FRANK MIRER, PHD, CIH, is a professor in the CUNY School
of Public Health in New York. He can be reached at (212) 396-7782 or
Getting Particular
What IHs Can Learn from Environmental Studies BY FRANK MIRER
The toxic potential and potency of many substances of concern in the general environment were first observed in the occupational environment. The formaldehyde off-gassing from building products—maybe in your home—is now “known” to cause leukemia, based on occupational studies. The diesel particulate matter you breathe in the street is “known” to cause lung cancer based on occupational studies, with potency estimates of risk calculated from these studies. I’m not aware of exposures for which the toxic potential in the occupational environment was identified from studies in the community. The exception is particulate matter. It’s ti​me the occupational health professions paid attention to “recent” evidence (since 1995) regarding particulate hazards and drastically reduced the OELs for particulate not otherwise classified (PNOC), perhaps limited to poorly soluble particulates. I’m suggesting that a precautionary evaluation criterion for a hazardous exposure to any respirable particulate should be closer to 100 µg/m3 than the 5000 µg/m3 PEL value it has been stuck at since 1971 or even the 1000 µg/m3 limit proposed in a commentary in the July 2013 Annals of Occupational Hygiene. Above 100 µg/m3, practical engineering controls should be installed to reduce exposure, and respiratory conditions in overexposed workers should be expected to arise from the occupational exposure. This conclusion requires connecting a few dots. The most recent authoritative evidentiary shoe to drop on particulates was IARC Monograph 109, which classified outdoor air pollution as “known to be carcinogenic to humans,” causing lung cancer, based on many cohort and case control studies. The working group noted that virtually all of the studies were done in areas where annual average levels of PM2.5 range from about 10 to 30 µg/m3, which represents approximately the lower third of exposures worldwide. Nevertheless, increased risk of lung cancer was observed even in those areas where PM2.5 concentrations are less than the current health-based guidelines. As of mid-June the full text of the monograph had not been posted, but an e-book with extensive background data is available. The evolution of understanding of particulate health effects was described by Doug Dockery in the April 2009 Annals of Epidemiology. Dockery reports a 1979 issue of the American Journal of Epidemiology devoted to a review sponsored by the American Iron and Steel Institute in which leading epidemiologists reported there was “no reasonable evidence” to support a particulate standard under the Clean Air Act. By 1995, as EPA conducted legally required periodic reviews of the National Ambient Air Quality Standards (NAAQS), two key studies by Dockery and coworkers emerged. These found not only increasing respiratory disease and lung cancer with increasing community particle levels, but also increasing cardiovascular disease. The association with cardiovascular disease, which very much increases the burden of disease preventable by controlling exposure, was affirmed by the American Heart Association. The studies were attacked by industry forces, which demanded access to the raw data. Senator Richard Shelb​y of Alabama, at the request of industry, added the Shelby Amendment to the federal budget, subjecting investigators with government funding in ‘‘institutions of higher education, hospitals, and other non-profit organizations’’ (but not private companies) to the Freedom of Information Act. Eventually, the Health Effects Institute (HEI), a non-profit corporation funded by EPA and the motor vehicle industry, acted as a mediator. EPA proceeded to promulgate, in 1997, an NAAQS of 15 µg/m3 annual average for PM2.5 and below, and 65 µg/m3 for a 24-hour average. In 2006, the standard for 24-hour PM2.5 was reduced to 35 µg/m3, and in 2012, the NAAQS for annual av​erage exposure to PM2.5 was reduced to 12 µg/m3. A massive compendium of scientific data behind these numbers is the EPA Integrated Science Assessment. By contrast, the OSHA PEL for respirable particulate of 5000 µg/m3 was adopted in 1971, from the then-TLV for nuisance aerosol, and remains in effect today. ACGIH later adopted a TLV for Particles Not Otherwise Regulated (PNOR) of 3000 µg/m3 for the respirable fraction. These values are thinly documented. DERIVING OELS FROM NAAQS The community exposure limit differs from the OSHA-ACGIH OELs in time weighting and particle size. While NAAQS limits are intended to protect “sensitive” populations, it’s not clear that working-age adults should be exposed to more. The NAAQS is derived from a large data set of health effects arising from a particular particle composition, but it’s