Toxicology Research and Application

Article

Toxicology Research and Application

Volume 3: 1–9

142 ACGIH Threshold Limit Values® ª The Author(s) 2019

Article reuse guidelines:

(TLV®s) established from 2008-2018

sagepub.com/journals-permissions
DOI: 10.1177/2397847318822137
journals.sagepub.com/home/tor

lack consistency and transparency

Carr J Smith1,2 and Thomas A Perfetti3

Abstract
The American Conference of Governmental and Industrial Hygienists (ACGIH®) states “ . . . the TLV®-CS Committee
preferably relies on published, peer-reviewed literature available in the public domain.” Many studies used to document
threshold limit values (TLVs) are not peer reviewed, with many unpublished. The academic studies published in the peer-
reviewed literature upon which ACGIH relies to determine TLVs rarely report data not already statistically transformed
and thus incalculable. From 2008 to 2018, ACGIH established or reviewed TLVs for 145 different substances or groups of
substances. Studies underlying development of these 145 TLVs demonstrate limited reporting of data sufficient to
facilitate independent statistical analysis. None of the 145 TLVs showed sufficient data to independently substantiate
ACGIH’s evaluations. In the majority of cases, sentinel studies upon which calculation of the TLV was based cannot be
determined. Occupational Safety and Health Administration (OSHA) recommends workplaces rely on ACGIH TLVs and
National Institute for Occupational Safety and Health recommended exposure limits rather than older OSHA permis-
sible exposure limit values to optimize worker safety. Inaccurate or poorly substantiated determination of TLVs can
adversely affect both human health and commerce, as net US sales of the 145 substances represent an estimated
USD$189 billion, about one-fourth of the USD$800 billion chemical market in the United States. ACGIH should provide
the studies relied upon, individual data values supporting the TLVs, and statistical calculation methods underlying the
development of a TLV. The current system of poorly supported and underreported TLVs has the potential to propagate
errors into the standard setting process of other regulatory organizations both nationally and internationally, as the
ACGIH TLVs are used by many foreign regulatory bodies. Increased transparency by ACGIH is consistent with recent
reforms implemented by US EPA.

Keywords
ACGIH, TLV, years 2008–2018, economic impact, non-transparency, inconsistency

Date received: 15 October 2018; accepted: 8 December 2018

Background
The American Conference of Governmental and Industrial
Hygienists (ACGIH®)) has been setting threshold limit val- 1
Albemarle Corporation, Panama City, FL, USA
ues (TLVs ®) for chemicals used in industry for over 2
Department of Nurse Anesthesia, Florida State University, Tallahassee,
72 years. The TLVs are reviewed and updated annually FL, USA
3
by a committee of health professionals who have stated Perfetti & Perfetti, LLC, Winston Salem, NC, USA
that, “These values are based on the best available infor- Corresponding author:
mation from industrial experience, from experimental stud- Carr J Smith, 6400 Brindlewood Court, Mobile, Alabama 36608.
ies and, when possible, from the combination of both”.1 Email: carr.smith@albemarle.com

