Objective. To test the hypothesis that high-dose vitamin A supplements will enhance recovery of children hospitalized for the treatment of community-acquired pneumonia.
Design. We conducted a randomized, double-blind, placebo-controlled clinical trial of high-dose vitamin A supplements among children 3 months to 10 years of age (N = 95) admitted to hospital with community-acquired pneumonia in Lima, Peru. Children ≤1 year of age received 100 000 IU of water-miscible vitamin A on admission to the hospital and an additional 50 000 IU the next day. Children >1 year of age received 200 000 IU on admission and 100 000 IU the next day.
Results. Children receiving vitamin A (n = 48) had lower blood oxygen saturation (the mean difference on day 3 in hospital was 1.1%), higher prevalence rates of retractions (37% in the vitamin A group vs 15% in the placebo group on day 3), auscultatory evidence of consolidation (28% in the vitamin A group vs 17% in the placebo group on day 3), and were more likely to require supplemental oxygen (21% in the vitamin A group vs 8% in the placebo group on day 3) than children in the placebo group (n = 47). Adjustment for baseline severity of disease and nutritional status did not alter the association of vitamin A with increased clinical severity, although the difference in blood oxygen saturation was no longer statistically significant. No differences were seen in duration of hospitalization or in chest x-ray changes 14 days after admission. No deaths occurred, and toxicity of vitamin A was not seen.
Conclusions. This study indicates that high-dose vitamin A supplements cause modest adverse effects in children recovering from pneumonia and should not be used therapeutically in such patients unless there is clinical evidence of vitamin A deficiency or concurrent measles infection.
Vitamin A deficiency is an important public health problem in many areas of the world1 that often leads to increased mortality from common childhood infections,2,3particularly in areas with minimal coverage by health services and vaccination programs.4 Vitamin A deficiency presumably leads to increased severity of certain infections because of its central role in regulating the immune response to pathogenic organisms and in repairing protective mucosal epithelium damaged by infection.5 To test the hypothesis that pharmacologic doses of vitamin A might be useful in speeding recovery from measles, Hussey and Klein6 conducted a clinical trial in which two water-miscible, high-dose vitamin A supplements were given to children admitted to hospital with severe cases of measles, which often is complicated by secondary bacterial infections. This supplementation regimen reduced mortality and speeded recovery from postmeasles pneumonia and diarrhea. High-dose vitamin A supplements are now widely recommended for children with measles or complications of measles.7 More recently, a placebo-controlled clinical trial of a single, fat-soluble, high-dose vitamin A supplement among Guatemalan children with community-acquired pneumonia did not show any enhancement in recovery.8 However, in view of the previous success in reducing the duration of postmeasles pneumonia and of work documenting the presence of subclinical vitamin A deficiency in Peru,9 we conducted a randomized, double-blind, placebo-controlled clinical trial using two doses of water-miscible vitamin A among children hospitalized with x-ray-confirmed, community-acquired pneumonia in Lima, Peru. We found that this supplementation regimen had a modest, negative impact on recovery from pneumonia, as described below.
SUBJECTS AND METHODS
Study Population and Recruitment
Children from 3 months to 10 years of age with a principal diagnosis of pneumonia (confirmed by x-ray) admitted as inpatients to the pediatrics ward of Universidad Peruana Cayetano Heredia (UPCH) Hospital from July 4, 1994, through October 31, 1995, were eligible for inclusion in the study. Full-time study personnel included two physicians (L.M.F. and H.H.) and four trained study nurses. At least one study nurse was present on the ward at all times. Before admission, an initial diagnosis of pneumonia was made by an attending physician (not associated with the study) in the outpatient service based on the clinical history and a physical examination (ie, presence of cough, fever, rapid breathing, respiratory distress, and rales and elevated bronchial breath sounds on auscultation). A chest x-ray was performed immediately and read by the attending physician. Children were then admitted to the pediatrics ward, and those with pulmonary infiltrate or consolidation confirmed by x-ray were eligible for recruitment into the study. A purified protein derivative (PPD) skin test was applied at the time of admission to the pediatrics ward and was read 48 hours later. Criteria for exclusion from the study included a previously diagnosed immunodeficiency of any origin, regular use of vitamin A supplements, a weight-for-height below the 70th percentile of the National Center for Health Statistics reference standards, a history of bronchial asthma or repeated bronchospasm, suspicion of clinically active tuberculosis (including a positive PPD skin test with a diameter of >10 mm), and the presence of underlying chronic disease. Informed, written consent was solicited at the time of admission to the ward by one of the study nurses from a parent or guardian of the potential subject after admission criteria (except PPD test reading) had been verified by the nurse. All admission criteria were reviewed and confirmed by the study coordinator, typically within 24 hours.
