The perpetuation of the misconception that rats receive a 3–5 times lower lung tissue dose than humans at the same ozone concentration

Abstract

This paper highlights the pervasive misconception concerning 1994 findings from Hatch et al. about ozone (O₃) tissue dose in humans versus rats. That study exposed humans to 0.4 ppm and rats to 2 ppm ¹⁸O-labeled O₃ and found comparable incorporation of ¹⁸O into bronchoalveolar lavage constituents. However, during O₃ exposure humans were exercising, which increased their ventilation rate five-fold, while rats were at rest. This resulted in similar O₃ tissue doses between the two species, and predominantly explained the comparable ¹⁸O incorporation at five-fold different concentrations. The five-times higher exercising human inhalation rate offset the five-times lower concentration, producing the same human dose expected at rest at 2 ppm (i.e. 0.4 ppm × 4686 L/2 hour ≈ 2 ppm × 998 L/2 hour). In 2013, Hatch et al. showed that resting humans and resting rats experienced fairly comparable ¹⁸O incorporation at the same O₃ exposure concentration and activity state into BALF cells. Despite these findings, we show here that in the peer-reviewed literature a substantial proportion of researchers continue to perpetuate the misunderstanding that human lung tissue doses of O₃ are simply 3–5 times greater than rat doses at the same O₃ concentration, due to interspecies differences, and not considering activity state. It is important to correct this misconception to ensure an appropriate understanding of the implications of O₃ studies by the scientific community and policy experts making regulatory decisions (e.g. the US Environmental Protection Agency’s National Ambient Air Quality Standards for O₃).

Keywords:

• Ozone
• environmentally-relevant
• respiratory
• tissue doses
• human equivalent concentration

Introduction

Ozone (O₃) is a photochemical air pollutant that arises secondarily from a series of complex reactions in the troposphere between nitrogen oxides, volatile organic compounds, and the ultraviolet spectrum of sunlight (Costa, 2008a). Historically, toxicological literature has shown that relatively high levels of O₃ (>1 parts per million) contribute to laboratory animal morbidity and mortality (Stokinger, 1957), and in humans lower levels can cause diminished lung function (Schelegle, 2001) and may exacerbate asthma (Figure 1). O₃ is one of six “criteria” air pollutants for which regulatory standards have been set by the US Environmental Protection Agency (USEPA) through the National Ambient Air Quality Standards (NAAQS) program. The O₃ NAAQS level is currently set at 0.070 parts per million (ppm) for an annual fourth-highest daily eight-hour maximum, averaged over three years. Historical values for the O₃ NAAQS have been 0.12 ppm for a one-hour maximum averaging time, and 0.08, 0.075, and currently 0.070 ppm for an eight-hour maximum averaging time. Typically, the justifications for these NAAQS have been an extrapolation from available controlled O₃ exposures of exercising human subjects, with support from human epidemiology and laboratory animal toxicity studies (USEPA, 2013). These animal toxicological studies utilize relatively high exposure levels of O₃ (0.2–3 ppm) to demonstrate a range of acute and chronic inhalation toxicities.

Figure 1. O₃ concentration-response from animal studies – the dose makes the poison. O₃ concentration (in ppb) versus severity of response Depuydt et al., 1999; Dye et al., 1999; Mittler et al., 1956; Scheel et al., 1959; Schelegle et al., 2001; Stokinger, 1957.

By contrast, according to the latest O₃ Integrated Science Assessment (ISA; USEPA, 2013), from 2007 to 2009 the median eight-hour daily maximum concentration across all US sites was 0.040 ppm, which is 5–75 times lower than the animal toxicity study concentrations mentioned above. While the ISA (USEPA, 2013) generally considers pollutant concentrations within one or two orders of magnitude of ambient conditions to be relevant, experimental animal exposure concentrations appreciably higher than those experienced by humans are not environmentally relevant or particularly useful in identifying health effects likely to occur in humans exposed under ambient conditions. This is because the mechanisms that cause toxicity at high concentrations may be entirely different than those operating at lower concentrations. This is the concept of dose-dependent transitions in mechanisms of toxicity (Slikker et al., 2004). For this reason, a correct understanding of the relevance of experimental animal O₃ exposure concentrations to humans exposed to O₃ environmentally (i.e. cross-species extrapolation) is crucial.

