Vitamin A status and therapy in childhood pulmonary tuberculosis

Low plasma vitamin A levels (mean, 18.1 +/- 10.3 micrograms/dl, 62% below normal) were demonstrated in South African children with pulmonary tuberculosis. More extensive or severe disease (e.g., additional extrapulmonary tuberculosis) and low levels of retinol binding protein, prealbumin, and albumin were associated with low vitamin A levels. High-dose vitamin A therapy had no effect on disease outcome.

Vitamin A status, other risk factors and acute respiratory infection morbidity in children

OBJECTIVE

This study evaluated the association between vitamin A status and the severity of acute respiratory infections (ARIs) in children, controlling for the influence of other known ARI risk factors.

DESIGN

Case control study.

SETTING

Ambulatory and hospital-based study.

PATIENTS

Severe cases (N = 35) were children with ARI who were admitted to hospital for inpatient treatment, while mild cases (N = 32) were children with ARI who were treated as outpatients. The control group (N = 54) was selected from children with non-infectious diseases attending the outpatient department. Cases and controls were matched for age and area of residence.

MAIN OUTCOME MEASURES

Serum vitamin A levels and analysis of ARI risk factors.

RESULTS

The mean (SD) vitamin A levels were 22.09 (7.27) micrograms/dl for the controls, 20.27 (11.11) micrograms/dl for the mild cases and 13.79 (7.60) micrograms/dl for the severe cases. All pairwise comparisons of levels of the three patient groups achieved statistical significance-severe and mild (P < 0.01), severe and control (P < 0.001) and mild and control (P = 0.03). After vitamin A levels were dichotomised, the odds ratios (and 95% confidence intervals) for severe versus mild cases were 2.1 (0.8-5.6), for mild versus controls 2.9 (0.8-10.5) and for severe versus controls 6.0 (2.0-19.4). A chi 2 for trend across the three groups was 13.2 (P = 0.001). Risk factors significantly associated with disease status included a history of hospital admission in the preceding 6 months, absence of a clinic card, poor housing and lack of electricity for indoor fuel use. Factors associated with poor vitamin A status included low weight for age, previous diarrhoeal disease and poor housing. Vitamin A status was independently associated with disease status in logistic regression modelling.

CONCLUSION

Vitamin A status has a strong association with severity of infection. The gradient of that association suggests a dose-response effect. The multifactorial nature of ARI severity and vitamin A status highlights the need for a comprehensive approach to public health programmes to address ARI. The role of vitamin A supplementation for at-risk groups is supported by this study, but needs to be clearly defined within a broader approach to health.