Fungal Contamination in Public Buildings: Health Effects and Investigation Methods

2. Health Effects of Indoor Molds (Continued)

2.4.2.2 Outcome assessment

Objective assessment of health outcomes is another weakness of many epidemiological studies on health effects of mold exposure, since most studies rely on subjective assessments by questionnaires, which once again render the drawing of firm conclusions more difficult. Objective measures of health outcomes do exist, but incorporating them into studies greatly increases study costs.

2.4.2.3 Confounding factors

Damp conditions favourable to mold growth are also favourable to other biological agents known to be allergenic, such as dust mites and gram-negative bacteria.

Unlike mold, bacteria are not visible and, therefore, their presence can be assessed only by air or dust sampling. Therefore, the association observed between mold exposure and allergic responses could be explained in part by confounding bacteria or dust mites being associated with both the exposure to mold and the outcomes considered. This may explain the findings, in some studies (Williamson et al. 1997; Norbäck et al. 1999), of stronger associations between dampness and asthma than between visible mold and asthma. However, in a cross-sectional study where bacterial endotoxins and dust mites were actually measured, controlling for these other allergen levels did not affect the association between indoor mold and respiratory symptoms (Dales and Miller 1999). Moreover, a case-control study revealed a significant association between mold growth and asthma after controlling for visible dampness (Thorn et al. 2001). Also, experiments in animal models showed that mold antigens are able to induce allergic responses in the absence of endotoxin or other biological agents (Alonso et al. 1997, 1998; Cooley et al. 2000).

Chemical exposures may also be confounders in at least one of the studies summarized above. In Smedje and Norbäck's (2001) cohort study, both airborne fungi and formaldehyde were significant risk factors for incident asthma, but could not be included together in multivariate models because of their strong mutual association.

Other potential confounders in respiratory disease epidemiology, such as socio-economic status, smoking and environmental tobacco smoke exposure, have been controlled for in the majority of cross-sectional and case-control studies reviewed, and are therefore unlikely to explain the findings.

2.4.2.4 Bias

There may be a reporting bias in some studies, as there is an increasing awareness in the population that molds are suspected to cause respiratory health effects. People with mold problems may pay more attention to symptoms experienced by their children or themselves. This is likely to have occurred in the Finnish cross-sectional study that found an association between mold and backaches and stomachaches (Pirhonen et al. 1996). As well, people with respiratory health problems may pay more attention to the presence of mold, as physicians investigating asthma or other respiratory diseases commonly ask patients if they have been exposed to mold or dampness, but this bias was eliminated in many studies by having houses inspected by an investigator blind of participants' case or control status. In the Williamson et al. (1997) case-control study, where the possibility of such a bias was reduced by a case or control classification based on hospital records and exposure assessment based on home visits, an odds ratio of 1.7 (but non-significant) was found between severe dampness and asthma.

2.4.2.5 Study design

To our knowledge, only one cohort study was published on health effects of indoor, non-occupational exposure to molds (Smedje and Norbäck 2001). At the time of writing, another cohort study is being conducted in Canada, the Prince Edward Island infant health study. The evidence linking mold to health effects arises mostly from cross-sectional and case-control studies. These two designs are generally considered weaker than cohort studies for investigating the etiology of disease, since it is difficult to ascertain that the suspected cause actually preceded the disease under study. However, though asthma and allergy are chronic conditions, asthma symptoms can improve when exposure to allergens and/or irritants that induce broncho- constriction is removed. A cross sectional or case-control study finding an association between "mold and/or dampness" and chronic wheezing does not demonstrate that mold has caused the onset of asthma, but it may indicate that either mold or dampness induces respiratory symptoms in asthmatics (assuming, of course, that both exposure and outcome assessments are accurate; see previous sections). On the other hand, cohort studies of home indoor environments and respiratory/allergic diseases (preferably with objective assessment of exposure and outcome, such as home inspection and physical assessments) are needed to ascertain the existence of a causal link between mold and respiratory diseases.