In speaking with O₃ researchers about their work, it became apparent that there is a common misconception about what O₃ concentrations produce equal lung tissue doses in rats and humans. Some researchers seem to erroneously believe that simply due to interspecies differences, rats must be exposed to O₃ concentrations 3–5 times higher than humans to produce the same dose to the lungs. The misconception is that human lung tissue doses of O₃ are 3–5 times greater than rat lung tissue doses at the same O₃ concentration due to interspecies differences in dosimetry. Furthermore, in speaking with O₃ researchers, it became apparent that the basis of this misconception (and its perpetuation) lies in a narrow interpretation of the findings of Hatch et al. (1994).

Briefly, Hatch et al. (1994) exposed eight male human volunteers (18–35 years of age) to 0.4 ppm ¹⁸O-labeled O₃ (¹⁸O₃) for two hours with intermittent 15-minute periods of heavy treadmill exercise. Additionally, two dose groups of resting male rats (n = 19 and 20) were exposed to 0.4 and 2 ppm ¹⁸O-labeled O₃ for two hours, respectively. Tissue dose of O₃ was assessed using microgram per gram (μg/g) ¹⁸O incorporation into bronchoalveolar lavage fluid (BALF) constituents from rats and humans collected 0–1 hour after exposure. BALF is a well-established specimen that can safely be collected, and has been used for dose–response studies (e.g. release of inflammatory cells into the BALF). Hatch et al. considered the BALF cells and surfactant to have an alveolar origin, based on histological evidence, so in the following text we refer to this as the alveolar region (Hatch et al., 1994, 2013; Pinkerton et al., 2015). In general, BALF found in distal airspaces (i.e. alveolar region) in the lower part of the lung collected from healthy, never-smoker individuals has a characteristic profile of cellular and acellular constituents. Likewise, it is common for acutely and chronically diseased individuals (e.g. acute interstitial pneumonia, diffuse alveolar damage, sarcoidosis) to have a characteristic profile of cellular and acellular constituents found in their BALF as well (Meyer & Raghu, 2011). Hatch et al. (1994) found that exercising humans had 4–5 times higher ¹⁸O concentrations in their BALF constituents than did resting rats at the same O₃ concentration of 0.4 ppm, with the implication being that O₃ toxicity in resting rats underestimates dose in exercising humans. Citing Hatch et al. (1994), Barreno et al. (2013) correctly reported that the dose of O₃ delivered to the lower lungs is the product of O₃ concentration, exposure time, and minute ventilation (V_E); thus, exercising during exposure increases the V_E of subjects and subsequently, the total dose of O₃ delivered to the lungs. However, many other researchers do not caveat nor document the importance of the difference in activity state when citing Hatch et al. (1994), but rather simply imply that rats, across the board, must be exposed to a 3–5 times higher O₃ concentration than humans to achieve an equal dose, due to interspecies differences in dosimetry. This is in contrast to Hatch et al. (1994), who suggested that exercise may be more important than species differences in determining alveolar O₃ dose. Consistent with this conclusion, the 4–5 times higher alveolar O₃ dose observed in exercising humans compared to resting rats in Hatch et al. (1994) is most simply explained primarily by the approximately five-fold increase in ventilation rate in the human subjects induced by exercise. Hatch et al. (2013) repeated this experiment with humans at rest, and found comparable incorporation of ¹⁸O into BALF cells between rats and humans exposed to 0.4 ppm ¹⁸O₃. They found twice as much incorporation of ¹⁸O into the BALF surfactant in humans compared to rats, but the implications of this for lung cell dose is uncertain, and the authors, as well as ourselves, chose to use the BALF cell (not surfactant) ¹⁸O incorporation as the measure of alveolar tissue dose. Hatch et al. (2013) noted that the BALF cell ¹⁸O incorporation was five-times higher in the exercising human subjects, compared to those at rest. They were exercising intermittently for two hours and breathed a total of 4686 L (i.e. 3876 L for 60 total minutes of intermittent exercising + 810 L for 60 total minutes of intermittent rest between exercise periods = 4686 L total for the two-hour intermittent exercise scenario; note the addition error in Table 3 of Hatch et al., 2013) compared to only 998 L total for two hours at rest, resulting in a 4.7-fold difference and almost perfectly explaining the difference in alveolar tissue ¹⁸O incorporation.

In this paper, we investigated the prevalence of the misconception that rats must simply be exposed to a concentration of O₃ 3–5 times higher than humans to achieve a comparable tissue dose, through a search of the scientific, peer-reviewed literature. We present here the finding that this misconception is quite prevalent among those who cite the work of Hatch et al. (1994).

