| Definition |
Confounding variables are factors - such as overcrowding and malnutrition, in the case of childhood respiratory infections - which are (a) known to increase the risk of ALRI and (b) also associated with the type of fuel and stove used in the home. Thus, poor homes are more likely to use smoky biomass stoves than better-off ones, and more likely to suffer overcrowding and malnutrition. If studies do not adequately deal with these confounding factors (which requires that the factors are both measured to begin with, and adjusted for in the analysis), it may be very difficult to know whether an apparent increase in risk of the disease (e.g. ALRI) is due to the smoke, or to one or more of these other factors, or indeed (as is probably most common) a little of both |
