 22 23 Reports of beneficial effects of vitamin A supplementation on diarrhoea and evidence that vitamin A has a physiological role in maintaining epithelial integrity and in stimulating specific and non-specific immune functions suggest that it may prove an effective adjuvant to antibiotic treatment in shigellosis.24 25 26 27 28 29 30 31 However, the role of vitamin A in the treatment of acute shigellosis in children has not been evaluated before.

This study aimed to determine whether treatment with vitamin A together with a standard antibiotic would reduce the severity of shigellosis. We carried out a randomised double blind controlled trial in hospital to determine the treatment effect of a large single oral dose of vitamin A in children with proved shigellosis.



The study was organised and conducted at the Dhaka hospital of the International Centre for Diarrhoeal Disease Research, Bangladesh. Altogether 90 subjects were enrolled in the trial during the period September 1992 to December 1994. The study was reviewed and approved by the research review committee and ethics committee of the International Centre for Diarrhoeal Research, Bangladesh. The parents and guardians of all the study children gave written informed consent.

Recruitment and enrolment

Children were eligible for the trial if they were (a) aged between 1 and 7 years, (b) had no other acute or chronic illnesses, (c) presented to the outpatient department because they had been passing bloody mucoid stools (for less than 72 hours), and (d) if microscopic examination of stools showed ≥20 pus cells and erythrocytes per high power field and no trophozoites of Entamoeba histolytica.

Children were excluded from the trial if (a) they had been given antibiotics for the present illness, (b) they had received vitamin A during the previous 3 months, (c) their weight for age was ≤75% of the national child health statistics growth reference median, (d) one of the Shigella sp was not isolated from the baseline stool or rectal swab samples, (e) subsequent microscopic examination of stools showed trophozoites of E histolytica.

The final trial subjects comprised those children in whom infection with Shigella spp had been proved by culture of the stool or rectal swab using a standard technique.32

Study design and randomisation

A randomised double blind controlled design was used for this trial. A computer generated randomisation plan with a block size of six was used to assign individual subjects to a vitamin A or a control preparation. A senior member of staff in the clinical sciences division of the International Centre for Diarrhoeal Disease Research, who was not directly involved in recruiting subjects, subject evaluation, or data analysis, created the list of subjects' identification numbers and corresponding treatment assignments. To conceal treatment allocation, this person provided the appropriate identification number and medication from the list when a subject was ready to be enrolled.

Study intervention

At enrolment, subjects were assigned to treatment with a single oral dose of 200 000 IU of water miscible vitamin A in the form of retinyl palmitate plus 25 IU of vitamin E or a control preparation of 25 IU of vitamin E. The treatment and control preparations were identical in colour, taste, flavour, and volume. Both preparations were kept in an amber coloured bottle, and neither the research team nor the children's parents or guardians knew which preparation the bottles contained. To ensure compliance, the trial treatment and control preparations were administered by a research assistant. On no occasion were details of treatment assignment disclosed; the code was broken only after the final analysis.

Evaluation of patients

At enrolment, a medical history and physical examination were completed by a research assistant (physician) and recorded on a standardised form. Before subjects were randomised to treatment group, stool and rectal swab samples were taken for culture of Shigella spp, Salmonella spp, Vibrio cholerae O1, Plesiomonas shigelloides, and Campylobacter jejuni; a full blood count was done; and serum electrolytes were determined.

Each day fresh stool and rectal swab samples were sent for culture of Shigella spp. Physical examinations and symptoms were recorded every 8 hours by research assistants. Stool frequency and consistency, including the presence of blood or mucus, or both, were determined by a nurse who counted and examined soiled diapers and recorded findings immediately on a tally sheet.

Medical care

Each child was given nalidixic acid (55 mg/kg body weight daily, every 6 hours), but after 48 hours the antibiotic was changed if the sensitivity test suggested this was advisable. Children were admitted to hospital for 5 study days after they had been given the trial treatment.