Methods

To assess the pervasiveness of the misunderstanding that rats must be exposed to a concentration of O₃ 3–5 times higher than a human in order to achieve a similar alveolar tissue dose, we first searched for the Hatch et al. (1994) manuscript in Google Scholar (17 May 2016) and used the “cited by” function to identify all of the published papers that cited the Hatch et al. (1994) manuscript. These papers comprised a broad spectrum of studies (i.e. in vitro to in vivo), which were further screened to exclude those articles that did not specifically address the comparability of/differences in rat versus human tissue doses (e.g. referenced Hatch et al. for methods only) (Figure 2). Then the papers were categorized as “Correct”, “Incorrect” or “Uncertain” (Table 1) based on their use/interpretation of the findings of Hatch et al. (1994) regarding rat versus human alveolar doses. “Correct” articles were those wherein study text reflected a proper understanding of the influence of activity state in discussing species differences in alveolar dose at given exposure concentrations (e.g. studies that had an appreciation for increases in ventilation rate and corresponding alveolar dose due to exercise, and/or acknowledged the importance of considering rest versus exercise when comparing doses across species, either explicitly or simply through appropriately caveated text). For example, the following sentence is properly caveated as to activity state (although “effects” should arguably more appropriately be “doses”), and so is deemed to reflect a “correct” understanding of Hatch et al. (1994), “The effect of O₃ exposures in resting rats underestimate those observed in exercising human” (Pryor et al., 1996). “Incorrect” articles were ones wherein study text perpetuated an iteration of the misunderstanding that the human alveolar dose, across the board, is simply 3–5 times greater than rat tissue doses at the same O₃ concentration (e.g. Miller et al., 2016). The authors ascribed this to some difference (e.g. interspecies, sensitivity, age, cardiac condition) in dosimetry, did not explicitly or implicitly consider activity state as a factor (e.g. Stiegel et al., 2016, Table 1), and/or assumed comparability/equivalency of rodent exposure at rest to exercising human exposure at the same concentration. For some articles, it was not clear as to whether there was a proper appreciation of the influence of activity state on alveolar dose in different species. Thus, these studies were deemed “uncertain”. For example, in discussing their rodent study exposure concentration, Elder et al. (2000) indicate, “The exposure concentration of O₃ (1 ppm) is high, but is similar to an exposure to 0.25 ppm in an exercising human (Hatch et al., 1994)”. Given that rodents in the Elder et al. study were not made to exercise, for purposes of the current study it was assumed that this sentence is implicitly discussing resting rodent versus exercising human dose, although this was not explicitly stated by the authors (i.e. 1 ppm not caveated by “at rest”). The prevalence of the three categories was then calculated as the percent of published papers citing Hatch et al. (1994) regarding rat versus human alveolar dose with either a “correct”, “incorrect” or “uncertain” interpretation (Table 1). We assessed all included studies based on these criteria, regardless of whether the subject of the study’s investigation was rats or other mammals, or was even an in vitro study. That is because we were assessing the prevalence and perpetuation of the stated misconception in the literature, not how often this misconception was used to explicitly choose O₃ concentrations in a rat study.

Figure 2. Database search results.

Table 1. Dose-related quotes from papers referencing Hatch et al. (1994).