Forty-seven subjects were admitted to the vitamin A group and 49 to the placebo group. Three subjects in the placebo group subsequently developed a PPD skin test of >10 mm and were withdrawn from the study. These subjects were included in analysis of the admission history and physical examination data, but were excluded from all other analysis. One subject who should have been excluded from the study (because of Downs syndrome) was inadvertently admitted to the placebo group but was subsequently excluded from all data analysis. Thus, admission history and physical examination data from 47 subjects receiving vitamin A and 48 receiving placebo were analyzed (excluding the subject with Down's syndrome) for assessment of the adequacy of the randomization protocol. Clinical outcome data from 47 subjects in the vitamin A group and 45 in the placebo group (excluding the 3 subjects with positive PPD tests) were analyzed to determine the efficacy of the vitamin A supplementation protocol. Complete clinical data were collected on 89 of these 92 subjects. Of the remaining 3, 2 subjects were withdrawn from additional data collection because of complications of their infection at day 10 (placebo) and day 14 (vitamin A). One subject was withdrawn from the placebo group during the first hospital day because of medical complications unrelated to his infection or the study protocol. Data from these subjects were included in all analysis until the time of withdrawal. This research protocol was reviewed and approved by the human subjects committees both at the University of Alabama at Birmingham and at UPCH.
Vitamin A and Placebo
At the time of admission to the study, subjects were randomized to the treatment or placebo group using a list of sequential numbers in which treatment or control had been randomly assigned within blocks of four numbers. Bottles containing the vitamin A and placebo preparations were prenumbered from the same list. The study was double-blind in that neither the investigators nor the subjects had knowledge of which numbers corresponded to treatment or control. Vitamin A and placebo preparations were administered orally. Aquasol A (Astra USA, Westborough, MA), a water-miscible preparation of retinol containing 50 000 IU/mL, was used as the treatment, and an identical placebo syrup was prepared by the University of Alabama at Birmingham Hospital Pharmacy using the formula provided by the manufacturer. The vitamin A and placebo preparations were indistinguishable to a panel of blinded tasters on the basis of appearance, viscosity, color, taste, and smell. Bottles, caps, labels, fluid volumes, and all other aspects of the vitamin A and placebo were identical. Children <1 year of age received 100 000 IU (2 mL) of vitamin A on admission to the study and 50 000 IU (1 mL) on the second day of hospitalization. Children ≥1 year of age received 200 000 IU (4 mL) on the first day and 100 000 IU (2 mL) on the second day. Children in the placebo groups received the appropriate volume of placebo syrup on both days.
Children were monitored closely for signs of toxicity, including nausea, vomiting, headache, dizziness, drowsiness, strabismus, bulging fontanelle, increased cranial circumference, desquamation, and papilledema. Nausea was observed in one child from each group on the first day of the study. Vomiting was seen in one child in the vitamin A group, also on the first day of the study. This child vomited within 1 hour of administration of the syrup, and the treatment was administered again.
Baseline Data Collection
On admission to the study, the following baseline data were collected for each subject from a parent or guardian: age, sex, duration of respiratory symptoms before admission, presence of cough, anorexia, fever in the previous 24 hours, respiratory distress, and a record of antibiotic use before admission. Weight, height or recumbent length, and auscultatory sounds (wheezing and unilateral or bilateral rales or bronchial breath sounds) during a physical examination by a study physician also were recorded. Using baseline data, a presumptive diagnosis of viral (upper respiratory signs or symptoms, wheezing and rales, and hyperinflation and/or interstitial infiltrate on chest x-ray) or bacterial (no wheezing, crepitant rales and/or bronchial breath sounds, and consolidation and/or effusion on chest x-ray) pneumonia was made. A hematocrit and total white blood cell count were also obtained.
