Cooking smoke can increase the risk of tuberculosis

by Vinod K Mishra*, PhD, MPH, Population and Health Studies, East-West Center, 1601 East-West Road, Honolulu, HI 96848-1601, USA. Phone: (808) 944 7452. FAX: (808) 944 7490. Email: [email protected]

Robert D Retherford, PhD, Population and Health Studies, East-West Center, Honolulu, HI

Kirk R Smith, PhD, MPH, School of Public Health, University of California, Berkeley, CA; and Environmental Studies, East-West Center, Honolulu, HI

* for correspondence and reprints of full article

Abstracted from Mishra, Vinod K, Robert D Retherford, and Kirk R Smith. 1999. Biomass cooking fuels and prevalence of tuberculosis in India. International Journal of Infectious Diseases 3(3): 119 129.

Background

In developing countries, air pollution tends to be highest indoors, where the burning of biomass fuels such as wood, animal dung, crop residues, and grasses produces noxious components, including respirable suspended particulates, carbon monoxide, nitrogen oxides, formaldehyde, and polyaromatic hydrocarbons such as benzo(a)pyrene. (1)

According to India's 1992-93 National Family Health Survey (NFHS), about three-quarters of households use unprocessed biomass as their primary fuel for cooking food. Cooking areas in many Indian households tend to be poorly ventilated, and about one-half of all households do not have a separate kitchen. Under these conditions, household members, especially women, tend to have high levels of exposure to cooking smoke. In developing countries such as India, daily air pollution exposures from cooking with biomass typically exceed relevant health-based guidelines by factors of 20 or more (2).

Tuberculosis is a major health problem in India. It is estimated that more than half of India's adult population is infected with the tuberculosis bacterium, Mycobacterium tuberculosis (3). Typically 5-10% of those infected eventually become ill with active tuberculosis. It accounts for nearly 5% of the national burden of disease in India, higher than in any other major world region. Approximately 500,000 persons die from tuberculosis each year in the country. In recent years, the growth of drug-resistant tuberculosis and the rapid spread of HIV/AIDS have contributed to the resurgence of tuberculosis in India and in other parts of the world (4).

Cooking smoke is a known risk factor for a number of respiratory diseases, such as acute respiratory infection, chronic obstructive lung disease, cor pulmonale, and lung cancer. The mechanism by which cooking smoke can increase the risk of tuberculosis is, however, not well understood. It is plausible that cooking smoke increases the risk of tuberculosis by reducing the respiratory system's ability to resist infection by Mycobacterium tuberculosis, or to resist development of active tuberculosis in already infected persons.

Cooking smoke has been shown to weaken the immune system in laboratory animals. Benzo(a)pyrene, a known carcinogen, is found in large quantities in cooking smoke from biomass fuels. There is evidence that exposure to benzo(a)pyrene can weaken the immune system in both animals and humans. Cooking smoke also tends to increase coughing, which can contribute to the spread of infection. It should also be noted that treatments of chronic respiratory diseases brought on by cooking smoke include certain chemical agents, such as corticosteroids, which are known to weaken the immune system and may increase the likelihood that an inactive tuberculosis infection, if present, will become active.

This study examines the relationship between cooking smoke from biomass fuel use and the prevalence of active tuberculosis among persons age 20 years and older in India, after statistically controlling for the effects of several variables which may bias the effect of cooking smoke (Figure 1). The analysis focuses on persons age 20 years and older because in India the prevalence of tuberculosis is relatively low below this age.

Figure 1: Prevalence of active tuberculosis by age, India 1992-93

Method

India's National Family Health Survey (NFHS), conducted during 1992-93, provided the data for this study. The NFHS collected a range of social and health information from a sample of 88,562 households, comprising 514,827 persons, of whom 260,162 were over 20 years of age. The NFHS asked several questions on the current health status of household members, including the prevalence of tuberculosis within the household. The household head or other knowledgeable adult in the household reported whether or not each household member currently had active tuberculosis. No effort was made to clinically test for tuberculosis.

Exposure to cooking smoke is measured indirectly by the main type of cooking fuel used by the household. The nine types of fuels that were asked about are grouped into two categories - biomass fuels (wood or dung) and a residual category of cleaner fuels (charcoal, coal/coke/lignite, kerosene, electricity, petroleum gas, or biogas). The small category of 'other fuels' is excluded from the analysis.

Because several demographic and socio-economic factors may also affect the prevalence of tuberculosis, the independent effect of cooking fuel on the prevalence of active tuberculosis is estimated after statistically controlling for such factors. The control variables included in the analysis are: age, sex, education, religion, membership in a scheduled caste or scheduled tribe (groups that the Indian government defines as disadvantaged), urban/rural residence, geographic region, house type (indicating quality of construction of roof, walls, and floor), crowding (measured by number of persons per room in the household), and availability of a separate kitchen in the house. These variables are held constant when estimating the independent (adjusted) effect of cooking fuel on the prevalence of active tuberculosis.