Ozone studies	Dose-related	Correct	Incorrect	Uncertain
Pryor et al. (1996) (p. 155)	“However, the effect of O3 exposures in resting rats underestimate those observed in exercising human”.	✓
Zhao et al. (1998) (p. L45)	“In as much as rodents are less sensitive to O3 than humans by a factor of four to five, 0.6 ppm may reflect a human response at 0.12–0.15 ppm, a concentration comparable to the NAAQS”.		✘
Elder et al. (2000) (p. 229)	“The exposure concentration of O3 1 ppm is high, but is similar to an exposure to 0.25 ppm in exercising humans (Hatch et al., 1994) and was chosen to mimic high episodic concentrations in urban air …”.			?
Huffman et al. (2001) (p. 24)	“In addition, the biologic effects of a short-term exposure to 2 ppm O3 in rats appear to be relatively equivalent to exposure of 0.4 ppm O3 in exercising human subjects”.	✓
Kleeberger et al. (2001) (p. 719)	“It is important to note that actual delivered dose of O3 to the lung of rodents may be considerably less than may be expected by concentration × time considerations in human subjects, because rodents are obligate nose breathers”. AND “Support for the enhanced scrubbing efficiency of upper airways of rodents compared with humans come from Hatch et al. (1994)”.		✘
van Bree et al. (2001) (p. 704)	“Interspecies comparisons for the modeled or measured delivered O3 tissue dose in the respiratory tract indicate (1) that total uptake efficiency in rats is larger than in humans, (2) that there is a similarity between rats and humans for the D-R pattern for both acute inflammation and chronic alveolar interstitial thickness and (3) that humans may be more responsive than rats. (Hatch, 1994;…) These findings indicate that rats may be considered as a sensitive human model for O3 toxicity and that effects observed in rats may very well occur in humans, even at lower O3 levels (Hatch, 1994;…)”.		✘
Kumarathasan et al. (2002) (p. 196)	“The dose of ozone in the lung parenchyma of resting rats is about four to five times lower than the dose in the parenchyma of an exercising human subject”. “In other words, the ozone mass transfer rate in the lung parenchyma of our animals should approach realistically the ozone mass transfer rate in a human lung in a scenario of high environmental exposure such as jogging during an episode of severe air pollution”.	✓
Soulage et al. (2004) (p. 983)	“The effects observed in exercising humans are very similar to that produced in rat exposed at rest to a five-fold higher O3 concentration“.			?
Thomson et al. (2004) (p. 80)	“…the known interspecies differences in sensitivity to air pollutants (with humans being more responsive than rats …)”.		✘
Nadadur et al. (2005) (pp. 1719–1720)	“…, we have shown that the impact of acute exposure to O3 at 0.4 ppm with intermittent heavy exercise in humans resulted in lung tissue dosimetry approximately equal to that of the rat exposed sedentary to 2 ppm for the same 2-hr period”.	✓
Tsujino et al. (2005) (p. 483)	“For instance, Hatch et al. (1994) reported lower concentrations of O^18-labeled O3 in BAL constituents of a rat than those of a human following inhalation of the same concentration”.		✘
Huffman et al. (2006) (p. 125)	“Thus the dose used in this study is a biologically relevant human ozone dose”.		✘
Cho et al. (2007) (p. 830)	“…dosimetry studies in which rodents require 4 to 5 fold higher doses of O3 than humans to create an equal deposition and pulmonary inflammatory response”.		✘
Costa (2008a) (Book Ch. 28)(p. 1144)	“It is important to consider the role of exercise in a study of O3 or any inhalant before making cross-study comparisons, especially if that comparison is across species”.	✓
Wagner et al. (2007) (p. 1177)	"Pulmonary O3 dosimetry studies demonstrate that humans have 4–5 times greater pulmonary deposition of inhaled O3 than Fisher F344 rats. Furthermore, the Brown Norway rat is much less sensitive to O3 than F344 rats. As such, 1 ppm O3 in BN rats provides a reasonable exposure scenario to what might occur in humans”.		✘
Costa (2008b) (p. 108)	“…while the rats were exposed only at rest to the same O3 concentration or that concentration estimated to equate the total inhaled under exercise-induced ventilation. Reported that authors found that dose to the cells was similar between the human and rat only if the exposure concentration to the rat were increased by a factor of 5. Hence a simple use of exposure concentration would distort comparability of dose and effect”.	✓
Crémillieux et al. (2008) (p. 776)	“Based on comparative dose-effect studies in humans and rats, 0.5 ppm O3 exposure in rats can be extrapolated in to 0.1 ppm (200 μg/m^3) O3 concentration exposure in human”.		✘
Boussouar et al. (2009) (p. 76)	“Although this level of O3 is about 2X as high as that measured in polluted urban areas, the rat respiratory system is less sensitive to O3 than that of humans and a concentration of at least 0.5 ppm is required to mimic the pathological effects on the human lung”.		✘
Haque et al. (2009) (p. 2)	“In general, experimental animals require significantly higher doses of O3 exposure than humans to reach comparable amounts of O3 concentration in the distal lung”. AND "Therefore, it is necessary that rodents be exposed to high O3 concentration to better enable extrapolation of findings from animal studies to human”.		✘
Perpu et al. (2010) (pp. 62–63)	“Interspecies comparisons of O3 exposure have led to the following: (1) the total uptake efficiency in rats is larger than in humans; (2) there is a similarity b/w rodents and human for the D-R pattern for both acute inflammation and chronic alveolar interstitial thickness; (3) humans may be more responsive than rodents. These findings indicate that rodents may be considered as a sensitive human model for O3 toxicity and that effects observed in rodents may very well occur in humans, even at lower O3 levels”.		✘