After admission to the study, the following data were collected by study nurses every 4 hours: temperature, heart rate, respiratory rate, presence or absence of retractions, occurrence of central cyanosis, percent of blood oxygen saturation by noninvasive oximetry (Nellcor Inc, Pleasanton, CA), and the use of supplemental oxygen (yes or no). The decision to use supplemental oxygen was made by attending physicians, not study personnel, and was based on standard clinical criteria, including blood oxygen saturation levels <90%. After meals, the amount of food consumed from the standard hospital meal was recorded (nothing = 0, <50% of the meal = 0.25, 50% consumed = 0.5, >50% consumed = 0.75, all of the meal consumed = 1). The mother's impression of each subject's appetite (worse = 0, no change = 0.5, better = 1.0) also was recorded. Nurses were specifically trained during a pilot study of 10 patients in which repeated clinical measurements were made on the same subjects. Interobserver variation was not substantially greater than intraobserver variation; thus, data collected by different nurses were judged to be comparable over time. Results of daily auscultatory examinations by a physician also were recorded.
We developed a numerical severity score based on the principal outcome variables to monitor clinical recovery. The calculation of this score is outlined in Table 1.
Antibiotic Treatment Regimens
Antibiotic treatment regimens are standardized within the pediatrics unit at UPCH. Decisions regarding selection of antibiotics and dosage were not controlled by study personnel and were made by attending physicians.
Bacterial and Viral Diagnostics
A blood sample was collected aseptically into sterile blood culture bottles (Fisher Scientific, Atlanta, GA) at the time of admission for culturing and identifying Haemophilus influenza and Streptococcus pneumonia using standard techniques.10 Within 24 hours of admission, a nasopharyngeal lavage specimen was collected. Cells were washed, spun onto a glass slide, and examined for the influenza A; influenza B; parainfluenza types 1, 3, and 4; adenovirus; and measles by immunofluorescent antibody analysis using commercial reagents (Chemicon International, Inc, Temecula, CA).
Statistical analysis was performed using SPSS software version 4.01 (SPSS Inc, Chicago, IL). Kruskal–Wallis one-way analysis of variance (ANOVA) was used for comparing continuous data between two groups. Pearson's χ2 was used to compare categorical data, except when the expected number of observations in a particular cell was less than 5, in which case Fisher's exact test was used. Comparisons of continuous variables (eg, blood oxygen saturation) between groups over time were made by analysis of covariance. Comparisons of discrete variables (eg, presence of retractions) between groups over time were made by logistic regression. In both cases, time since admission and group × time interaction terms were always included in the model. Significant differences indicated in the text for these analyses refer to P < .05 for either the group or group × time variables; if one of these variables is not identified, it was not significant. When other factors (eg, baseline weight for height) were included in the analysis, they are indicated in the text. Duration of hospitalization was evaluated using the Lee–Desu statistic to compare survival experience.
Comparability of the Groups at Baseline
Physical examination and medical history data collected at the time of hospital admission were compared between the placebo and control groups to assess the success of randomization in producing comparable groups with regard to demographic variables, nutritional status, and indicators of disease severity. Nineteen of these variables are shown in Table 2, and no significant differences were seen in 17 of the variables. Statistically significant differences were seen in two variables. Neither group suffered from chronic or acute malnutrition; the mean height-for-age, weight-for-age, and weight-for-height values were all within one z score (ie, standard deviation [SD] unit) of the US reference population, although the vitamin A group had a lower mean weight-for-height score (P = .022) than did the placebo group. Mean serum retinol concentrations were quite low, as expected during a febrile infection,11 and serum C-reactive protein concentrations were quite high. The prevalence or mean values for all of the indicators of disease severity did not differ between the groups. Nor were there differences in the etiology of pneumonia between the groups, the prevalence of antibiotic use before admission (Table2), or in type or duration of antibiotic used (data not shown). Although the mean duration of respiratory symptoms before admission did not differ between groups (5.9 ± 5.3 days for the vitamin A group vs 5.9 ± 2.3 days for the placebo group; P > .05), this variable was not normally distributed. When analyzed as a categorical variable, the distributions differed significantly (Table2), with the vitamin A group having more subjects with both shorter (<1 week) and longer (>2 weeks) duration of symptoms. Admission chest x-rays also were compared, as shown in Table3. No significant differences were found.