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons
Attribution-NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial
use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open
Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 Toxicology Research and Application
Additionally, ACGIH has stated that these limits are set at a
level to protect “nearly all workers” from adverse health
effects over their working lifetimes.2 However, the docu-
mentation supporting the implementation of this position is
quite limited.3,4 In addition, the subjective considerations
by ACGIH illustrated by their phraseology “nearly all
workers” and “a small percentage of workers may experi-
ence discomfort from substances at concentrations at or
below the threshold limit . . . ” leave room for interpretative
license from substance to substance and expert panel to
expert panel.
For at least 30 years, scientific articles have been criti-
quing ACGIH regarding the methods employed to develop
TLVs.5–7 There has also been commentary regarding the
inadvisability of adopting TLVs as official limits by the
Occupational Safety and Health Administration (OSHA)
and other countries around the world.5 Additionally, some
authors have posited that industry has played too impor-
tant a role in the setting of TLVs.5–8 In 1990, Roach and
Rappaport postulated that TLVs were being driven direc-
tionally more by considerations of actual exposures in
industry rather than by strict assessment of human health
hazards. 7 Specifically, as exposures in industry were
reduced via improved technologies or better workplace
practices, TLVs were concomitantly lowered. Roach and
Rappaport7 and Nelson9 showed that TLVs reflect attain-
able exposures at the time of implementation and that the
trajectory of TLVs tracks the decline of exposures to
industrial chemicals over time.
Roach and Rappaport7 reviewed the 1976 and 1986
Documentation of the TLVs for Chemical Substances
(TLV-CSs) to determine the basis for the TLVs.3,4 Upon
evaluating the documentation, they found that the TLV
Committee has traditionally emphasized studies involving
human exposures. In the 1976 Documentation,3 225 of the
488 (46%) TLVs listed were at least partly based on human
exposures. In contrast, in the 1986 Documentation,4 only
127 of the 600 (21%) chemicals listed were at least partly
based on human exposures. The trend toward decreasing
emphasis being placed on human workplace exposures
continues today.
Careful analysis by Roach and Rappaport7 of the data
available from the original references cited by the TLV
Committee found that 17% of employees exposed to a
concentration at or below the 1976 TLV3 and 14% of
employees exposed at or below the 1986 TLV4 reported
adverse effects. These results suggest an overall risk of
exposure beyond the recommended workplace air level of
between 1 in 6 and 1 in 7. According to Roach and Rappa-
port,7 this result was inconsistent with the TLVs, represent-
ing levels sufficient to protect “nearly all workers” from
adverse health effects over their working lifetimes.1
Rappaport8 also showed that the TLVs positively corre-
lated with the levels of exposure reported in the studies
cited by the TLV committee in its documentation. Based
on this colinearity, Roach and Rappapport7 concluded that
the TLVs reflected the exposure levels at the time the limits
were adopted. Castleman and Zeim5 have alleged that the
colinearity between the achievability and the actual thresh-
old value set suggests that corporations unduly influenced
the setting of TLVs through personal communications with
the ACGIH Committee. (A less pejorative interpretation of
this colinearity is that requiring achievement of an expo-
sure standard for which technology is not currently avail-
able within a feasible cost range is not practical if the
compound under consideration is necessary and cannot be
replaced. In these instances, use of personal protective
equipment by all workers within the exposure zone can
sometimes be needed.) The ACGIH Board of Directors
responded by stating that they solicited information from
all possible sources through the Notice of Intended
Changes.1 The Board also stated that the analysis of Rap-
paport8 suffered from bias introduced by reliance on a
“small non-representative sample of TLVs . . . ”.10 Rappa-
port and Roach11 countered that their statistical sample was
not biased merely because of limited size because inclusion
and consideration of all pertinent studies cited in the Doc-
umentation represented a generalizable subsample of TLVs
as a whole.
Rappaport8 extended this line of investigation and ana-
lyzed reductions in TLVs as a function of time for 27 sub-
stances in the “Notice of Intended Changes (for 1991–
1992)”. 12 The result of his analysis showed overall
median-fold reductions of 2.0–2.5 between 1946 and
1989. In 1993, Nelson9 completed a detailed analysis of all
630 TLVs as a whole, industry-specific TLVs, and TLVs
for known or suspected carcinogens. For each category of
TLV, Rappaport8 determined the rate of change of TLVs
between 1946 and 1990 and found the median-fold reduc-
tion for all TLVs to be 2.0. For most industries in this
analysis, the TLV reduction was also 2.0-fold, and for
known or suspected carcinogens, the TLV reduction was
2.4-fold. The rate of change for all TLVs decreased with
time (21.7% reductions per year in 1951 to 13.3% per year
in 1990). For most industries, the reduction rate was 13.3%
reductions per year, while for known or suspected carcino-
gens, it was 16% per year (over this period of time). He
postulated that the results of this analysis could be used to
infer past exposures.
Over the last several years, the existence of a replicabil-
ity crisis in the biomedical literature has become widely
recognized.13 The replicability crisis has elevated the
essentiality of transparency in any data-based process upon
which important decisions rely. To the best of our knowl-
edge, the current investigation introduced below represents
the most comprehensive analysis of the lack of transpar-
ency in the ACGIH TLV process. Recently, we addressed
the lack of transparency in the ACGIH TLV process.14 The
current analysis extends the previous study by examining
every TLV set during the 10-year interval between 2008
and 2018.
Smith and Perfetti
Introduction
ACGIH TLVs levels are extremely influential. ACGIH
TLVs were in widespread industrial hygiene practice for
many years before any other workplace exposure limits
were established.15 The OSHA requires the ACGIH TLV
to be listed on Safety Data Sheets.16 OSHA recommends
that its own legally binding permissible exposure limits
(PELs) not be considered but rather that other exposure
limit recommendations including the ACGIH TLV guide
workplace safety standards.17 The ACGIH TLV directly
influences the following: selection and use of personal
protective equipment, installation of workplace ventila-
tion equipment, monitoring of workplace air levels, deci-
sions by customers as to the particular chemical selected
for a given industrial application or process, decisions by
chemical manufacturers regarding whether to produce a
particular chemical or keep producing a particular chem-
ical, and decisions by chemical company research and
development departments as to whether a chemical sub-
stitute should be sought for a chemical with a TLV sug-
gestive of suboptimal toxicity.
Despite the serious adverse consequences of an inaccu-
rately determined TLV, the current TLV determination
process is primarily dependent on an academic peer-
reviewed literature in the midst of an irreproducibility crisis
not yet fully explained or addressed.14 Also, a significant
number of studies relied upon by ACGIH are not published
and the data are not accessible to the public, thereby further
exacerbating deficiencies in the TLV process. In addition, a
well-established bias against the publication of negative
results can additionally skew evaluations of the toxicologi-
cal profile of a chemical toward overestimate of hazard.14
In this analysis, we evaluated the transparency and consis-
tency of the data relied upon by ACGIH to set TLVs for
145 substances or groups of substances between the years
2008 and 2018.
Methods
Construction of Supplemental Table 1 on TLVs
for chemicals adopted or reviewed from 2008
to 2018
To determine the chemicals for which TLVs have been
established and the supporting documentation, it was nec-
essary to purchase the 2018 edition of the 7th Edition,
ACGIH TLVs and Biological Exposure Indices (BEIs®).18
A search of the book/CD was conducted. On page 11, a
section on adopted TLV values is presented in tabular form.
The relevant table was analyzed, and the establishment
dates of the adopted TLVs from 2008 to 2018 were recov-
ered and placed into a table. Information on the 145 che-
micals was then collected for the purpose of further
examination. The information included the following:
chemical name and Chemical Abstracts Service Registry
3
Table 1. Distribution of different biological activities considered
for the 145 chemicals.
TLV basis–critical effecta Percent
Irritation 30.4
CNS effects 12
Respiratory effects 8.8
Liver effects 8.7
Blood effects 6.4
Kidney effects 4.7
Skin effects 3.8
Cancer 3.7
Sensitization 2.8
All other effects 18.7
ACGIH: American Conference of Governmental and Industrial Hygienists.
a
Merkle SE: ACGIH’s Role in the Development of Exposure Assessment
Guidelines for Occupational Hygiene. AIHA Carolinas Section, Spring
Conference, Charlotte, NC, March 13–15, 2002. http://www.google
.com/url?sa¼t&rct¼j&q¼&esrc¼s&source¼web&cd¼1&ved¼2ahUKE
wjn0bG3vfzdAhUQnFkKHWlCAvIQFjAAegQICRAC&url¼http%3A%
2F%2Fwww.aiha-carolinas.org%2Fmeeting-info%2Ftlvs.ppt&usg¼AOv
Vaw3SknfQmKftcnD-zueeN74R
Number(CASRN), TLV from ACGIH, year the TLV was
adopted, human health basis for the TLV, references to
sentinel studies used to form the recommendations for the
TLV (as well as the range of dates for the sentinel articles),
genotoxicity end points important in the TLVs, information
on the nature of supporting data present in the documenta-
tion (raw data, statistically transformed retrievable data,
statistically transformed data that were irretrievable, tabu-
lar data, graphical data, other data), and the page references
for the information retrieved from the ACGIH 7th Ed.
Documentation on each chemical.
All documentation for each chemical was reviewed
thoroughly. Information in the TLV Recommendations was
scrutinized, and all references used in the determination of
the TLV were recorded. The TLV Basis was recorded. The
section on genotoxicity studies was reviewed. Information
from these sections of the documentation was placed in
Table 1 for each of the 145 chemicals. The references were
examined to determine the following: whether they were
derived from peer-reviewed journals, irretrievability of
URLs, findability of unpublished references, accessibility
of the journals and books to the general public, relevance of
the content of the reference or whether it was peripheral,
and so on. The citations were all corrected (where neces-
sary) and retyped into the uniform Vancouver style.
Construction of Supplemental Table 2 on
carcinogenicity indices for top 145 chemicals
reviewed and adopted by ACGIH between 2008
and 2018
Information on the carcinogenicity indices came from four
sources: chemical name and CASRN, TLV, ACGIH
Cancer Rating obtained from the 2018 edition of the
4 Toxicology Research and Application