Gackière et al. (2011) (p. 962)	In reference to using 0.5 and 2 ppm: " …this choice was justified when considering that O3 toxicity in resting rats is known to underestimate effects in exercising humans. Moreover, our primary goal was not to mimic air pollution peaks, but rather to provide fundamental data to describe the neural pathways activated by O3"	✓
Shore et al. (2011) (p. 7)	“…concentration is higher than the ambient O3 concentrations that would be observed even in the most polluted cities…; because rodents are much less susceptible to O3 than humans …; rodents require much higher concentrations of O3 to achieve the same O3 dose as humans…"		✘
Barreno et al. (2013) (pp. L126–L127)	“As the dose of O3 delivered to the lungs is the product of O3 concentration, exposure time, and minute ventilation (VE), exercising during exposure increases the VE of these subjects and, subsequently, the total dose of O3 delivered to their lungs”.	✓
Cho et al. (2013) (pp. 6 and 10)	“…results from dosimetry studies in which rodents require 4–5-fold higher doses of O3 than humans to create an equal deposition and pulmonary inflammatory response, either level of O3 used in the current study is a reasonable exposure level which is comparable with human exposures”.		✘
Hatch et al. (2013) (p. 63)	“Results confirm the use of VE as a factor in the extrapolation of inhaled dose of O3 at different levels of physical activity, and suggest that higher and more continuous activity levels will yield significant effects at even lower ambient levels of O3”.	✓
Martinez-Lazcano et al. (2013) (p. 346)	“Hatch et al. (1994) demonstrated that rodents require much higher doses compared with the exposure doses reported in humans in which neurological disorders due to O3 exposure were observed…”		✘
Oakes et al. (2013) (p. 31)	“Although our study used a high concentration of O3, exposure to 2 ppm O3 is comparable to 0.2 ppm O3 in a susceptible human subject”		✘
González-Guevara et al. (2014) (p. 1091)	“In animal models, the O3 is usually administered in high doses, i.e. the rats need be exposed to 2.0 ppm O3 in order to achieve immunological changes in the bronchoalveolar lavage fluid (BALF) as those found in human exposed to 0.4 ppm (Hatch et al., 1994)”		✘
Mikerov et al. (2014) (p. 334)	“The rationale for this dose is based on a study by Hatch et al. (1994) in which they reported that experimental animals required higher O3 concentrations than human to deliver comparable amounts of O3 to the distal lung”.		✘
Dye et al. (2015) (p. 23)	“First, the concentrations of O3 used were higher than ambient levels in most metropolitan areas. However, previous investigations have shown that obligate nasal-breathing, diurnally exposed, non-exercising rats require ≥5-fold higher O3 concentrations in order to achieve comparative O3 lung deposition (and oxidant-induced lung effects) compared with mouth-breathing humans undergoing intermittent exercise during experimental O3 exposure”.	✓
Miller et al. (2015) (p. 78)	“The 1 ppm O3 concentration used in this study for most analysis is considered high compared to human relevant dose. However, it has been shown that rats require 3–4Xs the concentration of O3 to acquire an equivalent dose to humans (Hatch, 1994). Thus, the concentrations of O3 used in our study, although higher than ambient levels, are appropriate for rat exposure”.		✘
Vella et al. (2015) (p. 1021)	“As a consequence, the toxicity of O3 observed for a given concentration in rats strongly underestimates the effect observed for the same concentration in humans, and thus, the effects observed in human are comparable to those produced in rats exposed to a five-fold higher O3 concentration”.		✘
Ward et al. (2015) (p. 85)	“Since exercise might increase lung deposition in humans inhaling 0.4 ppm O3 (Hatch et al., 2013), and it has been predicted that humans retain higher dose of O3 upon inhalation than rodents (Hatch et al., 1994), we believed that the concentration of 1.0 ppm in the present study was suitable for examining gene expression profiling”.		✘
Yang et al. (2015) (p. 572)	“That a 3-ppm inhaled dose in rodents results in O3 concentration in the lungs relevant to human exposure levels has been experimentally also validated by other by using oxygen-18-labeled O3. For example, in a study by Hatch et al., it was demonstrated that exposure to ^18O3 (0.4 ppm for 2 hours) caused 4- to 5-fold higher ^18O3 concentrations in human subjects than in rats. Rats exposed to 2.0 ppm had still less ^18O3 in BALF than human subjects exposed to 0.4 ppm”.		✘
Miller et al. (2016) (p. 314)	”It has been shown that in rats once inhaled, O3 deposition in the lung is 3–4 times less that what humans will experience (Hatch et al., 1994)”.		✘
Ong et al. (2015) (p. 332)	“For example, it takes approximately 4 to 5 times the concentration of O3 to induce pulmonary inflammation in rodents equivalent to that documented in exercising human subjects under similar controlled exposure conditions”.	✓
Snow et al. (2016) (p. 320)	"It is also known that the dose encountered in clinical studies involving intermittent exercise during a 0.2–0.4 ppm O3 exposure in humans results in a dose that is 3–4 times that experienced by non-exercising rodents (Hatch et al., 1994, 2013), indicating that although our high dose is not necessarily environmentally relevant, it is appropriate for examining O3-induced toxicity in a rodent model”.	✓
Stiegel et al. (2016) (p. 16)	“…used a high concentration of O3" AND "exposure to 2 ppm of O3 in a rodent was shown to be comparable to 0.2 ppm of O3 in a susceptible human subject”.		✘
Zhong et al. (2016) (p. 384)	“It has been experimentally documented that it takes approximately 4 to 5 times the concentration of O3 to induce pulmonary inflammation in laboratory rodents that is similar to that in exercising human subjects under similar controlled acute exposure conditions (Hatch et al., 1994, 2013). The airborne concentration that was chosen for this exposure was equivalent to an exercising human exposure of 0.15 ppm or approximately twice the concentration of the current U.S. National Ambient Air Quality Standard (NAAQS) for O3 (0.070 ppm)”.	✓