Effects of cooking smoke

The analysis shows that a household's use of a biomass cooking fuel has a strong effect on the prevalence of active tuberculosis among adult household members (Figure 2). When measured without controls, the prevalence of active tuberculosis is 3.6 times higher among adults in households that use biomass fuels (1,046 per 100,000) than among adults in households that use cleaner fuels (296 per 100,000). Adjusting for the ten control variables reduces this difference somewhat, but the independent effect of cooking fuel remains strong. After adjusting for the other variables, the prevalence rate of active tuberculosis is 2.6 times higher among adults in households that use biomass fuels (969 per 100,000) than among adults in households that use cleaner fuels (378 per 100,000).

Figure 2: Unadjusted and adjusted prevalence of active tuberculosis by type of cooking fuel: person aged 20 years and older, India 1992-93

The effect is greater for women than for men (Figure 3). For men, the prevalence rate is 2.4 times higher among those in bio-mass-fuel-using households than among those in households using cleaner fuels. For women, the prevalence rate is 2.7 times higher among those in biomass-fuel-using households than among those in households using cleaner fuels. One expects the effect of cooking smoke from biomass fuels to be greater for women, because women are more exposed than men to cooking smoke (Figure 4).

Figure 3: Adjusted prevalence of active tuberculosis by type of cooking fuel and gender: persons aged 20 and older, India 1992-93

Figure 4: Women are more exposed than men to cooking smoke as women tend to be indoors more than men

India E1.02
IT/Rita Nandu

The effect is also greater in rural areas than in urban areas (Figure 5). In urban areas, the prevalence rate is 2.3 times higher among those in biomass-fuel-using households than among those in households using cleaner fuels. In rural areas, the prevalence rate is 2.6 times higher among those in biomass-fuel-using households than among those in households using cleaner fuels. One expects the effect to be higher in rural areas, because medical services to treat tuberculosis are less widely available in rural areas.

Figure 5: Adjusted prevalence of active tuberculosis by type of cooking fuel and residence: persons aged 20 years and older, India 1992-93

Proportion of all tuberculosis in India attributable to cooking smoke

In addition to estimating the effects of smoke from biomass fuels on the prevalence of active tuberculosis, one can estimate the proportion of tuberculosis prevalence that is due to smoke from biomass fuels. This proportion depends not only on the effect of biomass smoke on tuberculosis but also on the extent to which biomass fuels are used in the population.

This proportion can be interpreted as the reduction in tuberculosis prevalence that would occur if everyone were to use cleaner fuels instead of biomass fuels with no other changes, for example in income or nutrition. It indicates (after appropriately including the control variables) that the prevalence of active tuberculosis among persons age 20 and older in India could be reduced by 51% if everyone were to use cleaner fuels.

The proportion of active tuberculosis that can be attributed to cooking smoke is much higher in rural areas (59%) than in urban areas (23%). This urban/rural difference stems from two factors. First, the effect of biomass fuel use on tuberculosis prevalence is stronger in rural areas than in urban areas (as seen earlier in Figure 5), and, second, the proportion of households using biomass fuels is much higher in rural areas (93%) than in urban areas (32%).

Conclusions

The finding that the prevalence of active tuberculosis among persons age 20 and older in India could be reduced by about half if everyone were to use cleaner fuels instead of biomass fuels may be a finding of enormous public health importance. It suggests that tuberculosis prevalence could be reduced substantially in India-and probably in many other developing countries - by lowering exposure to cooking smoke from biomass fuels.

A large proportion of Indian households, however, cannot afford to buy cleaner cooking fuels. Although a long-run policy of reducing exposure to biomass cooking smoke should be followed, a more feasible policy in the short run would be to strengthen programmes that promote improved cookstoves that burn biomass fuels more efficiently and that produce less smoke indoors.

Further research is needed to validate the findings of this study. The use of different types of fuels is not an ideal measure of exposure to cooking smoke, and reports of tuberculosis by household heads or other informants are not as accurate as clinical tests. Studies are needed that are based on better measures of these key variables.

Vinod Mishra works on population and health problems in Asia, population and environment relationships, and environmental impacts on health.

Robert D Retherford is Coordinator, Population and Health Studies, at the East-West Center, Honolulu, Hawaii USA. He conducts research on a wide variety of population and health topics, focusing especially on India and Japan.

Kirk R Smith, Professor and Chair of Environmental Health Sciences at the University of California Berkeley and Senior Fellow at the East-West Center Honolulu, has worked on household fuel issues in developing countries since 1978.

References

1. Smith KR. Indoor Air Pollution in India. Natl Med J India. 1996; 9(3): 103-4.

2. Smith KR, Liu Y. Indoor Air Pollution in Developing Countries. Epidemiol Lung Cancer. 1994; 74: 151-84.

3. WHO (World Health Organization). WHO Report on the Tuberculosis Epidemic 1997. Geneva: World Health Organization, 1997.

4. Piot P. Tuberculosis and AIDS: The Dual Epidemic. UNAIDS: Points of View. 1997. Geneva: UNAIDS.