Table 2. Carcinogenicity indices from NTP, ACGIH, and IARC.

NTP, US Department of
Health and Human Services ACGIH IARC
Group K: known to be Group A1: confirmed human carcinogen Group 1: carcinogenic to humans
human carcinogens
Group R: reasonably Group A2: suspected human carcinogen Group 2A: probably carcinogenic to humans
anticipated to be human
carcinogens
Group A3: confirmed animal carcinogen with unknown Group 2B: possibly carcinogenic to humans
relevance to humans
Group A4: Not classifiable as a human carcinogen. Group 3: The agent (mixture, exposure
“There are inadequate data on which to classify the circumstance) is not classifiable as to its
agent in terms of its carcinogenicity in humans and/or carcinogenicity to humans.
animals.”
Group A5: Not suspected as a human carcinogen. Group 4: The agent (mixture, exposure
circumstance) is probably not carcinogenic
to humans
ACGIH: American Conference of Governmental and Industrial Hygienists; NTP: National Toxicology Program; IARC: International Agency for Research
on Cancer.

7th Edition, and ACGIH TLVs and BEIs.18 The National

Toxicology Program (NTP) carcinogenicity rating was
extracted from the NTP website (https://ntp.niehs.nih.gov/
pubhealth/roc/index-1.html).19 The International Agency
for Research on Cancer (IARC) carcinogenicity rating was
taken from the IARC website (https://monographs.iarc.fr/
agents-classified-by-the-iarc/) 20 The carcinogenicity
index/ranking of the chemical was based on the process
developed by Smith et al.21

