Formaldehyde and cancer: a critical review

Abstract

Formaldehyde is a naturally occurring chemical found in every human cell. It has been in widespread use for over a century as a disinfectant and preservative agent, and more recently in a number of industrial products. Animal studies indicate that formaldehyde is a rat carcinogen at high levels (>_ 10 ppm) of exposure. Results for lower levels of exposure show less clear-cut carcinogenic effects, and some species, such as mice and hamsters, appear much less sensitive to any carcinogenic potential of formaldehyde. Epidemiologic studies of the effects of formaldehyde exposure among humans provide inconsistent results. In general, these nonexperimental studies suffer from a number of biases and flaws. The epidemiologic studies fall into three categories: formaldehyde industry workers, case-control studies, and studies of professionals who use formaldehyde. Studies of industry workers with known exposure to formaldehyde report little evidence of an excess cancer risk. Nasopharyngeal cancer, the one cancer considered most strongly linked to formaldehyde among humans, appears after close examination to be likely a result of multiple subgroup analyses and misclassification. The case-control studies usually lack any direct measure of formaldehyde exposure and rely instead on hypothetical exposure based on occupational exposure matrices. Most of these studies, after adjustment for confounding factors, fail to find a significant association with putative formaldehyde exposure. The studies that do report a significant association suffer from methodologic problems limiting their interpretation. The investigations of professionals who use formaldehyde in their work, such as embalmers, pathologists, and anatomists, have the advantage over case-control studies of a much higher likelihood of actual formaldehyde exposure. The findings among these individuals, however, are at odds with those of the other two groups, with excesses of deaths from cancer of the brain and leukemia. The inconsistency between professionals and formaldehyde industry workers in cancer risk patterns suggest that formaldehyde is not the etiologic agent. When the epidemiologic data on formaldehyde and human cancer are examined in light of the widely accepted causal criteria of strength of the association, consistency and specificity of results, dose-response effects, and biologic coherence and plausibility, the studies published so far fail to provide credible causal evidence.