Estimating the induction period of pleural mesothelioma from aggregate data on asbestos consumption

Our plain-English summary of Estimating the induction period of pleural mesothelioma from aggregate data on asbestos consumption, by M. Nurminen, A. Karjalainen, K. Takahashi (Journal of Occupational and Environmental Medicine, 2003).

The study

• Authors: M. Nurminen, A. Karjalainen, K. Takahashi
• Journal: Journal of Occupational and Environmental Medicine
• Published: 2003;45(10):1107-15
• Cited by: 12 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

This study aimed to estimate the induction period from causal action of asbestos exposure to the manifestation of mesothelioma. We included the 9 countries for which we could find published aggregate data on the use of raw asbestos for a relevant time period. We extracted the annual numbers of cases of pleural cancer among men from the World Health Organization mortality database for those years using the International Classification of Diseases, 9th revision, classification. For the Scandinavian countries, we used published national cancer incidence data. In autoregressive Poisson regression modeling, we invoked different time lags of the mean annual use of asbestos to specify which time span produced the best correlation between the 2 time series. The ecologic analysis suggested that the most probable estimate for the mean induction period (use versus morbidity at society level) is approximately 25 years.

In plain English

This was an ecologic study that compared national totals rather than individual people. The authors used published figures on raw asbestos use in nine countries (Denmark, Finland, France, Germany, Great Britain, Italy, Japan, New Zealand and Sweden). They paired these with counts of pleural cancer among men aged 15 and over, taken from the World Health Organization mortality database or, for Denmark, Finland and Sweden, from national cancer incidence data. Both sets of figures were grouped into five-year periods. Pleural cancer was used as a stand-in for pleural mesothelioma. The authors tested delays of 20 to 50 years between asbestos use and disease. Delays of 25, 30 and 35 years fitted best. In models that also accounted for calendar year, the link was statistically significant only for the 25-year delay. The authors concluded that the most likely induction period, from the causal action of asbestos to disease at the level of society, was approximately 25 years, with a possible range of 25 to 35 years. They noted large variation around their models and the known limitations of ecologic studies.

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: Ecologic study of national aggregate time-series data, analysed with autoregressive Poisson regression testing different time lags between asbestos use and pleural cancer
• Who was studied: Men aged 15 years and older in 9 countries: Denmark, Finland, France, Germany, Great Britain, Italy, Japan, New Zealand and Sweden. Pleural cancer counts (ICD-9 code 163) came from the WHO mortality database, and from national cancer incidence data for Denmark, Finland and Sweden. Both the asbestos-use and the pleural cancer data were grouped into 5-year periods. The time series covered 1900 through 1995 for asbestos use and 1950 through 1995 for incidence. Australia was excluded because its consumption data were grouped by decade.
• Exposure: Total annual national consumption of raw asbestos (kilogram per adult man), lagged by 20, 25, 30, 35, 40, 45 and 50 years. Fiber types were not separated, and occupation or industry was not restricted.
• Main results:
  • In models including calendar year, the coefficient for asbestos use was positive (indicating excess risk) for induction periods of 25, 30 and 35 years. It was statistically significant only for the 25-year period, with a P value printed as 0.013; the comparison symbol is missing in the text.
  • In a repeat analysis using the same number of observations for every lag, the 25-, 30- and 35-year induction periods again gave the best fits, and the risk parameter was statistically significant only for 25 years (P value printed as 0.049).
  • In models with time-lagged asbestos use as the only predictor, the 25-, 30- and 35-year induction periods were the best predictors, with P values of 0.006, 0.059 and 0.044 respectively.
  • Pleural cancer rates were still rising linearly in 1990 through 1995 in all countries except Denmark, Germany and Sweden, and the fitted model showed no sign that incidence had leveled off during the study period.
  • Applied to Australian asbestos consumption in the 1960s (data not used to build the model), the model predicted approximately 209 annual pleural cancer cases in Australian men 30 years later. Assuming a 1.6:1 ratio as in Britain, this corresponds to about 334 mesotheliomas, compared with approximately 300 annual notifications of male mesothelioma to the Australian Mesothelioma Register in 1990 through 1999. The authors noted that this point estimate is not as accurate as it would appear.
• The authors' conclusion: The authors concluded that the best estimate of the induction period of pleural mesothelioma at the society level is approximately 25 years, with a possible range of 25 to 35 years, and that 25 years (the lower limit) is the best estimate. They noted that this is 10 years shorter than the Australian register-based estimate of 35 years, a difference they attributed probably to differing definitions (time since causal action versus first exposure) and types of data.
• Limitations the authors state:
  • Ecologic studies have well-known limitations.
  • Pleural cancer was used as a surrogate for mesothelioma. Mesothelioma diagnosis is often difficult and misdiagnosis was common in the past.
  • The coding practice for pleural cancer may have been different in Japan, where incidence was much lower.
  • Only data on raw asbestos use were available. Imported asbestos products and exposure from building renovation, maintenance and demolition were not captured.
  • No incidence data corresponded to asbestos use in the early decades of the 1900s, and early asbestos-use data were likely less reliable, so the model's intercept requires caution.
  • There was rather large variation around the fitted models, and observed values deviated considerably from expected ones.
  • The induction period could be modified by asbestos type or exposure intensity, which the total-use data could not separate.

Topics

The National Library of Medicine indexes this study under subjects our site covers: Mesothelioma.

Medical terms in this study, explained in our Lexicon: Pleural Mesothelioma, Asbestos Exposure, Mesothelioma.

Related studies

• Predictions of mortality from pleural mesothelioma in Italy: a model based on asbestos consumption figures supports results from age-period-cohort models — A. Marinaccio, F. Montanaro, et al. (2005)
• Some limitations of aggregate exposure metrics — L.A. Cox, D.A. Popken (2007)
• Asbestos Consumption and Malignant Mesothelioma Mortality Trends in the Major User Countries — C. Gariazzo, A. Gasparrini, et al. (2023)
• Estimating the incidence of malignant mesothelioma in Vietnam: a pilot descriptive cancer registration study — M.J. Soeberg, M.A. Luong, et al. (2016)
• Ecological association between asbestos-related diseases and historical asbestos consumption: an international analysis — R.T. Lin, K. Takahashi, et al. (2007)
• Different measures of asbestos exposure in estimating risk of lung cancer and mesothelioma among construction workers — K. Koskinen, E. Pukkala, et al. (2002)

Browse more research in Journal of Occupational and Environmental Medicine, or return to medical research.