Construction of Supplemental Table 3 on TLVs,
economic value of the TLV-chemicals, and exposure
information on the top 145 chemicals produced
or imported during 2017
Information on the TLVs, economic value of the TLV-
chemicals, and exposure information included the chem-
ical name and CASRN and TLV value.18 The data on the
pounds of each chemical produced or imported into the
United States were obtained from 2017 data from the US
EPA’s Chemical Data Reporting (CDR) database 22
(http://www.epa.gov/chemical-data-reporting and http://
epa.gov/cdr/).
Data on the costs per metric ton were obtained from
Intratec at https://www.intratec.us/indexes-and-pricing-
data/chemicals-pricing-data-store23 and http://www.icis.
com/about/free-trials/24 or were obtained from a search of
the Internet. The estimated exposure data were taken from
Labrèche et al. at http://www.irsst.qc.ca/media/documents/
PubIRSST/R-895.pdf 25 and anonymous https://submis
sions.swa.gov.au/SWAforms/wes/Documents/exposure-
standards-discussion-paper.pdf.26 The data for all other
tables come from the 2018 edition of the 7th Edition,
ACGIH TLVs and BEIs18 or searches of the Internet. All
sources are noted at the bottom of each table.
Figure 1. Frequency diagram of chemical TLVs adopted or
reviewed from 2008 to 2018. TLV: threshold limit value.
Results
In total, ACGIH has issued approximately 657 chemical
listings (i.e. TLVs), representing more than 700 chemi-
cals.18 TLVs have been issued for a variety of situations
including one unique organic chemical, one unique inor-
ganic chemical, a chemical family (e.g. butanes or
butenes), or an undefined group of chemicals, for example,
soluble inorganic salts of beryllium. Since many of the
chemical listings were issued many years ago, only the
145 chemicals for which TLVs were issued over the last
10 years (2008–2018) were analyzed as representative of
current ACGIH practices (Table 1). Table 1 and references
in Table 1 can be found in Online Supplemental Material.
Between the years 2008 and 2018, 145 chemical listings
were adopted or reviewed. Figure 1 shows the number of
chemical listings adopted or reviewed per year. On aver-
age, about 13 chemical listings were adopted or reviewed
per year, ranging from 8 to 20 chemical listings per year.
As previously mentioned, the information in Table 1
includes the following: chemical name and CASRN, TLV
from ACGIH, year the TLV was adopted, human health
Smith and Perfetti
Figure 2. Frequency diagram of references to sentinel studies per
year for chemicals with TLVs adopted or reviewed from 2008 to
2018. TLV: threshold limit value.
basis for the TLV, references to sentinel studies used to
form the recommendations for the TLV (as well as the
range of dates for the sentinel articles), genotoxicity end
points important in the TLVs, information on the nature of
supporting data present in the documentation (raw data,
statistically transformed retrievable data, statistically trans-
formed data that were irretrievable, tabular data, graphical
data, other data), and the page references for the informa-
tion retrieved from the ACGIH 7th Ed. Documentation on
each chemical.
ACGIH considers a wide variety of biological end
points for the human health basis for the TLV in setting
the TLV. Table 1 shows the distribution of different biolo-
gical activities considered for the 145 chemicals. The fol-
lowing potential toxicities and the percentage of chemicals
considered for that end point in TLV setting are as follows:
irritation (30.4%), Central Nervous System (CNS) effects
(12%), respiratory effects (8.8%), liver effects (8.7%), blood
effects (6.4%), kidney effects (4.7%), skin effects (3.8%),
cancer (3.7%), sensitization (2.8%), and all other effects
(18.7%).27 Of the 145 chemicals (Table 1), only 38 (27%)
relied upon a single toxicity end point in setting the TLV.
There were 1257 sentinel studies references by ACGIH
for the 145 chemicals with TLVs adopted or reviewed from
2008 to 2018.18 The median publication date for these arti-
cles was 1988 + 17 years. The earliest publication was
1919 and the latest was 2016 (Figure 2).
Genotoxicity data were reported on most but not all of the
145 chemicals.18 About 8% (12 of 145) had no genotoxicity
data. Typically, results for some combination of Ames
assays, micronucleus assay, sister chromatid exchange chro-
mosomal aberrations, mouse lymphoma assay, and/or
dominant-lethal assay were provided. The most abundant
information was on Ames test results. About 75% of all the
chemicals were tested via the Ames assay. Of these 69% (75
of 108) were Ames negative, 17% (18 of 108) were Ames
positive, and 14% (15 of 108) had equivocal Ames results.
For the 145 chemicals (Table 1), there was no case in
which a single sentinel study could be identified as the sole
5
basis for setting the TLV. Five of 145 (3.5%) chemical
listings reported but a single reference, i.e. carfentrazone-
ethyl, 1-chloro-1-nitropropane, methyl isopropyl ketone,
silicon tetrahydride, and sulprofos. One chemical listing
provided no references, that is, boron trifluoride. None of
the 145 chemical listings reported the underlying methods
employed for statistical analysis in the documentation
related to determination of the TLV. None of the 145 chem-
ical listings presented data in the supported documentation
that could be used to facilitate outside independent statis-
tical analysis.
As noted previously, 5 of the 145 (3.7%) chemicals in
Table 1 incorporated rodent tumor (cancer end point) data
into the determination of the TLV. In the limited number of
cases dependent on cancer end points (Table 2), the chemical
listing might not cite the relevant National Cancer Institute
(NCI) study even when they exist, for example, vinyl acet-
ate.28–30 When an NTP study is cited, for example, with
vanadium pentoxide, the NTP report does contain tumor
count data amenable to independent statistical analysis.31
However, that data are not in the ACGIH Documentation
and independent statistical analysis of these few cases where
“raw” NCI or NTP data are available would be superfluous
as NCI and NTP data analyses are conducted by professional
statisticians. Statistical analysis is not the basis upon which
NCI or NTP studies are usually criticized.32 Table 2 can be
found in the Supplemental Information.
Examination of Table 2 (carcinogenicity indices for top
145 chemicals reviewed and adopted by ACGIH between
2008 and 2018) shows that 128 of 145 chemical listings
have either 3 or 4 entries notated as not otherwise speci-
fied (n.o.s.; as carcinogenic) or 1 or 2 entries notated as
n.o.s. with entries for IARC, NTP, ACGIH, or Smith et al.
carcinogenicity indices.21 In total, these 128 chemical
listings show a high degree of consistency in terms of their
