Objectives To determine whether (i) supplementation of oral 100 000 iu of vitamin D3 (cholecalciferol) along with antibiotics will reduce the duration of illness in children with pneumonia; (ii) supplementation will reduce the risk of repeat episodes.
Methods Double-blind individually randomised placebo-controlled trial in an inner-city hospital in Kabul, of 453 children aged 1–36 months, diagnosed with non-severe or severe pneumonia at the outpatient clinic. Children with rickets, other concurrent severe diseases, very severe pneumonia or wheeze, were excluded. Children were given vitamin D3 or placebo drops additional to routine pneumonia treatment.
Results Two hundred and twenty-four children received vitamin D3; and 229 received placebo. There was no significant difference in the mean number of days to recovery between the vitamin D3 (4.74 days; SD 2.22) and placebo arms (4.98 days; SD 2.89; P = 0.17). The risk of a repeat episode of pneumonia within 90 days of supplementation was lower in the intervention (92/204; 45%) than the placebo group [122/211; (58%; relative risk 0.78; 95% CI 0.64, 0.94; P = 0.01]. Children in the vitamin D3 group survived longer without experiencing a repeat episode (72 days vs. 59 days; HR 0.71; 95% CI 0.53–0.95; P = 0.02).
Conclusion A single high-dose oral vitamin D3 supplementation to young children along with antibiotic treatment for pneumonia could reduce the occurrence of repeat episodes of pneumonia.
Effets de la supplémentation en vitamine D chez les enfants atteints de pneumonie à Kaboul: un essai contrôlé randomisé
Objectifs: Déterminer si (1) la supplémentation orale de 100.000 UI de vitamine D3 (cholécalciférol), en même temps que des antibiotiques réduirait la durée de la maladie chez les enfants atteints de pneumonie, (2) la supplémentation réduirait le risque d’épisodes répétés.
Méthodes: Essai randomisé placebo contrôlé en double aveugle dans un hôpital à l’intérieur de la ville de Kaboul, sur 453 enfants âgés de 1 à 36 mois, avec un diagnostic de pneumonie sévère ou non sévère dans le service des patients ambulants. Les enfants souffrant de rachitisme, d’autres maladies concurrentes sévères, de pneumonie ou de respiration sifflante très sévère, ont été exclus. Les enfants ont reçu des gouttes de vitamine D3 ou de placebo en plus du traitement de routine de la pneumonie.
Résultats: 224 enfants ont reçu la vitamine D3 et 229 un placebo. Il n’y avait pas de différence significative dans le nombre moyen de jours requis pour la guérison entre le groupe sous vitamine D3 (4,74 jours; SD= 2,22) et le groupe placebo (4,98 jours; SD= 2,89; p = 0,17). Le risque d’épisodes répétés de pneumonie dans les 90 jours de supplémentation était plus faible dans le groupe d’intervention (92/204, 45%) que dans le groupe placebo (122/211 [58%]; risque relatif: 0,78; IC95%: 0,64-0,94; p = 0,01). Les enfants dans le groupe de la vitamine D3 sont restés plus longuement sans éprouver un épisode répété (72 jours vs 59 jours; HR: 0,71; IC95%: 0,53 - 0,95, p = 0,02).
Conclusion: Une dose élevée unique par voie orale de supplémentation en vitamine D3 chez les jeunes enfants sous traitement antibiotique pour une pneumonie pourrait réduire la survenue d’épisodes répétés de pneumonie.
Efecto de la suplementación con vitamina D en niños con diagnóstico de neumonía en Kabul: ensayo aleatorizado y controlado.
Objetivos: Determinar sien niños con neumonía, (1) la suplementación oral con 100,000 iu de vitamina D3 (Cholecalciferol) junto con antibióticos reduce la duración de la enfermedad; (2) si la suplementación reducirá el riesgo de episodios repetidos.