Outcome measures

The primary end points were achievement of clinical cure on study day 5 and of bacteriological cure. Subjects were considered cured clinically if they passed three or fewer formed stools in a day without any visible blood or mucus, were afebrile, and had no abdominal pain or tenderness on that study day. Bacteriological cure was defined as the continuing absence of Shigella spp in both stool and rectal swab samples from study day 3 onwards. Stool frequency was defined as the number of stools passed each study day. Stools that could be poured or predominantly contained blood or mucus, or both, were considered liquid stools. Complaints of abdominal pain or the presence of abdominal tenderness and a rectal temperature ≥37.8°C during any 8 hour study period were defined as pain and fever respectively for that study day. Study days were deemed to begin at the time trial treatment had been given.

Estimates of sample size

The pretrial sample size was based on a previous study in which 70% of subjects achieved clinical and bacteriological cure with a standard treatment on study day 5.33 Power and sample size calculations were made by computer program.34 For this trial to show a relative risk of 1.4 in the vitamin A group compared with the control group in relation to clinical and bacteriological cure, a maximum sample size of 41 was considered necessary for each group, assuming α=0.05 and power=90%.

Statistical analysis

The stata statistical program (release 4.0) was used for data entry and analysis. Continuous variables that were normally distributed were evaluated by Student's t tests, and the Wilcoxon rank sum test was used for data with a skewed distribution. Bivariate variables were compared by χ2 test. Risk ratios and 95% confidence intervals were calculated with the EpiInfo software package (version 6). Kaplan-Meier survival curves and the log rank test were used to evaluate the difference in time taken to achieve a clinical cure in the groups.


Comparability of groups at baseline

Ninety subjects were initially enrolled in the study, but seven were excluded subsequently. Three control subjects were excluded—one developed haemolytic uraemic syndrome and in two Shigella spp could not be cultured in the initial stool or rectal swab. Four subjects in the vitamin A group were excluded—in two children Shigella spp could not be cultured in the initial stool or rectal swabs, one left hospital against medical advice on study day 2, and one child had E histolytica in the stool sample. The final analysis included 83 subjects–42 in the vitamin A group and 41 in the control group. Baseline characteristics of the trial subjects were similar for each of the treatment groups (1).

Baseline characteristics of trial subjects in relation to treatment group

View this table:

Analysis of clinical cure

Forty five per cent (19 of 42) of children in the vitamin A group and 20% (8 of 41) in the control group were cured clinically by day 5 (χ2=5.14, 1 df, P=0.02; risk ratio=0.68 (95% confidence interval 0.50 to 0.93)). Kaplan-Meier survival curves showed a higher rate of clinical cure in the vitamin A group from day 2 (1). The null hypothesis that the two survival curves are equal was rejected by a log rank test (χ2=4.27, 1 df; P=0.04).


Analysis of bacteriological cure

No significant difference in bacteriological cure between the groups was found. This was achieved by 38% (16 of 42) in the vitamin A group and 39% (16 of 41) in the control group (χ2=0.02, 1 df; P=0.93; risk ratio=0.98 (0.57 to 1.68)).


This is the first published study in which vitamin A has been used as an adjunct to antibiotic treatment in shigellosis. A single oral dose of 200 000 IU of vitamin A accelerated clinical cure in children with shigellosis, and no adverse effects were noted.

Vitamin A might have reduced the severity of shigellosis via two main mechanisms. Firstly, vitamin A is adequately absorbed across the intestinal epithelium, even during episodes of acute diarrhoea, and is probably available to the colonic epithelium almost immediately after absorption.35 36 37 The rapid delivery of vitamin A to the affected site in the intestine may have enhanced repair of the micro-ulcers in the gut epithelium. Secondly, animal studies show that a high dose of vitamin A stimulates phagocytosis and cell mediated killing of pathogens. Vitamin A may therefore reduce the severity of shigellosis by stimulating the immune system.30 31

Because shigellosis affects epithelial tissues extensively and causes an intense immune response, treatment with vitamin A might be expected to reduce the severity of the disease. Although the exact mechanism by which vitamin A reduces the severity of other diseases is not yet known, it has been suggested that it is related to an improvement in immune function and a rapid regeneration of epithelial tissues.38 39

A recent meta-analysis of all the randomised controlled community trials of vitamin A supplementation reported a 39% reduction in mortality from diarrhoeal disease.40 This reduction did not seem to result from a reduction in the incidence of diarrhoeal diseases so much as from a reduction in the severity of attacks. This finding is consistent with our results.

