Some limitations of aggregate exposure metrics

Our plain-English summary of Some limitations of aggregate exposure metrics, by L.A. Cox, D.A. Popken (Risk Analysis, 2007).

The study

• Authors: L.A. Cox, D.A. Popken
• Journal: Risk Analysis
• Published: 2007;27(2):439-45
• Cited by: 6 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Aggregate exposure metrics based on sums or weighted averages of component exposures are widely used in risk assessments of complex mixtures, such as asbestos-associated dusts and fibers. Allowed exposure levels based on total particle or fiber counts and estimated ambient concentrations of such mixtures may be used to make costly risk-management decisions intended to protect human health and to remediate hazardous environments. We show that, in general, aggregate exposure information alone may be inherently unable to guide rational risk-management decisions when the components of the mixture differ significantly in potency and when the percentage compositions of the mixture exposures differ significantly across locations. Under these conditions, which are not uncommon in practice, aggregate exposure metrics may be "worse than useless," in that risk-management decisions based on them are less effective than decisions that ignore the aggregate exposure information and select risk-management actions at random. The potential practical significance of these results is illustrated by a case study of 27 exposure scenarios in El Dorado Hills, California, where applying an aggregate unit risk factor (from EPA's IRIS database) to aggregate exposure metrics produces average risk estimates about 25 times greater - and of uncertain predictive validity - compared to risk estimates based on specific components of the mixture that have been hypothesized to pose risks of human lung cancer and mesothelioma.

In plain English

This article was a mathematical analysis, with an illustrative case study, of 'aggregate' exposure measures that add different types of particles or fibers into a single count. Using worked examples and two theorems, the authors argued that when the parts of a mixture differ in potency, aggregate exposure alone cannot show which of two sites has the higher risk. They also argued that cleanup rules based on it can do worse than picking sites at random. In a hypothetical four-site example, cleaning the site with the highest aggregate exposure removed a risk of 6, compared with an expected 15 for a random choice. The case study used EPA Region IX data on 27 dust exposure scenarios in El Dorado Hills, California. It compared an EPA approach based on phase contrast microscopy equivalent (PCME) fibers with a Berman and Crump (BC) approach that counts only 'long protocol structures'. Only five of 4,306 fibers were long protocol structures. BC-based risks averaged 4.1% of EPA-based risks. The authors concluded that aggregate metrics may be 'worse than useless' for guiding risk-management decisions.

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: Mathematical (theoretical) analysis with hypothetical examples, illustrated by a simulation-based case study using existing EPA exposure data
• Who was studied: 27 dust exposure scenarios in the El Dorado Hills region of California, studied by EPA Region IX (data reported in Ladd, 2005); the scenarios were activities such as children's baseball, trail jogging and playground use
• Exposure: Airborne asbestos fibers counted two ways. The EPA index counts PCME fibers: length greater than 5 µm, width from 0.25 to 3 µm, aspect ratio greater than 3. The Berman and Crump index counts only long protocol structures: length greater than 10 µm, width less than 0.40 µm. No specific fiber types are named.
• Main results:
  • Hypothetical four-site example: cleaning the one site with the highest aggregate exposure removed a risk of 6, compared with an expected risk removed of 15 when a site was picked at random
  • Theorem 1: from aggregate exposure information alone, when different detailed exposure patterns with different risks can share the same aggregate value, it is impossible to determine which of two sites has the higher risk
  • In the EPA data set for El Dorado Hills, only five fibers out of 4,306 were long protocol structures
  • Across the El Dorado Hills scenarios, risks under the Berman and Crump (BC) assumptions averaged only 4.1% of the risks under the EPA assumptions, about 25-fold lower
  • Scenario-specific risks were not significantly different from zero for most scenarios under the BC assumptions, but could be unacceptably high (for example, greater than 1E-6) under the EPA assumptions
• The authors' conclusion: The authors concluded that total or aggregate exposure metrics may give no useful information about where risks are highest when particle types differ greatly in potency. They concluded such metrics may be 'worse than useless' (worse than random selection) for guiding risk-management decisions, and that component-specific exposure and potency information is the only way to tell whether this is the case.
• Limitations the authors state:
  • The authors noted that the lower BC risk estimates at El Dorado Hills do not carry over to other regions: elsewhere the BC assumptions do not necessarily give lower risks than EPA's

Topics

Medical terms in this study, explained in our Lexicon: El Dorado Hills, Lung Cancer, Mesothelioma.

Related studies

• Stoking Ground Zero health worries — R. Harris (2002)
• Regulation of priority carcinogens and reproductive or developmental toxicants — K. Hooper, J. LaDou, et al. (1992)
• Cytotoxicity assessment of some carbon nanotubes and related carbon nanoparticle aggregates and the implications for anthropogenic carbon nanotube aggregates in the environment — L.E. Murr, K.M. Garza, et al. (2005)
• Immunotoxic effects of chemicals: A matrix for occupational and environmental epidemiological studies — A. Veraldi, A.S. Costantini, et al. (2006)
• Environmental aftermath — L. Claudio (2001)
• Commentary: environmental disease--a preventable epidemic — P.J. Landrigan (1992)

Browse more research in Risk Analysis, or return to medical research.