Métodos: Ensayo doble ciego, aleatorizado individualmente, controlado con placebo, de 453 niños con edades comprendidas entre los 1-36 meses y diagnosticados con neumonía no severa y neumonía severa en consultas externas de un hospital de la ciudad de Kabul . Se excluyeron los niños con raquitismo, otras enfermedades graves concurrentes, neumonía muy severa o sibilancias. A los niños se les dió Vitamina D3 o placebo en gotas, además del tratamiento rutinario para la neumonía.
Resultados: 224 niños recibieron vitamina D3 y 229 recibieron placebo. No había una diferencia significativa entre el número medio de días hasta la recuperación entre los brazos de vitamina D3 (4.74 días; SD 2.22) y placebo (4.98 días; SD 2.89; p=0.17). El riesgo de un episodio repetido de neumonía dentro de los 90 días de suplementación era menor en el grupo de la intervención (92/204; 45%) que en el grupo placebo (122/211; (58%; riesgo relativo 0.78; 95% IC 0.64, 0.94; p=0.01). Los niños en el grupo de la vitamina D3 tenían una mayor supervivencia sin experimentar un episodio recurrente (72 días vs. 59 días; HR 0.71; 95%IC 0.53 a 0.95; p=0.02).
Conclusión: En niños pequeños, la suplementación con una dosis oral alta y única de vitamina D3, junto con el tratamiento antibiótico para la neumonía, podría reducir el número de episodios recurrentes.
Pneumonia is the leading cause of childhood mortality, accounting for 19% of the 10.6 million deaths that occur each year (Bryce et al. 2005). Three hospital-based case–control studies from Ethiopia (Muhe et al. 1997) and India (Rehman 1994; Wayse et al. 2004) suggest that vitamin D deficiency may substantially increase the risk of severe pneumonia among children younger than five. In the study in Ethiopia, 42% of hospitalised pneumonia cases had rickets, compared to 4% of children admitted for other reasons (controls); the odds of having vitamin D deficiency were 13.4 times higher in pneumonia cases (95% CI 81, 24.2; P < 0.001) than in the control group (Muhe et al. 1997). In a study in India (Wayse et al. 2004), prevalence of subclinical vitamin D deficiency (serum concentrations of 25-hydroxyvitamin D3 < 22.5 nmol/l) was higher in patients with pneumonia than in healthy children attending clinic for immunisation (80%vs. 31%; P < 0.001). High rates of vitamin D deficiency and rickets have also been found among children admitted to hospital for pneumonia, ranging from 43% in Tehran (Salimpour 1975) and Kuwait (Lubani et al. 1989) to 50% in Yemen and Jordan (Najada & Habashneh 2004).
Vitamin D plays a crucial role in calcium and phosphorous homeostasis, and it is also important for skeletal mineralisation. Its deficiency causes rickets in children and osteomalacia in adults. Vitamin D also plays an important role in modulating the innate immune response against infections (White 2008). In vitro studies have shown that 1,25-dihydroxyvitamin D3, the active metabolite of vitamin D, is important for promoting and regulating immune responses (Rockett et al. 1998; Cantorna 2000; Pichler et al. 2002). Subclinical vitamin D deficiency has been associated with an increased risk of tuberculosis in adults through modification of polymorphisms in the vitamin D receptor (VDR) (Wilkinson et al. 2000).
In children, serum concentrations of 25-vitamin D [25(OH)D] <20 ng/ml (Misra et al. 2008) are considered to be a suboptimal index of vitamin D status, and concentrations <12 ng/ml (Pettifor 2000) are often associated with rickets. Prevalence of vitamin D deficiency is very high in Afghanistan. In the winter of 2005, the median serum concentration of 25(OH) D was 5 ng/ml (range 2–25 ng/ml) among 108 children aged 6–48 months in Kabul; 73% had concentrations <8 ng/ml, a level considered to be significantly deficient (Manaseki-Holland et al. 2008). Thus, we undertook a randomised controlled trial (RCT) in Kabul to assess the effects of vitamin D3 supplementation to children seen at a hospital for treatment of pneumonia. We hypothesised that supplementation of 100 000 iu of vitamin D3 (Cholecalciferol), along with antibiotic treatment, will shorten the duration of illness in children with pneumonia and reduce the risk of repeat episodes of pneumonia, over the next 3 months.