carcinogenicity index rating. For example, a group R rat-
ing from NTP is analogous to a group A2 rating from
ACGIH and a group 2A rating from IARC. A group A3
rating from ACGIH is analogous to a group 2B rating from
IARC (Table 2).
The remaining 17 chemical listings in Table 2 are all
considered to be more carcinogenic. Of these 17 chemical
listings, 8 (47%) showed consistency across the recom-
mending bodies in their carcinogenicity ratings. The car-
cinogenicity ratings on the remaining nine chemical
listings were inconsistent with the ACGIH rating being
less carcinogenic in all but a single case. Overall, 136 of
the chemical listings (94%) showed consistency in their
carcinogenicity indices.
Table 3 lists information on the TLV, cost and exposure
information on the 145 chemicals possessing TLVs
adopted or reviewed between 2008 and 2018 and produced
or imported during 2017 (information on 2018 was not yet
available). Table 3 can be found in the Supplemental Infor-
mation, which lists the chemical name, TLV, pounds of
chemical produced or imported annually in the United States
6 Toxicology Research and Application
(2017), average cost per pound of the chemical, metric tons
of each chemical produced or imported, sales in US dollars
(USD) per metric ton of chemical, and the estimated annual
chemical exposure for males and females in all job categories
in the United States. The information on TLVs was obtained
from the 2018 7th Edition ACGIH Book of TLVs and BEIs.18
The data on the poundage of each chemical produced or
imported into the United States were obtained from 2017
data from the US Environmental Protection Agency (EPA)’s
CDR database22 (http://www.epa.gov/chemical-data-report
ing and http://epa.gov/cdr/).
Data on the costs per metric ton were obtained from
Intratec at https://www.intratec.us/indexes-and-pricing-
data/chemicals-pricing-data-store23 and http://www.icis
.com/about/free-trials/24 or were obtained from a search
of the Internet. The estimated exposure data were excerpted
from Labrèche et al. at http://www.irsst.qc.ca/media/docu
ments/PubIRSST/R-89525 and anonymous https://submis
sions.swa.gov.au/SWAforms/wes/Documents/exposure-
standards-discussion-paper.pdf.26
The total sales of the 145 chemicals listed in Table 3
sum to USD$189,909,385,400. (It should be noted that
metals and gases were not included. Some price and ton-
nage data were not available for certain chemicals.) The
estimated sales for the total US chemical industry are
approximately 800 billion USD. So, the chemical listings
in Table 3 represent approximately 24% of the total sales of
the US chemical industry.
Some degree of chemical exposure is common, although
biologically significant chemical exposures are less com-
mon. Exposure to clinically significant levels of chemicals
is more common in occupational settings than in the gen-
eral population. As of October 2018, there are approxi-
mately 150,000,000 jobs in the United States, https://
www.bls.gov.ces and https://www.deptofnumbers.com/
employment/us/.33 The US chemical industry directly
employs over 800,000 workers with an additional
2,700,000 workers indirectly employed by industry
suppliers.34 The US chemical sector has more than 10,000
firms that produce over 70,000 different chemicals.34 The
145 chemicals listed represent only a small fraction of
the total number of chemicals produced or imported into the
United States. It is extremely difficult to accurately estimate
the percentage of the population exposed to a particular
chemical, although estimates are calculated and published.
For the 145 chemicals considered in this analysis, the per-
centage of the population estimated to be exposed to one or
more of these 145 ranges from 0.1% and 33.0% or 150,000
to 49,500,000 people, respectively.25,26,34
Discussion
Role of ACHIH within the regulatory framework
OSHA and the National Institute for Occupational Safety
and Health (NIOSH) were created on December 29, 1970,
when President Richard M. Nixon signed the Occupational
Safety and Health Act.35 The relative youth of OSHA and
NIOSH can be contrasted with the formation of the prede-
cessor of ACGIH in 1938.36 In 1941, the TLV-CS Com-
mittee was established and became a standing committee in
1944. In 1946, ACGIH adopted the first list of 148 expo-
sure limits, at that time termed maximum allowable con-
centrations (MACs). In 1956, the term MAC was changed
to TLV. Currently, ACGIH provides TLVs for over 700
chemical substances and physical agents.37,38 In January
2016, ACGIH became a 501(c)(3) charitable scientific
organization and is not a regulatory agency of the United
States or any other country. Therefore, from a legal per-
spective, ACGIH TLVs are nonbinding recommendations
and not legally binding regulations.39
The term “recommendation” belies the historical, regu-
latory, public health, and economic impact of the ACGIH
TLV. The centrality of the role played by TLVs is due to
the limited number of authoritative bodies in the United
States, issuing similar exposure limits. OSHA issues leg-
ally binding PELs.40,41 However, OSHA PELs are rarely
updated and are not generally relied upon. In a January
2016 interview, then OSHA administrator David Michaels
stated that
Many of these PELs are dangerously out of date and do not
adequately protect workers. Past efforts to update our PELs
have largely been unsuccessful. Since 1971, OSHA has suc-
cessfully established or updated PELs for only about 30 che-
micals. We have issued only one new exposure limit since the
year 2000. As a result, many workers are currently being
exposed to levels of chemicals that are legal, but not safe.42
Due to OSHA’s lack of confidence in its own PELs, the
agency developed an online resource, known as the Anno-
tated Permissible Exposure Limits (or annotated PEL)
tables (https://www.osha.gov/dsg/annotated-pels/index.
html). The stated purpose of the Annotated PELs is “to
enable employers to voluntarily adopt newer, more protec-
tive workplace exposure limits than those required by
OSHA.” The Annotated PEL tables display a side-by-side
comparison of OSHA PELs for general industry; NIOSH
recommended exposure limits, California OSHA PELs,
and ACGIH TLVs.43 Therefore, the agency of the US Fed-
eral Government with primary responsibility for worker
safety recommends that workplaces rely on the TLVs set
by ACGIH.
Recommendations for improving the TLV-setting
process
A thorough analysis covering the last 10 years and 145
substances or groups of substances demonstrates that the
process by which ACGIH sets TLVs could be improved by
increasing transparency. First, the purchase of an expensive
book/CD is required to initiate the evaluation of the degree
Smith and Perfetti 7