Results

In total, 127 articles were identified that referenced the Hatch et al. (1994) study. Of these, 88 articles referenced Hatch et al. but did not address interspecies differences in alveolar dose specifically. For example, this includes referencing for methods, experimental design, and/or the analysis/incorporation of ¹⁸O into the BALF and its constituents, and so were unrelated to the question at hand and were excluded from further analysis. Thirteen of the remaining 39 articles (33%) properly referenced Hatch et al. (1994) in regards to the importance of activity state/ventilation rate (i.e. exercising humans versus resting rats) in estimating relevant alveolar doses (Table 1). Twenty-four articles (62%) that referenced Hatch et al. (1994) perpetuated the misconception that human alveolar doses of O₃ should simply be expected to be 3–5 times greater than rat doses at equal exposure concentrations (Table 1). In addition, one article, Oakes et al. (2013), went as far as to say that a rat O₃ exposure concentration of 2 ppm was “comparable” to a human exposure concentration of 0.2 ppm in susceptible human subjects. It is not clear how the authors reached this conclusion (i.e. a 10-fold higher rat exposure concentration is comparable), but they cited Hatch et al. (1994). The two remaining articles (5%) were deemed not clear in the researchers’ citation/interpretation of Hatch et al. (1994) and were placed in the “uncertain” category (Table 1). Overall, the available articles included both in vivo (i.e. human, rats, mice, monkeys) and in vitro assays, and covered a range of dates from 1998 to 2016.

Our review found that most authors who cited Hatch et al. (1994) for their findings of ¹⁸O incorporation in rats versus humans used it to inappropriately explain or justify an O₃ animal exposure scenario based on the misconception (i.e. interspecies differences in O₃ dose, with humans experiencing higher doses than rats). As the studies did not use a different metric of tissue dose and merely used concentration as a surrogate, results for/discussion of other dose metrics is not relevant for the limited purpose of this paper.

Discussion

We found that a misunderstanding/misapplication of the findings of Hatch et al. (1994) is pervasive, with researchers not properly understanding and/or appropriately considering that the original Hatch et al. (1994) study exposed humans and rats under very different activity states (i.e. resting versus exercising), which explains the differences in measured alveolar dose. This misconception occurred in 62% of the studies identified (24 of 39 studies; Table 1), published up to present day. Thus, these studies continue to perpetuate the misconception that rats receive a 3–5 times lower alveolar dose than humans at the same O₃ concentration simply due to interspecies differences in dosimetry.