The study was approved by the Ethics and Review Board of the Ministry of Public Health of Afghanistan. Thumbprint or signature consent was obtained from one of the child’s parents at outpatients if the child met the study criteria and after either the parent read the Dari consent form or it was explained to him/her by the doctor.
Study sites and sample population
The trial was conducted at Maywand Hospital, which serves the central city districts of Kabul that constitute the socio-economically deprived population of Kabul, many of whom live in high-walled mud houses. All children between 1 week and 3 years of age from this population diagnosed clinically with ‘pneumonia’ (Box 1) at the local Maywand Teaching Hospital were eligible for inclusion in the trial. Children who had clinical signs of rickets or were known to have received high-dose vitamin D treatment in the past 3 months (one child) had severe vomiting (one child) or pronounced wheeze (10 children) were excluded from the study. (Children who developed wheeze after enrolment were not excluded.) Thirteen children with very severe pneumonias and nine children with other severe illnesses (meningitis, heart or renal disorders, measles, severe malnutrition and suspected tuberculosis) were also excluded. Finally, one child from a family that was likely to migrate out of the study area within 3 months was excluded from the trial.
Randomisation and allocation of study groups
The children were individually randomised into intervention or placebo groups using a random number sequence generated in an Excel spreadsheet with no restrictions, and 460 placebo and 100 000 iu of vitamin D3 (Cholecalciferol; Sinochem Ningbo Laboratory, China; quality certified by Ministry of Health of Pakistan) doses in 1 ml of olive oil were individually packaged into sealed 2-ml plastic syringes at the Department of Pharmacy, Aga Khan University Hospital, Karachi [Joint Commission International Accreditation of Hospitals (JCIA) accredited], Pakistan, and labelled with an unique ID number (only office aware of randomisation codes) and stored in manufacturer’s recommended conditions in a dry, cool environment for 2–8 weeks (depending on the date of recruitment). Placebo (containing olive oil alone) and vitamin D syringes looked the same and the contents tasted the same. None of the investigators, staff in Kabul and caretakers of children, were aware of the study groups. Three paediatricians well trained in the standard operating procedures of the trial enrolled children from 9th December 2006 to 10th February 2007. Children were treated with antibiotic according to the national pneumonia treatment protocol [based upon Integrated Management of Childhood illnesses (IMCI) guidelines] and after obtained consent from parents, children were given either vitamin D3 or placebo orally by the blinded doctors choosing the next syringe with a randomisation code. On random questioning of staff and parents, there were no indications at any stage that families or doctors knew which child may have received placebo or vitamin. As there was only one administration of the intervention, there were no deviations from protocol.
Study children were given an ID card to facilitate follow-up at home and revisits to the local study hospital. The children were followed daily, up to 10 days, either at the study hospital by paediatricians or at home by medical doctors if discharged to assess the resolution of signs and symptoms of the first episode of pneumonia. Thereafter, the children were followed fortnightly up to 90 days (ending mid-May 2007) by trained female medical doctors to assess any illness and to refer to the study hospital if necessary. All doctors involved were trained in IMCI and examination of the study signs and symptoms and their work in the clinics or follow-up were monitored through random observations by a supervisor on weekly basis. The families were encouraged to come to the study hospital for any illnesses of the child, and the treatment was free of charge to children during the study period.
Severity of pneumonia was categorised using WHO’s IMCI criteria (Box 1). Absence of pronounced wheeze was added to define pneumonia because wheeze could be caused by viral infections and would reduce the specificity of the diagnosis of pneumonia. Respiratory rate was measured twice for one full minute using stopwatches, and axillary temperature was measured twice using electronic thermometers (Thermoval Classic, Paul Hartmann AG, Germany).
Box 1 Definitions of outcomes
Pneumonia: (i) Age-specific tachypnoea (>60/min if <2 months; >50/min if 2–11 months; >40 if 12–24 months) and (ii) absence of wheeze (with or without fever).
