Epidemiological studies on ingested asbestos

Our plain-English summary of Epidemiological studies on ingested asbestos, by J. Hopkins (Food and Chemical Toxicology, 1984).

The study

• Author: J. Hopkins
• Journal: Food and Chemical Toxicology
• Published: 1984;22(2):179-8l
• Cited by: 1 article, as counted by NIH iCite
• Read it: PubMed · Publisher

In plain English

This short review, written for an information section by an author at BIBRA, looked at epidemiological studies of cancer in US communities whose drinking water contained chrysotile asbestos. In Connecticut, where water ran through asbestos-cement pipes, Harrington et al. found no evidence of an effect on gastric, colonic and rectal cancers. Meigs et al. found scattered significant associations with no consistent pattern; the only association seen in both time periods was pancreatic cancer in males. In the Puget Sound region, Polissar et al. reported more than 300 odds ratios, and no more were significant than chance would predict. In the San Francisco Bay region, Conforti and colleagues reported significant associations (P < 0.01) with cancers including the oesophagus, stomach and pancreas in white males and white females. Other researchers criticised how that work handled confounding factors. The author judged the data preliminary and pointed to uncertain exposure estimates. The author concluded that waterborne asbestos remained "an unconvicted risk" and that confirmatory studies were needed.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Review of epidemiological studies of cancer and asbestos in drinking water
• Who was studied: Populations covered by published studies in three US areas: Connecticut, Puget Sound (western Washington) and the San Francisco Bay region. Earlier work in Duluth, Minnesota is also mentioned.
• Exposure: Chrysotile fibres in drinking water. In Connecticut they came from asbestos-cement pipes (up to about a million fibres per litre). In San Francisco Bay they came from serpentine rock (up to 36 million fibres per litre). In Puget Sound they came from the Sultan River (mean 206 million fibres per litre). The Duluth water contained amphibole-type fibres.
• Main results:
  • Connecticut (Harrington et al.): no evidence that asbestos exposure influenced the incidence of gastric, colonic and rectal cancers from 1935 to 1973.
  • Connecticut (Meigs et al., 169 townships, 1955-1974): several statistically significant associations between cancer sites and asbestos-related measures, but no consistent pattern.
  • Connecticut (Meigs et al.): only pancreatic cancer in males showed positive associations in both analysis periods, and different asbestos indicators were involved in each period.
  • Connecticut (Meigs et al.): the authors of that study estimated that, for each sex and decade, they had a more than 90% probability of detecting a 50% additional total cancer risk in the highest- versus lowest-asbestos towns.
  • Puget Sound (Polissar et al.): of more than 300 odds ratios, about half were above unity, and no more differed significantly from unity at the 5% level than chance would predict.
  • Puget Sound (Polissar et al.): there was a nearly consistent positive association with tumours of the small intestine in both sexes and of the thyroid, eye, testes and prostate in males. Odds ratios for brain tumours and leukaemia were consistently below one.
  • San Francisco Bay (Conforti et al., 400 census districts, population 3 million, over 50,000 cancer cases, 1969-1974): significant relationships (P < 0.01) between asbestos content and cancer of the digestive tract, oesophagus, stomach and pancreas in white males. In white females, significant (P < 0.01) associations were found with cancer of the oesophagus, stomach, digestive organs and pancreas.
  • Critics of the San Francisco Bay work said its methods did not fully account for confounders such as social class or occupation. The original authors accepted this criticism as having some validity. Another critic suggested that population density could explain the pattern.
• The authors' conclusion: The author concluded that the current methods could only clearly reveal a powerful carcinogenic effect, and that by that standard asbestos in water supplies remained an unconvicted risk. The author also stated that the Conforti group's findings would need confirmation before ingested asbestos could seriously be considered demonstrably carcinogenic in humans.
• Limitations the authors state:
  • The author stated that confidence in the accuracy or relevance of the asbestos exposure estimates in these studies must be lower than confidence in the cancer registry data.
  • The author stated that asbestos measured in the 1970s may not reflect exposure over the previous 20-30 years.
  • The author stated that the studies assumed everyone in an area was exposed to the same asbestos level, which may misclassify individuals' exposure. Case-control studies would be needed to reduce this problem.

Related studies

• Distribution of chemical carcinogens in aquatic environments — H.F. Kraybill (1976)
• Ingested mineral fibers: elimination in human urine — P.M. Cook, G.F. Olson (1979)
• The contamination of Lake Superior with amphibole gangue minerals — A.M. Langer, C.M. Maggiore, et al. (1979)
• Asbestos-contaminated drinking water: its impact on household air — J.S. Webber, S. Syrotynski, et al. (1988)
• Asbestos in public water supplies: discussion of future problems — J.P. Hills (1979)
• Epidemiological Research on Occupational and Environmental Carcinogens — C. Ledda (2021)

Browse more research in Food and Chemical Toxicology, or return to medical research.