considerably more robust, sensitive, and specific than occupational studies because of the large populations involved and long duration of observation. Equating the limits by exposure time is straightforward. The NAAQS applies to 24/7/365 exposures over 70 years compared to the occupational standard of 8/5/250 over 45 years. Factoring work ventilation rates (instead of the average daily ventilation rates to which the NAAQS apply) and duration of exposure yields an occupational equivalent of the PM2.5 annual standard of approximately 50 µg/m3, with a daily limit of approximately 70 µg/m3. Comparison of occupational exposures to these limits could be assessed by using a PM2.5 impactor rather than the traditional respirable cyclone collecting at PM4. We could simply apply the time-modified PM2.5 levels from NAAQS. Assessment of PM2.5 by direct reading instruments and higher volume sampling pumps would be substantially more convenient than respirable sampling. Relating size distribution of respirable particulate to measurements of PM2.5 and below is variable. Respirable sampling (PM4) is intended to correspond to deposition of particles in the pulmonary system. Environmental sampling recognizes that outdoor aerosols consist of a mixture of mechanically generated (crustal) particles, which distribute around 10 microns, and combustion-derived aerosols distributed around 1 micron. A PM2.5 (and below) sampler somewhat separates these two components, although the low end of the mechanical distribution will be captured with PM2.5, and the fractions by size will vary with conditions and source of the aerosol. Thus, for a welding aerosol, side-by-side respirable and PM2.5 would be expected to be close to equal, while for dust on a demolition site respirable sampling results could be materially higher than side-by-side PM2.5. Accounting for particle composition is still more complex. First, the PM2.5 outdoor aerosols upon which the NAAQS and IARC assessments are based are mostly atmospheric and combustion-derived particles distributed around 1 micron in size. There is evidence that these smaller particles are more potent (more effect per milligram), but this is not proven. Second, outdoor PM2.5 aerosols are about half ammonium sulfate and nitrate, 40 percent carbonaceous (elemental and organic), and the balance crustal. Generally, insoluble or poorly soluble particles are thought to possess more toxic potential and potency than soluble (less durable) particles. This suggests an occupational particle mix would be more potent and thus require a stricter control limit than an outdoor mix. Very limited published side-by-side data show occupational respirable results to be less than twice occupational PM2.5 results. These considerations tend to cancel each other, leaving a suggested OEL of 100 µg/m3, with little precautionary extrapolation. ACTION There’s certainly enough evidence to change our professional view of exposure levels that pose a health hazard—where a worker’s respiratory illness is plausibly related to exposure, where control measures are justified. The “art” and “science” parts of industrial hygiene come together. It will take some new thinking for IHs to defer to or even recognize the authority of EPA and ATSDR reference concentrations (RfCs). We repeat that air quality standards, RfCs, and minimal risk levels (MRLs) are “intended” to protect children, the elderly, and sick people, while OELs can expose adults who are well enough to work to higher levels (while at work). But that outlook should depend on the particular substance. The NAAQS for PM2.5 (and below) is actually in the range of an effect level. Unit risk values for carcinogens do not take sensitive populations into account. It will take a lot more ammunition to get OSHA or NIOSH to consider adding a radically stricter enforceable limit to the Z-2 table. Still, there’s more than enough evidence for the TLV committee or NIOSH to consider a new health-based limit for particulate exposure. In addition, our understanding of the biology of these exposures would be improved by laboratory studies of chronic exposure. The IARC determination of “known” applies to an aerosol mix that is half ammonium sulfate and ammonium nitrate, yet we lack laboratory studies of the carcinogenicity of these agents. Long-term inhalation bioassays are needed to complete our assessment.
  • Annals of Epidemiology: “Health effects of particulate air pollution” (April 2009).
  • Annals of Occupational Hygiene: “Low-Toxicity Dusts: Current Exposure Guidelines Are Not Sufficiently Protective” (July 2013).
  • Circulation: “Particulate Matter Air Pollution and Cardiovascular Disease: An Update to the Scientific Statement from the American Heart Association” (June 1, 2010).
  • The Lancet Oncology: “The carcinogenicity of outdoor air pollution” (Dec. 2013).
LINKS •Departments HomeTable of Contents