Trials of vitamin A supplementation in Brazil and Ghana have reported that it reduced the incidence or prevalence of severe diarrhoea.41 42 Our results are consistent with these findings and suggest that some beneficial effects of vitamin A in the prevention of diarrhoea may be related to its effects in reducing the severity of dysentery due to Shigella spp. However, no benefit of vitamin A treatment was found in children with acute watery diarrhoea due to enterotoxigenic Escherichia coli or rotavirus.43 Differences in the pathogenesis and duration of episodes of shigellosis compared with enterotoxigenic E coli and rotavirus diarrhoea may explain this variation in the response to treatment with vitamin A. In our study the rate of both clinical and bacteriological cure was lower by day 5 than in the previous studies; this may be explained by the higher resistance of the isolated Shigella spp to nalidixic acid (64%) during the trial period.22 23

The strengths of our trial include the randomised double blind controlled design; successful concealment of treatment allocation; complete follow up of all trial subjects (except one in each group) for 5 days; and the hospital setting, which allowed us to make accurate assessments of the clinical and bacteriological variables. Furthermore, the similarity of treatment groups in terms of baseline characteristics shows that random allocation of subjects to treatment groups was effective.


These results indicate that treatment with a single oral dose of 200 000 IU vitamin A along with a standard antibiotic reduces the severity of acute shigellosis in children aged more that 1 year. However, vitamin A supplementation does not accelerate bacteriological cure. As with other infectious diseases, shigellosis probably increases the requirement for vitamin A.44 45 Among children in countries such as Bangladesh, where vitamin A deficiency is a major public health problem and shigellosis is endemic, vitamin A supplementation should be added to the standard treatment for acute shigellosis.46 This treatment would help to hasten clinical cure and probably restore the vitamin A status. A reduction in the severity of shigellosis will reduce the level of childhood morbidity and mortality in these countries.


We thank Dr Jenny Porteous for reviewing the manuscript and offering helpful suggestions; Dr Michael Coory for statistical analysis; Ms Samsun Nahar and Ms Sakina Begum, who provided nursing care and assisted in data collection; Mr Meer Ramzan Ali for efficient logistic support; and ACME Laboratories (Bangladesh Limited) for supplying the vitamin A and control preparations.

Funding: This research was funded by the United States Agency for International Development under grant no DPE-5986-A-00-1009-00 with the International Centre for Diarrhoeal Disease Research, Bangladesh. The centre is supported by the aid agencies of the governments of Australia, Bangladesh, Belgium, Canada, Japan, the Netherlands, Norway, Saudi Arabia, Sri Lanka, Sweden, Switzerland, the United Kingdom, and the United States; international organisations including Arab Gulf Fund, Asian Development Bank, European Union, the United Nation Children's Fund, the United Nations Development Programme, and the World Health Organisation; private foundations including Aga Khan Foundation, Child Health Foundation, Ford Foundation, Population Council, Rockefeller Foundation, Thrasher Foundation, and the George Masion Foundation; and private organisations including East West Inc, Helen Keller International, International Atomic Energy Centre, Lederle Praxis, New England Medical Centre, Procter Gamble, RAND Corporation. Social Development Center of Philippines, Swiss Red Cross, the Johns Hopkins University, the University of Alabama at Birmingham, UCB Sidac, Wander AG, and others.

Conflict of interest: None.


Contributors: SH, the principal investigator, initiated the study hypotheses, designed the protocol, and did statistical analysis under the supervision of DM and MD. SS and IK participated in protocol design and medical care of subjects. RB recruited subjects, documented data, and entered data into the computer and also gave medical care to subjects. MD, DM, and GF helped in the interpretation of data. The paper was written jointly by SH, MD, and GF. Guarantors: SH, DM.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
View Abstract