At the beginning of the clinical observation period in hospital, before the administration of vitamin A or placebo, there were no significant differences between the groups in the mean values or prevalence rates for the individual clinical indicators of disease severity (Table4). The prevalence rates of central cyanosis and the mean values for respiratory rate and the aggregate clinical severity score (calculated as shown in Table 1) all were marginally higher (0.05 ≤ P < .10) in the vitamin A group than in the placebo group, suggesting that the children in this group tended to have more severe disease at admission.
Antibiotic Treatment of Pneumonia
Antibiotics were prescribed according to standard practices on this unit and were not proscribed by the study protocol. The percentage of children receiving antibiotics did not differ between the vitamin A (39/47, 83%) and placebo groups (38/45, 84%; P = 1.00). Nor did the types of antibiotic given (P= .64) or duration of use (P = .68) differ between the groups. Penicillin was the most common antibiotic used (vitamin A, 34%; placebo, 36%), followed by ampicillin or amoxicillin (vitamin A, 26%; placebo, 29%) and chloramphenicol (vitamin A, 13%; placebo, 16%).
Comparison of Clinical Outcomes During Recovery
We monitored changes in 12 clinical indicators of disease severity during hospitalization. Because treatment effects could change over time, our initial analysis took into account study group (the group term), time in hospital (the time term) and a group × time interaction term. The 12 variables analyzed in this manner included the clinical severity score, its eight individual components (see Table 1), the need for supplemental oxygen, and two auscultatory signs (crepitations and bronchial breathing). Eight of these 12 variables showed significant differences between the treatment groups during hospitalization, all indicating the development of more severe disease in the vitamin A group than in the placebo group. As shown in Fig1, mean blood oxygen saturation during the first 16 hours of hospitalization did not differ between the vitamin A and placebo groups. However, by 24 hours, the mean for the placebo group had increased to a higher level than had the mean for the vitamin A group, and this difference persisted through hospitalization (P = .003 for the group term). Although the absolute differences in mean percent of saturations were small at any given time (∼0.5% to 1%), the corresponding differences seen in other clinical variables suggest that this difference in blood oxygen saturation had clinical ramifications. For example, the prevalence of retractions was ≥60% in both groups on admission and decreased to ≤10% during hospitalization, as shown in Fig 1. However, the rate of decrease was greater in the placebo group, resulting in a significantly lower prevalence of retractions in the placebo group than in the vitamin A group when the curves were compared over time (P < .0005 for group × time term). Similarly, the mean respiratory rate (P < .0005 for the group term) and heart rate (P = .001 for the group term) in the vitamin A group were significantly greater than in the placebo group when compared over time, as was also seen with the aggregate severity score (P = .021 for the group term) (data not shown). The clinical importance of the resulting difference between the two groups also was seen in the resulting need for supplemental oxygen. The percent of subjects receiving oxygen initially was the same in the two groups, but the rate increased rapidly in the vitamin A group, with the result that significantly more children in the vitamin A group than in the placebo group eventually required supplemental oxygen (P < .0005 for the group × time term), as seen in Fig 1. In addition, the prevalence of auscultatory evidence of consolidation initially was the same in the two groups (Fig 2), but the prevalence decreased steadily in the placebo group while staying nearly constant in the vitamin A group, resulting in a significant difference between the groups (P = .0014 for the group × time term).