Very severe pneumonia: criteria of pneumonia plus at least one of the danger signs (central cyanosis, sever respiratory distress [head nodding, nasal flaring, grunting], inability to drink, convulsions, vomiting) (World Health Organisation, 1995).
Fever: Axillary temperature >37.50 °C (age 1 week–3 months) or >38.0 °C (2–23 months).
Recovery: For two consecutive days, respiratory rate <40/min, no danger signs or subcostal recession, and no fever.
Failure to treat: No reduction in the resting respiratory rate over a 72 h period compared to that detected at enrolment after allowing for a variability of ±5 breaths/min of the baseline respiratory rate (Brooks et al. 2004, 2005).
Repeat episodes of pneumonia: An episode of pneumonia 14 days after the last day of illness of the previous episode of pneumonia.
Sample size and statistical analysis
To detect a 20% difference in the mean duration of pneumonia between the vitamin D arm and placebo arm [5 days vs. 6 days (SD 2)] allowing Type I error of 5% and Type II error of 10%, 137 per group (274 total) were required. Allowing for 37% failure of response to the treatment regime and loss to follow-up during the 3 months, 450 were to be recruited (453 children were actually enrolled).
Given the short period of the study follow-up and single intervention, there was no stopping rule or interim analyses. All children randomised were included in the analysis on an intention-to-treat basis unless the outcome measures were missing (n = 2) or reported to have recovered or lost within 24 h (n = 22). Mean time to recovery for the episode of pneumonia at recruitment was compared for the vitamin D group and placebo group.
For the repeat episodes of pneumonia, the aforementioned 24 children were excluded from analysis as were those lost to follow-up before 10 days post-treatment (n = 14). Time at risk of repeat episodes of pneumonia was calculated from day 11 post-treatment with vitamin D or placebo from the index episode of pneumonia at recruitment to the date of the first or only new episode of pneumonia, for those who experienced a new episode, or the last day of follow-up within the 90-day follow-up period for those who did not have a new episode. There were four children in intervention and four in the placebo arms whose index recovery time was after 10 days and up to 14 days.
Kaplan–Meier plots and log-rank tests were used to compare the time to recover from the index episode of pneumonia between the vitamin D and placebo groups. Incidence rates of pneumonia were calculated by dividing the number of new episodes of pneumonia by total time at risk for all children. Hazard ratios with 95% CIs were obtained with Cox proportional-hazards models to measure time to repeat episodes between treatment groups. The potential confounding of baseline characteristics on treatment effect was also assessed.
Three children died during the 90-day follow-up (Figure 1). No adverse events related to vitamin D3 were observed. The number of children lost to follow-up during the first 10 days of post-treatment follow-up was small and similar between the two groups (Figure 1). There was no statistically significant difference in any of the baseline characteristics between the groups (Table 1).
Table 1. Background and recruitment illness characteristics of study children at enrolment
Vitamin D group N = 225
Placebo group N = 229
*Treatment group – 211 analysed; placebo – 218 analysed.
Mean age in months (SD)
Never breast fed (%)
Mean age in months at weaning (SD)
<20 years (%)
20–34 years (%)
>34 years (%)
Secondary or above (%)
Mean number persons sleeping in the same room with the child (SD)
Recruitment illness characteristics
Severity of pneumonia (very severe excluded)
Mean RR (SD)
Mean temperature (SD)
The mean number of days to recovery from the index episode of pneumonia was the same for both the vitamin D group and the placebo group [4.74 (SD 2.22) vs. 4.98 (SD 2.89); P = 0.17] (Table 2). The risk of children having a repeat episode of pneumonia during the 90-day post-treatment period was significantly lower in the vitamin D group than in the placebo group (RR 0.78; 95% CI 0.64, 0.94, P = 0.01) (Table 3). Children in the vitamin D group survived without experiencing a repeat episode of pneumonia for a longer period than children in the placebo group, for the first or only episode of pneumonia (HR 0.71; 95% CI 0.53–0.95, P = 0.02) (Table 3, Figure 2). There was no confounding effect of baseline measures on risk of repeat pneumonia or time to repeat episode.