Although Hatch et al. (2013) specifically demonstrated that resting human subjects achieve a BALF cell ¹⁸O₃ dose (5.6 ± 1.7 μg/g dry weight) somewhat (i.e. 25%) lower than that of resting rats (7.5 ± 1.6 μg/g dry weight) at the same exposure concentration (and much lower than exercising humans), with two-fold higher ¹⁸O incorporation into human BALF surfactant (Table 4 of Hatch et al., 2013), the paper apparently accomplished little in terms of correcting the common misunderstanding based on the previous paper (Hatch et al., 1994), as evidenced by results of the current study. In our findings, more than half of the papers did not account for the critical consideration that Hatch et al. (1994) results from resting rats are being compared to those from exercising humans. Given the fact that exercising humans breathed approximately five times more air than those at rest (i.e. 3876 L for 60 total minutes of intermittent exercising + 810 L for 60 total minutes of intermittent rest between exercise periods = 4686 L total for the two-hour intermittent exercise scenario/998 L total for the two-hour resting scenario = 4.7-fold difference; note the addition error in Table 3 of Hatch et al., 2013) as well as other considerations (see below), it is not particularly surprising that exercising humans accumulated approximately five times more ¹⁸O in BALF cells than resting rats. Ultimately, this results in the incorrect conclusion that high, environmentally-irrelevant O₃ exposure concentration (e.g. >500 ppb) rat study results are directly relevant to environmentally-exposed humans, without accounting for human activity and simply ignoring the fact that exposure concentration, duration and exercise state are the primary determinants of the alveolar dose of O₃ in humans and rats, which are more important than potential interspecies differences in dosimetry.

In addition to the results of the Hatch et al. (2013) study itself, other considerations support that the human equivalent concentration (HEC)/effective dose corresponding to a rat study exposure concentration should not be five times lower at comparable activity levels. These considerations, which suggest that the rat may be an appropriate if not conservative model for humans (i.e. the HEC is not lower than the rat concentration), include:

• Interspecies differences in alveolar dose;

• USEPA interspecies dosimetric adjustment procedures; and

• Modeled predictions of O₃ absorption in respiratory tract regions of interest (i.e. alveolar region).

These topics and their implications for interspecies (rat versus human) dosimetric differences are briefly discussed below.

Interspecies differences in lung dose

Rats breathe more air per unit body weight (BW) than humans, which is a consideration judged by Hatch et al. (2013) to be relevant and appropriate for discussion. An example of this is provided in Appendix 2 of Hatch et al. (2013), which uses an allometric equation to calculate that the rat ventilation rate per unit BW is 2.82 times higher than that for humans. Based on this consideration, the authors acknowledge that rats should receive a higher alveolar dose and that this fact is contrary to the notion that rats may underestimate human O₃ dose. By corollary, it is also contrary to the idea that a five-fold higher exposure concentration is needed for rats to receive the same alveolar dose of ¹⁸O as humans. In fact, because rats should receive a higher alveolar dose, despite their discussion of Hatch et al. (1994) results (based on a comparison that does not consider the increased ventilation rate produced by exercise), the study authors conclude that, "The finding that human resting BALF ¹⁸O dose approximates that of the resting rat BALF ¹⁸O dose is unexpected". In other words, if any interspecies differences in dose would be expected, based on the knowledge that rats breathe more air per unit BW than humans, the study authors would expect the rat to overestimate human dose. Thus, in addition to the results of the Hatch et al. (2013) study itself, which should be given the greatest weight, this consideration judged to be relevant by Hatch et al. themselves suggests that the HEC corresponding to a rat study O₃ exposure concentration should not be lower at the same level of physical activity, much less five times lower. If anything, it suggests that humans should receive a lower dose.

Standard USEPA interspecies dosimetric adjustment procedures

This section discusses standard USEPA dosimetric adjustment method considerations to demonstrate general expectations about the interspecies extrapolation of dose. O₃ is a Category 2 gas based on its physical/chemical and toxicokinetic characteristics and their influence on sites of toxicity. More specifically, O₃ is a Category 2 gas because it has the ability to exert both portal of entry (POE) and systemic effects. For Category 2 gases, USEPA suggests different interspecies dosimetric adjustment procedures depending on where the specific adverse effect being evaluated occurred (i.e. POE versus systemic). That is, when a POE effect produced by a Category 2 gas is being evaluated, the default dosimetric adjustments for a Category 1 gas (which tend to be highly reactive and undergo rapid, irreversible reactions in the respiratory tract) are recommended, while the dosimetric adjustments for a Category 3 gas are recommended when a Category 2 gas has produced a systemic effect (pp. 4–17, USEPA, 1994). Accordingly, when a POE effect is induced in the pulmonary region, whether by a Category 1 or Category 2 gas, the dosimetric adjustments for a Category 1 gas are used. Thus, when O₃ exerts POE effects, the interspecies dosimetric adjustment procedure used is that for a Category 1 gas.