Although the clinical measures of disease severity showed that vitamin A had adverse effects on recovery, other indicators of disease severity were unaffected by vitamin A treatment. The mean duration of hospitalization did not differ between the vitamin A (5.9 ± 3.5 d) and placebo groups (6.0 ± 4.9; P = .74), nor were differences seen between the groups in degree of inflammation in chest x-rays taken at follow-up (Table 3). Finally, there were no deaths in either group and only two cases (one in each group) of pneumonia did not resolve adequately by the end of the study.
Adjusting Analysis for Baseline Severity of Disease, Age, and Diagnosis
The vitamin A group had a marginally higher clinical severity score at baseline (P = .05), lower weight-for-height z score on admission (P < .05), and different distribution of days with respiratory symptoms before admission (P < .05) than did the placebo group. Thus, we included these continuous variables as covariates in our analysis to adjust for these imbalances. These adjustments did not fundamentally alter the findings described above, although addition of these three variables caused the association between treatment and blood oxygen saturation, and between treatment and heart rate, to lose statistical significance (P > .05 for both group and group × time terms). However, the association of vitamin A treatment with a higher prevalence of retractions (P < .0005 for the group × time term) and greater need for supplemental oxygen (P < .005 for the group term) remained highly significant after adjustment, as did the difference in prevalence of auscultatory evidence of consolidation (P = .0014 for the group term and P = .0001 for the group × time term), indicating that treatment group, rather than baseline disease severity, had the stronger association with these and the other significant outcome variables. Thus, the vitamin A treatment itself, rather than a chance imbalance between the treatment groups in the baseline severity of disease, is the most probable cause of the greater clinical severity of disease seen in these subjects. In addition, inclusion of the subject's age or diagnosis of viral versus bacterial pneumonia as covariates, either alone or with the baseline severity of disease indicators, did not significantly alter these findings described above.
The results of this clinical trial show that high-dose vitamin A supplements are not beneficial in improving recovery from pneumonia in hospitalized children. Furthermore, the study provides clear and consistent evidence that this supplementation regimen produces adverse effects that are of clinical importance. These adverse effects included longer duration of clinical signs and a greater need for supplemental oxygen, which requires more nursing time and higher patient care costs. These adverse effects were not so severe as to require longer hospitalization or produce untoward clinical outcomes or to cause significant differences in chest x-ray findings at the follow-up examination. Although the randomization protocol produced clearly comparable groups at baseline for most variables, there were imbalances between the groups in weight for height, duration of disease before admission, and the baseline clinical severity score. Adjustment for these imbalances did not fundamentally alter the outcome of our analysis and thus does not detract from our conclusions.
Our findings differ from those of other clinical trials of vitamin A in children with pneumonia, which have found no consistent benefit to such supplements but, in contrast to our study, have not found evidence of harm. The first such study published was conducted among 263 Guatemalan children 3 months to 5 years of age hospitalized with community-acquired pneumonia (90% of subjects) or bronchiolitis.8 A single oil-based retinyl palmitate supplement (100 000 IU to those <1 year and 200 000 IU to older children) was given at admission. Similar exclusion criteria and endpoints were used, but no significant differences were seen between the vitamin A and placebo groups in any of these variables or in duration of hospitalization. Two additional placebo-controlled studies using higher doses of oil-based vitamin A (400 000 IU over 2 days in children >1 year of age and half this dose in younger subjects) have been completed recently in Brazil12 and Vietnam.13 In Brazil, 472 inpatients and outpatients 6 months to 5 years of age with a clinical diagnosis of pneumonia were studied. Fever, respiratory rate, and blood oxygen saturation were monitored daily in inpatients and on days 3 and 11 in outpatients. Duration of pneumonia was not affected by vitamin A, but the prevalence of fever was lower at day 3 in the vitamin A group than in the placebo group (16% vs 26%; P = .008). The clinical failure rate for the first-line antibiotic was 29% higher in the placebo group than in the vitamin A group, although the statistical significance of this finding was marginal (P = .054). In Vietnam, 592 inpatients 1 to 59 months of age with a clinical diagnosis of pneumonia were studied. No differences in duration of hospitalization or time to normalization of fever and respiratory rate were seen in the group as a whole or in subgroup analysis based on sex, age, or severity of disease at admission. However, in children with moderate malnutrition duration of hospitalization was shorter in the vitamin A group (6.8 days) than in the placebo group (8.6 days;P = .04). The real significance of this finding is unclear given its borderline statistical significance, particularly in light of the multiple comparisons made in that study. A principal methodology difference between our study and these other pneumonia studies is the type of vitamin A supplement used. We used water-miscible retinol (rather than oil-based retinyl palmitate) both because water-miscible preparations are absorbed more efficiently than oil-based preparations, thus giving higher peak serum concentrations as vitamin A is transported to the liver,14 and because water-miscible supplements were used in the trial of Hussey and Klein,6 which showed the clinical efficacy of such supplements.