of transparency of the TLV process. Second, an indepen- Perspective

dent outside third-party would not be able to evaluate the
The mission to promote health and safety in the workplace
scientific logic and statistical analysis underpinning the
undertaken by volunteer industrial hygienists and other
promulgated TLVs. Third, the criteria for deciding
occupational and safety professionals over many years
the relative weighting of the diverse biological end points
under the auspices of ACGIH is highly laudable. ACGIH’s
considered are not apparent. Compounding the already
TLVs and BEIs are not and have never been meant to be
extremely problematic situation, many studies conducted
consensus standards.49 TLVs and BEIs are health-based
over the last several years have demonstrated that a large
values intended for use in the practice of industrial hygiene
but not exactly determinable percentage of the results in the
as guidelines or recommendations to assist in the control of
peer-reviewed academic literature are not reproducible in
potential workplace hazards.49,50 TLVs and BEIs are only
another laboratory attempting to employ the original
one of multiple factors to be considered in evaluating spe-
protocol.14 Based on analyses conducted in different fields
cific workplace situations and conditions.50
of study, a reasonable estimate would be that about 40–50%
Under the auspices of ACGIH, the work of multidisci-
of the biomedical literature is not replicable.14,44–48 ACGIH
plinary committees that review and analyze published and
states that the primary resource upon which it relies in
peer-reviewed literature across various scientific disci-
setting TLVs is this same body of peer-reviewed academic
plines including industrial hygiene, toxicology, occupa-
literature in the midst of a replicability crisis. Given the
tional medicine, and epidemiology represents valuable
enormous number of workers whose safety is affected by
and unique expressions of science-based biomedical opin-
the TLV process (Table 3), and the hundreds of billions of
ion.51 Numerous benefits emanate from such voluntary
dollars of commerce involved (Table 3), the opaqueness
cooperation among scientists and clinicians including joint
of the current TLV setting process should be replaced by a
hypothesis generation, discovery of synergistic observa-
protocol that describes the selection of studies, numerical
tions, collaborative insights regarding risk characterization,
values extracted for consideration, statistical methods
elucidation of possible mechanisms and risk modifiers, and
employed, actual statistical calculations, raw data ana-
open discussions on impact of antecedent or intervening
lyzed, and logic employed. Implementation of these sug-
variables. Collective innovative thinking has historically
gested changes would represent a modernization of
resulted in better and more productive solutions to challen-
protocol but would not require outlay of additional
ging issues.52,53
resources. Hopefully, ACGIH, industry, and government
It is not our intent to, in the words of Murray Kempton,
can collaborate in improving the workplace air standard
to play the role of critic of this integral system by entering
setting process.
the battlefield to shoot the wounded after the war but rather
Many online open-access journals now accept supple-
to address and potentially correct shortfalls in the pro-
mental materials that can be accessed by interested readers.
cess.7,54,55 ACGIH did not set out to have OSHA become
Going forward, ACGIH might consider a similar online
dependent on the TLV process but rather the dependency
repository for the voluminous raw data underlying sentinel
developed due to problems experienced by OSHA with its
studies. The raw data would need to be provided by the
own PEL system.56 The planned updates and improve-
authors of the sentinel study, preferably as part of the orig-
ments in the OSHA PEL system should hopefully alleviate
inal publication. Since many older studies that are relevant
some of the over reliance on ACGIH TLVs and BEIs.57
to the TLV-setting process were published prior to the
implementation of supplemental materials, ACGIH could
improve transparency by explicitly elucidating the trans- Declaration of conflicting interests
formed data values and their location in the specific publi- The author(s) declared no potential conflicts of interest with
cations upon which the TLVs are derived. If the authors of respect to the research, authorship, and/or publication of this
a sentinel study are in possession of archived raw data, article.
these authors should be willing to share these data for the
purpose of independent outside statistical analysis. Results Funding
from independent statistical analysis would be shared with The author(s) received no financial support for the research,
all interested parties for the purpose of improving the qual- authorship, and/or publication of this article.
ity of information available to the ACGIH expert panelists.
In summary, the emphasis in TLV setting would shift from ORCID iD
peer-reviewed studies to transparent peer-reviewed studies
Carr J Smith https://orcid.org/0000-0002-8708-5208
and Good Laboratory Practices (GLP) reports facilitating
commentary by all interested parties. A transparent, more
science-based approach to TLV-setting with worker safety References
as the paramount goal would serve as an important adjunct 1. Anonymous. American Conference of Governmental Indus-
to the efforts of both national and international regulatory trial Hygienists (ACGIH) Board of Directors. Threshold limit
agencies. values: a more balanced appraisal. Appl Occup Environ Hyg
8 Toxicology Research and Application
1990; 5(6): 340–344. DOI: 10.1080/1047322X.1990.