Table 2. Outcomes of the index episode of pneumonia by treatment in intention-to-treat analysis
Vitamin D group N = 224*
Placebo group N = 229*
*Treatment group – 211 analysed; placebo – 218 analysed.
Recovered within 24 h of admission or date of recovery not recorded
Mean number of days to recovery* (SD)
Table 3. Risk of repeat episodes of pneumonia during 90 days post recovery from the index episode of pneumonia in intention-to-treat analysis
Vitamin D N = 204
Placebo N = 211
Risk ratio (95% CI)
Children with at least one episode of repeat simple or severe pneumonia
0.78 (0.64, 0.94)
Children with at least one episode of severe pneumonia
1.04 (0.77, 1.41)
Children with two or more episodes of repeat simple or severe pneumonia
1.4 (0.81, 2.48)
Children with at least two episodes of severe pneumonia
1.1 (0.15, 7.57)
Median time to 1st repeat episode
Incidence of at least one episode repeat pneumonia
7 per 1000 child days
10 per 1000 child days
Hazard ratio: 0.71 (0.54, 0.94)
The result of this RCT has shown for the first time that vitamin D supplementation, along with antibiotic treatment, significantly reduces new episodes of pneumonia over a 90-day period. However, this trial did not find any evidence for improved prognosis of the index episode of pneumonia by vitamin D supplementation along with antibiotic treatment. The numbers of severe repeat episodes or more than two repeat episodes did not seem different between the groups, but the study was underpowered to detect the effect of allocation on this outcome. From these results, we conclude that a high-dose supplementation of vitamin D may not immediately influence the recovery from a pneumonia case. Nevertheless, occurrence of pneumonia is a risk factor for the next episode (Lehmann et al. 1991), and children experiencing an episode of pneumonia are likely at a higher risk of repeat pneumonia because of underlying conditions, their socioeconomic or environmental risk factors (Behrman et al. 2003). Thus, preventing repeat episodes of pneumonia would likely improve their health outcomes and the overall burden of disease and deaths from pneumonia.
This was a randomised, well-conducted trial in a population at high risk of vitamin D deficiency. Study outcomes were ascertained by experienced doctors and the loss to follow-up was minimal. One possible source of imprecision in our study is the lack of x-ray confirmation of cases of pneumonia. However, the use of IMCI clinical definitions is comparable with other trials with pneumonia as an outcome in children (World Health Organisation, 1995; Banajeh 1998).
Distinction of viral causes of pneumonia was impossible in this study, but the exclusion of cases with wheeze reduced the chances of including viral respiratory diseases in the index cases. Misclassification of other causes of raised respiratory rate as pneumonia would be distributed equally between the two groups and therefore unlikely to bias the observed reduction in the incidence of repeat episodes of pneumonia.
A few children may have received treatment from health care providers other than the study doctors, such as private medical practitioners, for repeat pneumonia cases. However, the under reporting of cases would be similar between the two groups because the socio-economic characteristics were comparable. The high risk nature of this study population [a socioeconomically deprived population known to have a high rate of vitamin D deficiency (Manaseki-Holland et al. 2008)] means that generalisability of these findings may be limited to those children of similar age with high risk and especially to children who had an episode of pneumonia.
Furthermore, it was not possible to conduct quality control of the vitamin D3 preparation because testing of such samples was too costly and technically difficult for the resources available to this study; a measurement of vitamin D level in the serum achieved as the result of this supplementation (to ensure adequate supplementation or excessively high blood levels) was also not possible; these are the weaknesses of our study. Side effects were not detected, though the number of children and the length of follow-up involved in this study could only detect gross, common and early adverse effects.