Three recent clinical trials of vitamin A in subjects with respiratory syncytial virus (RSV) infection15 also have given mixed results. These studies involved substantially different patient populations than did the pneumonia trials. In general, the pneumonia subjects were older, had fewer underlying chronic health problems, and had pneumonia of multiple etiologies (eg, only 6% of our subjects had confirmed RSV infection). In comparison, the majority of RSV patients admitted to hospital are <6 months of age, may have been born prematurely, and are more likely to have underlying health problems, including chronic heart or lung disease. A small trial (n = 33) among children hospitalized in the United States15 found neither benefit nor detriment to the use of vitamin A, although those receiving vitamin A tended to have longer durations of hospitalization than did the placebo recipients (6.6 days vs 3.5 days; P = .08). Bresee et al16 studied 239 subjects 1 month to 6 years of age in a multicenter trial in the United States. Subjects received a single dose of retinyl palmitate in oil (50 000 IU for those <6 months, 100 000 IU for those 6 to 11 months, and 200 000 IU for those ≥12 months of age). No differences between the treatment and placebo groups were found for most clinical outcomes but, as in the previous study, subjects receiving vitamin A did had a longer duration of hospitalization than did subjects in the placebo group (5.0 days vs 4.4 days; P = .01). The same investigators conducted a parallel trial among 180 RSV-infected subjects in Santiago, Chile, using essentially the same study protocol and supplementation regimen.17This study found “no significant benefit from vitamin A treatment for the overall group in duration of hospitalization, need for supplemental oxygen or time to resolve hypoxemia.” However, these authors did find a more rapid recovery from tachypnea (P = .01) in the subgroup of children in the vitamin A group admitted with blood oxygen saturations <90% (15% and 26% of the placebo and vitamin A groups, respectively). Thus, the principal finding of these RSV trials was that vitamin A supplementation did not hasten recovery from infection. However, in the two trials in the United States,15,16 an increase in duration of hospitalization was seen.
Why do high-dose vitamin A supplements not promote recovery from community-acquired pneumonia (or RSV infection) when they are of benefit in measles? One possible explanation is that the subjects in the measles trial of Hussey and Klein6 may have been more severely vitamin A-deficient than were the subjects in our study (or other pneumonia studies) and thus may have benefitted more from the administration of vitamin A supplements. Although subjects with clinical evidence of vitamin A deficiency were excluded from all of these clinical trials, it is difficult to evaluate vitamin A status biochemically during acute infections, because serum retinol concentrations are depressed by the acute phase response.5,11 However, the mean admission serum retinol concentration in the measles subjects was 0.4 mol/L, slightly higher than that seen in our subjects, although somewhat lower than that seen in other studies.8 Thus, it seems unlikely that a substantial difference in vitamin A status accounts for the different results of these studies. A second possible explanation for the benefit seen in measles that has not been reproduced in pneumonia is simply that measles is different from pneumonia and that subjects recovering from measles benefit from vitamin A supplements whereas those recovering from other infections do not. For example, vitamin A supplements could enhance some aspect of the immune response (eg, antiviral natural killer cell activity), which would be of particular benefit in recovery from measles but of lesser importance in pneumonia and RSV infections.5