10389650.
2. Anonymous. American Conference of Governmental Indus-
trial Hygienists (ACGIH). Threshold limit values and biolo-
gical exposure indices for 1992-1993. Cincinnati: ACGIH,
1992, ISBN 0936712996.
3. Anonymous. American Conference of Governmental Indus-
trial Hygienists (ACGIH). Documentation of the threshold
limit values. 3rd ed. 3rd printing, Cincinnati: ACGIH, 1976.
4. Anonymous. American Conference of Governmental Indus-
trial Hygienists (ACGIH). Documentation of the threshold
limit values, 5th ed, Cincinnati: ACGIH, 1986.
5. Castleman BI and Zeim GE. Corporate influence on threshold
limit values. Am J Ind Med 1988; 13: 531–559.
6. Zeim GE and Castleman BI. Threshold limit values: histori-
cal perspectives and current practice. J Occup Med 1989; 31:
910–918.
7. Roach SA and Rappaport SM. But they are not thresholds: a
critical analysis of the documentation of threshold limit val-
ues. Am J Ind Med 1990; 17: 727–753.
8. Rappaport SM. Threshold limit values, permissible exposure
limits, and feasibility: the bases for exposure limits in the
United States. Am J Ind Med 1993; 23: 683–694.
9. Nelson RW. An analysis of changes in threshold limit values
over time. Thesis submitted to the Faculty of the University
of North Carolina at Chapel Hill, NC in partial fulfillment of
the requirements for the degree of Master of Science in Envi-
ronmental Engineering, 1993, pp. 1–156, http://www.dtic.
mil/dtic/tr/fulltext/u2/a267492.pdf (accessed 10 October
2018).
10. Breysse PA. ACGIH TLVs: a critical analysis of the docu-
mentation. Am J Ind Med 1991; 20: 423–428.
11. Rappaport SM and Roach SA. Response to ACGIH critical
analysis. Am J Ind Med 1991; 20: 429–430.
12. Anonymous. American Conference of Governmental Indus-
trial Hygienists (ACGIH). Threshold Limit values and biolo-
gical exposure indices for 1991-1992. Cincinnati: ACGIH,
1991.
13. Begley CG and Ellis LM. Raise standards for preclinical
cancer research. Nature 2012; 483: 531–533.
14. Smith CJ and Perfetti TA. Improving the ACGIH threshold
limit value (tlv) process. Tox Res Appl. 2018; 2: 1–22.
15. Paustenbach DJ, Cowan DM and Sahmel J. Chapter 20: The
history and biological basis of occupational exposure limits
for chemical agents. In: Rose VE and Cohrssen B (eds)
Patty’s industrial hygiene, Sixth ed. New York: Wiley &
Sons, 2011, pp. 865–954. DOI: 10.1002/0471435139.hyg041.
16. United States Department of Labor (US DOL). Occupational
Safety and Health Administration US DOL, OSHA Permis-
sible Exposure Limits – Annotated Tables. 2018. https://
www.osha.gov/dsg/annotated-pels/index.html) (accessed 13
October 2018).
17. Popendorf W. Industrial hygiene control of airborne chemi-
cal hazards. Boca Raton: CRC Press, 2006, pp. 1–728.
18. American Conference of Governmental Industrial Hygienists
(ACGIH ®). 2018 TLVs ® and BEIs ® with 7th Edition
Documentation, CD-ROM (Single User Version); Product
ID: 0118CD; ISBN: 978-1-607261-02-5.
19. National Toxicology Program website. 2018. https://ntp.
niehs.nih.gov/pubhealth/roc/index-1.html (accessed 13 Octo-
ber 2018).
20. International Agency for Research on Cancer IARC. 2018.
https://monographs.iarc.fr/agents-classified-by-the-iarc/
(accessed 13 October 2018).
21. Smith CJ, Perfetti TA, Ko GM, et al. Ames mutagenicity,
structural alerts of carcinogenicity, Hansch molecular para-
meters (ClogP, CMR, MgVol), tumor site concordance/multi-
plicity, and tumorigenicity rank in 2-year NTP studies. Tox
Res Appl 2018; 2: 1–14.
22. US EPA. Public version of the 2016 Chemical Data Report-
ing (CDR) database (an Access data file). 2016. https://www.
epa.gov/chemical-data-reporting and http://epa.gov/cdr/
(accessed 4 October 2018).
23. Intratec Solutions, LLC. 2017. https://www.intratec.us/
indexes-and-pricing-data/chemicals-pricing-data-store
(accessed 9 October 2018).
24. Independent Chemical Information Service (ICIS Pricing).
2017. http://www.icis.com/about/free-trials/ (accessed 11
October 2018).
25. Labrèche F, Lacourt A and Lavoué J. Chemical and biological
hazard prevention, studies and research projects, Report R-895,
occupational exposure to chemical and physical contaminants,
sex-differentiated analysis. Montréal: Institut de recherche
Robert-Sauvé en santé et en sécurité du travail (IRSST),
November 2015, pp. 1–66. http://www.irsst.qc.ca/media/docu
ments/PubIRSST/R-895.pdf (accessed 5 October 2018).
26. Anonymous, Safe Work Australia, Discussion Paper. The
role of chemical exposure standards in work health and
safety laws, November 2015, pp. 1–24. https://submissions.
swa.gov.au/SWAforms/wes/Documents/exposure-standards-
discussion-paper.pdf (accessed 13 October 2018).
27. Merkle SE. ACGIH’s role in the development of exposure
assessment guidelines for occupational hygiene. 2002. In: AIHA
carolinas section, spring conference, Charlotte, NC, 13–15
March 2002. http://www.google.com/url?sa¼t&rct¼j&q¼
&esrc¼s&source¼web&cd¼1&ved¼2ahUKEwjn0bG3
vfzdAhUQnFkKHWlCAvIQFjAAegQICRAC&url¼http%3
A%2F%2Fwww.aiha-carolinas.org%2Fmeeting-info%2Ftlvs.
ppt&usg¼AOvVaw3SknfQmKftcnD-zueeN74R”usg¼
AOvVaw3SknfQmKftcnD-zueeN74R (accessed 2 October
2018).
28. Lijinsky W and Reuber MD. Chronic toxicity studies of vinyl
acetate in Fischer rats. Toxicol Appl Pharmacol 1983; 68: 43–53.
29. National Cancer Institute (NCI). Bioassay of vinyl acetate in
F344 rats pathology report addendum. U.S. EPA/OTS public
files. Rockville: National Cancer Institute. FYI-AX-1082-
0197, 1982a.
30. National Cancer Institute (NCI). Chronic toxicity studies of
vinyl acetate in Fischer rats. US EPA/OTS public files. Rock-
ville: National Cancer Institute. FYI-AX-0782-0197, 1982b.
31. National Toxicology Program (NTP). Technical report on the
toxicology and carcinogenesis studies of vanadium pentoxide