In general, for pulmonary region effects (from the respiratory bronchioles to the alveoli) caused by either a Category 1 or 2 gas, the HEC is calculated as the animal (e.g. rat) exposure concentration multiplied by the regional gas dose ratio for the pulmonary region (RGDR_PU). The RGDR_PU accounts for animal-to-human differences in deposition and absorption within the pulmonary region. USEPA (2009) provides a rat-to-human RGDR_PU of 3.5. This indicates that not only is the HEC generally not five times lower than the rat exposure concentration for pulmonary effects, it is generally expected to be approximately 3.5 times higher. Thus, for example, the HEC that would correspond to a rat exposure concentration of 400 ppb for pulmonary effects would be approximately 1400 ppb using this standard procedure (i.e. HEC = animal exposure concentration × RGDR_PU = 400 ppb × 3.5 = 1400 ppb). Simply being aware of typical rat-to-human RGDR_PU values should lead a researcher to question and critically evaluate the basis for any expectation of a five-fold lower HEC, which would correspond to an RGDR_PU of 0.2 (e.g. HEC = animal exposure concentration × RGDR_PU = 400 ppb ×0.2 = 80 ppb, which would account for a five-fold higher dose in humans compared to rats). Thus, consideration of standard USEPA interspecies dosimetric adjustment procedures for POE effects suggests that the HEC corresponding to a rat study exposure concentration should be higher than the rat concentration, not lower, much less five times lower.

Modeled predictions of O₃ absorption

Furthermore, while the above discussion on interspecies dosimetric adjustment procedures is general in nature, O₃ absorption has been modeled in both humans and rats. Tables 3–5 of USEPA (2012) provides such modeling results with the aim of simulating how interspecies anatomical and physiological differences influence the transport of O₃ throughout the airways and alveoli. Briefly, contrary to the notion that rats may underestimate human dose by five-fold at the same exposure concentration, the modeling results showed that total absorbed amount of O₃ (g/kg BW) was found to be 30% higher in rats. More specifically and importantly, the amount of O₃ absorbed per surface area and unit time (g/cm²/min) in the alveolar region was reported to be slightly higher (11%) in the rat, although still comparable to that in humans. Similar to other considerations (e.g. empirical data from Hatch et al., 2013), this information suggests that despite previous erroneous notions about the rat underestimating human alveolar O₃ dose, the rat may be an appropriate if not conservative model for human alveolar dose. That is, the relevant considerations and information discussed above do not support that the HEC corresponding to a rat study exposure concentration should be lower than the rat concentration, much less five times lower.

Conclusions

In conclusion, our experience and research indicate that:

• It is common for researchers to assert that the HEC/effective dose for O₃ is 3–5 times lower than the rat exposure concentration (i.e. that rats must simply be exposed to O₃ concentrations 3–5 times higher than humans to produce the same dose to the lower lung);

• As a result, this misconception has essentially become “common knowledge” among O₃ researchers, even though it is erroneous;

• This erroneous assertion appears to have originated and persisted due to an inappropriate/incongruous comparison of study results for exercising humans versus resting rats (Hatch et al., 1994), which is often insufficiently caveated when cited or in some instances has not been made clear; and

• The misconception is now prevalent in the scientific literature.

This frequently accepted but erroneous notion is entirely inconsistent with expectations about potential interspecies differences in O₃ dose at the same air concentration, as well as with currently available data. These expectations are based on considerations of interspecies differences in alveolar dose, USEPA dosimetric adjustment procedures, and modeled predictions of O₃ absorption. More importantly, the erroneous notion on the part of 62% of the research papers citing Hatch et al. (1994) in regard to the HEC for a given rat O₃ exposure concentration being five-fold lower is contrary to the Hatch et al. (2013) demonstration of comparable ¹⁸O incorporation into alveolar BALF cells in resting humans and rats. Inopportunely, it appears that the comparison of mismatched exposure scenario results (exercising humans versus resting rats) from Hatch et al. (1994) has also resulted in many researchers incorrectly concluding that high, environmentally-irrelevant O₃ exposure concentration (e.g. >500 ppb) rat study results are relevant to environmentally-exposed humans, without considering physical activity. Thus, based on our review this misunderstanding is prevalent in the scientific literature and regrettably will be difficult to correct. Having said that, it is important for the scientists and policy makers responsible for making regulatory decisions (e.g. setting the O₃ NAAQS) to have a more refined and accurate understanding of the applicability of doses used in O₃ studies.

Disclosure statement

No potential conflict of interest was reported by the authors.