Our study raises one additional question: How could high-dose vitamin A supplements increase the severity of pneumonia? Two explanations seem plausible. First, high-dose vitamin A may enhance some aspect of the immune response, which would increase inflammation in the lungs, thereby increasing the severity of disease. High-dose retinoids can enhance some aspects of immunity mediated by Th2 cells and IL-4 in mice.18 Because eosinophil- and IgE-mediated responses are promoted by IL-4 and Th2 cells,19 a boost to pathogen-specific responses involving these effector mechanisms may have transiently increased the severity of lung inflammation in our subjects who received vitamin A. Such a phenomenon also might account for the longer duration of hospitalization seen in subjects with RSV infection who received vitamin A supplements.15,16 A second possibility is that our vitamin A regimen produced toxic side effects. Signs of acute vitamin A toxicity are produced, at least in part, by the damaging effects of retinol on biological membranes.20If our treatment regimen with oral water-miscible retinol produced particularly high serum retinol levels, this may have damaged endothelial cells in alveolar capillaries, particularly at sites of inflammation where increased permeability already would be present because of the inflammation associated with pneumonia. Such damage could result in increased effusion of fluids into alveoli, thus impairing oxygen diffusion into the blood and possibly increasing the prevalence of retractions and the need for supplemental oxygen, as seen in our subjects receiving vitamin A. Such signs of acute toxicity may not have been evident in the previous clinical trials of water-miscible vitamin A, which was given to subjects with measles,6because different clinical variables were monitored and because underlying lung inflammation is less severe in measles than it is in pneumonia.
In conclusion, this study demonstrates that high-dose vitamin A supplements cause modest adverse effects in children recovering from community-acquired pneumonia. Therefore, we recommend that vitamin A supplements not be used therapeutically in such patients to hasten recovery from pneumonia, unless there is clinical evidence of vitamin A deficiency (ie, xerophthalmia) or concurrent measles virus infection. Administration of high-dose vitamin A supplements after recovery from infection to prevent or treat vitamin A deficiency is not contraindicated by this study.
This study was supported by the United States Agency for International Development (PSTC Project No. 12.353).
We thank Anita Colarossi for making the serum retinol measurements, Sharon Blount for assistance with other laboratory assays, Steve Bernhardt for preparation of the placebo syrup, and Astra Pharmaceuticals for donation of the Aquasol A.
- Received September 8, 1997.
- Accepted December 12, 1997.
Reprint requests to (C.B.S.) Ryals 217, Department of International Health, School of Public Health, University of Alabama at Birmingham, Birmingham, AL 35294.
Dr Franchi's current address: Pediatric Pulmonary Service, Instituto Nacional del Niño, Lima, Peru.
- UPCH =
- Universidad Peruana Cayetano Heredia •
- PPD =
- purified protein derivative •
- ANOVA =
- analysis of variance •
- SD =
- standard deviation •
- RSV =
- respiratory syncytial virus
- ↵Underwood BA, Sporn MB, Roberts AB, Goodman DS, eds. Vitamin A in human nutrition: public health considerations. In: Retinoids: Biology Chemistry and Medicine. 2nd ed. New York, NY: Raven Press Ltd; 1994:211–227
- Glasziou PP,
- Mackerras DE
- Ramakrishnan U,
- Latham MC,
- Abel R,
- Frongillo EA Jr.
- Ross AC,
- Stephensen CB
- Salazar-Lindo E,
- Salazar M,
- Alvarez JO
- ↵Balows A, Hausler WJ Jr, Herrmann KL, Isenberg HD, Shadomy HJ.Manual of Clinical Microbiology. 5th ed. Washington, DC: American Society for Microbiology;
- Nacul LC,
- Kirkwood BR,
- Arthur P,
- Morris SS,
- Magalhaes M,
- Fink MCDS
- Johnson EJ,
- Krasinski SD,
- Howard LJ,
- Alger SA,
- Dutta SK,
- Russell RM
- Racke MK,
- Burnett D,
- Pak SH,
- et al.
- ↵Hathcock JN, Hattan DG, Jenkins MY, McDonald JT, Sundaresan PR, Wilkening VL. Evaluation of vitamin A toxicity. Am J Clin Nutr. 1990;183–202
- Copyright © 1998 American Academy of Pediatrics