Smith and Perfetti
(CAS No. 1314-62 -1) In F344/N Rats and B6C3F1 Mice
(Inhalation Studies), National Toxicology Program P.O. Box
12233 Research Triangle Park, NC 27709, NTP TR 507.
2002. NIH Publication No. 03-4441, U.S. Department of
Health and Human Services, Public Health Service, National
Institutes of Health, December 2002. https://ntp.niehs.nih.
gov/ntp/htdocs/lt_rpts/tr507.pdf (accessed 1 October 2018).
32. Smith CJ and Perfetti TA. The ‘false positive’ conundrum in
the NTP 2-year rodent cancer study database. Tox Res Appl
2018; 1–13.
33. US Department of Labor, Bureau of Labor Statistics. Current
employment statistics. https://www.bls.gov.cesandhttps://
www.deptofnumbers.com/employment/us/ (accessed 14
October 2018).
34. US Department of Commerce, International Trade Adminis-
tration, and Industry and Analysis Unit. Chemical Spotlight.
The Chemical Industry in the United States. 2018. https://
www.selectusa.gov/chemical-industry-united-states
(accessed 11 October 2018).
35. Occupational Safety and Health Act. 29 U.S.C. United States
Code, 2016 Edition, Title 29 – Labor Chapter 15 - Occupa-
tional Safety and Health Sec. 651 - Congressional statement
of findings and declaration of purpose and policy. From the
U.S. Government Publishing Office. www.gpo.gov (accessed
5 October 2018).
36. Bobst S. Chapter 4 - The history of risk assessment within
OSHA and ACGIH: Asbestos case study. In: Bobst S (ed.)
History of risk assessment in toxicology, History of toxicology
and environmental health. Cambridge MA: Academic Press,
2017, pp. 27–30. ISBN: 978-0-12-809532-4.
37. Corn M. Role of the American Conference of governmental
industrial hygienists and the American industrial hygiene
association in OSHA Affairs. Am Ind Hyg Assoc J 1976;
37(7): 391–394.
38. American Conference of Governmental Industrial Hygienists
(ACGIH). 2018. https://www.acgih.org/about-us/history
(accessed 28 September 2018).
39. American Conference of Governmental Industrial Hygienists
(ACGIH). 2018. TLVs and BEIs. https://www.acgih.org/
(accessed 28 September 2018).
40. Allen DT and Shonnard DR. Green engineering: environmen-
tally conscious design of chemical processes. Upper Saddle
River, NJ: Prentice Hall, 2001. ISBN-13: 9780130619082.
41. Occupational and Safety Administration (OHSA). 2018.
https://www.osha.gov/ (accessed 15 September 2018).
42. Johnson D. OSHA’s exposure limits are dangerously out of
date, industrial safety and hygiene news. 2016. https://www.
ishn.com/articles/103083-oshas-exposure-limits-are-danger
ously-out-of-date (accessed 15 September 2018).
43. United States Department of Labor (US DOL). Occupa-
tional safety and health administration US DOL, OSHA
permissible exposure limits – annotated tables. 2018.
https://www.osha.gov/dsg/annotated-pels/index.html)
(accessed 14 September 2018).
9
44. Open Science Collaboration (OSC). Estimating the reprodu-
cibility of psychological science. Science 2015; 349(6251):
aac4716.
45. Baker M. 1500 Scientists lift the lid on reproducibility. Sur-
vey sheds light on the ‘crisis’ rocking research. Nature 2016;
533(7604): 452–454.
46. Boutron I and Ravaud P. Spin in biomedical literature. P Natl
A Sci 2018; 115(11): 2613–2619.
47. Iqbal SA, Wallach JD, Khoury MJ, et al. Reproducible
research practices and transparency across the biomedical
literature. Vaux DL, ed. PLoS Biol 2016; 14(1): e1002333.
48. Li G, Abbade LPF, Nwosu I, et al. A systematic review of
comparisons between protocols or registrations and full
reports in primary biomedical research. BMC Med Res Meth-
odol 2018; 18(9): 1–20.
49. ACGIH. Physical and biological hazards of the workplace,
ACGIH policy statement (xxiii) and ACGIH statement of
position (xxv-xxvi). In: Stave GM and Wald PH (eds.) New
York: John Wiley & Sons, Technology & Engineering, 2016,
pp. i–608.
50. United States Department of Labor (USDOL). Occupational
Safety and Health Administration (OSHA). Important note
regarding the ACGIH TLV, policy statement on the uses of
TLVs® and BEIs®. 2002. https://www.osha.gov/dsg/anno
tated-pels/note.html (accessed 10 September 2018).
51. Ford PJ and Kubu CS. Stimulating debate: ethics in a multi-
disciplinary functional neurosurgery committee. J Med Ethics
2006; 32(2): 106–109.
52. Hamel PB, Riley CM, Hunter WC, et al. Monitoring bird
migration in the caribbean Basin: multi-national cooperation
can close the loop. In: Bird conservation implementation and
integration in the Americas: proceedings of the Third Inter-
national Partners in Flight Conference (eds RC John and DR
Terrell), Asilomar, California, Volume 2 Gen. Tech. Rep.
PSW-GTR-191. Albany, CA: U.S. Dept. of Agriculture,
Forest Service, Pacific Southwest Research Station, 20–24
March 2002, pp. 729–733. USDA forest service gen. Tech.
Rep. PSW-GTR-191. 2005. https://www.fs.fed.us/psw/publi
cations/documents/psw_gtr191/psw_gtr191_0729-0733_
hamel.pdf (accessed 10 September 2018).
53. Bennett LM and Gadlin H. Collaboration and team science:
from theory to practice. J Investig Med 2012; 60(5): 768–775.
54. Robinson JC and Paxman DG: Technological, economic, and
political feasibility in OSHA’s air contaminants standard.
J Health Polit Policy Law 1991; 16(1): 1–18.
55. Hogan TJ and Nalbone JT. Threshold limit values - The
development process: a TLV is not just a number. American
Society of Safety Engineers 2016; 4: 52–56.
56. Robinson JC, Paxman DG and Rappaport SM. Implications
of OSHA’s reliance on TLVs in developing the air contami-
nants standard. Am J Ind Med 1991; 19: 3–13.
57. United States Department of Labor (USDOL). Occupational
Safety and Health Administration (OSHA). Permissible expo-
sure limits – annotated tables. https://www.osha.gov/dsg/anno
tated-pels/index.html (2017, accessed 10 September 2018).