Although this is the first trial demonstrating the preventive effect of vitamin D supplementation upon pneumonia infection, it is in harmony with findings that vitamin D can enhance the immune function. The exact mechanisms at this stage can be speculated. There is increasing evidence suggesting that 1,25-dihydroxy vitamin D (1,25(OH)2 D, the biologically active metabolite of vitamin D acting via the VDR, plays an important role in the human innate immune system (Bikle 2008; White 2008). Innate immunity is responsible for host defence against infections through rapid production of antimicrobial proteins (AMPs) such as cathelicidin, after activation of toll like receptors (TLRs) which are responsible for recognition of pathogens in human macrophages, monocytes and epithelial cells. Activation of TLRs also results in increased expression of the lalpha-hydroxylase enzyme (CYP27B1) (Liu et al. 2006) responsible for local conversion of 25 hydroxyvitamin D to 1,25(OH)2 D. This local synthesis of 1,25 (OH)2D further enhances the expression of cathelicidin, thereby potentiating the host’s innate immunity against microbes. These investigators also observed that the induction of TLR-mediated induction of cathelicidin mRNA was blunted in African Americans, who had low serum concentration of 25(OH)D, the substrate necessary for local 1,25(OH)2 D synthesis. Addition of exogenous 25(OH)D to African American sera restored the induction of cathelicidin mRNA.
Altogether the results of these in vitro studies provide a scientific basis for the observations of increased susceptibility to infections in vitamin D-deficient individuals. They also provide an explanation for results for our trial which was undertaken among the children in whom we have previously reported profound and widespread vitamin D deficiency (Manaseki-Holland et al. 2008).
In spite of close monitoring, we observed no side effects from this supplementation. Although vitamin D overload is a theoretical possibility, it can be minimised by using clear guidelines for supplementation and 100 000 iu of vitamin D provide the best protection against vitamin D deficiency and no overload in high-risk 0–9 month infants with normal baseline ranges of vitamin D (Zeghoud et al. 1994). There are no known noticeable side effects from this supplementation. Single intramuscular injection of three times this dose (300 000 iu) was safe and effective in treating nutritional rickets in 6 to 30-month-old children residing in lower socioeconomic regions of sunny Istanbul (Kutluk et al. 2002). Higher than recommended doses of daily supplementation of vitamin D (500–1000 iu/day, adding up to 120 000 iu over 3 months) plus additional vitamin D fortified milk does not induce an overload (Vervel et al. 1997) in infants starting with normal ranges of vitamin D, even when supplementation continued during the summer (in France) and mothers who had antenatal vitamin D supplementation (Zeghoud et al. 1997). Thus, the evidence suggests that the proposed dosing regime is safe and effective.
We used one 3-monthly high-dose since in a pragmatic setting, it was more likely to enable compliance than a daily dose and was effective at maintaining the serum vitamin D level in normal ranges for 2–3 months in other French high-risk infant populations (Vervel et al. 1997). Although it is possible that a higher dose of supplementation may have lasted longer and been more effective in highly deficient individuals, this study was pragmatic, and as many hospitals in developing countries that have a population with high vitamin D deficiency did not have the possibility to test individuals for vitamin D status before supplementation. In this case, and given the lack of published evidence for safety of higher doses in infants, we could not justify giving higher doses to young children from the community diagnosed with pneumonia.
We acknowledge the families who took part in this study. We further thank Dr Abdul Malik Faize who assisted in recruitment, training and management of the project, all the project field staff, in particular Drs Abdul Rasheed Mansoor, Sayed Mujahid Hashimi and Ahmad Fayaz Iqbal, who recruited and clinically followed up the children and the 10 female field workers of the study. Support and guidance of Drs Alawi from the Ministry of Public Health of Afghanistan and Dr Kohdamani, the Director of the Maywand Hospital, were invaluable. We are grateful to Dr Latif Sheikh and his Pharmacy Department colleagues at the Aga Khan University Hospital Karachi who provided the randomised labelled syringe samples of vitamin D and placebo and to Prof Zulfiqar Bhutta for his coordination and facilitation of this arrangement. The study was generously funded by the New Zealand Aid Cooperation.