Exhaled Nitric Oxide and Pulmonary Oxygen Toxicity Susceptibility

Individual susceptibility to pulmonary oxygen toxicity (PO2tox) is highly variable and currently lacks a reliable biomarker for predicting pulmonary hyperoxic stress. As nitric oxide (NO) is involved in many respiratory system processes and functions, we aimed to determine if expired nitric oxide (FENO) levels can provide an indication of PO2tox susceptibility in humans. Eight U.S. Navy-trained divers volunteered as subjects. The hyperoxic exposures consisted of six- and eight-hour hyperbaric chamber dives conducted on consecutive days in which subjects breathed 100% oxygen at 202.65 kPa. Subjects’ individual variability in pulmonary function and FENO was measured twice daily over five days and compared with their post-dive values to assess susceptibility to PO2tox. Only subjects who showed no decrements in pulmonary function following the six-hour exposure conducted the eight-hour dive. FENO decreased by 55% immediately following the six-hour oxygen exposure (n = 8, p < 0.0001) and by 63% following the eight-hour exposure (n = 4, p < 0.0001). Four subjects showed significant decreases in pulmonary function immediately following the six-hour exposure. These subjects had the lowest baseline FENO, had the lowest post-dive FENO, and had clinical symptoms of PO2tox. Individuals with low FENO were the first to develop PO2tox symptoms and deficits in pulmonary function from the hyperoxic exposures. These data suggest that endogenous levels of NO in the lungs may protect against the development of PO2tox.


Introduction
Pulmonary oxygen toxicity (PO 2 tox) results from prolonged exposure to a hyperoxic atmosphere, with the severity of symptoms increasing progressively with elevation of the inspired oxygen partial pressure (PiO 2 ) and the duration of exposure [1].Symptoms of PO 2 tox include chest pain, tightness, cough, and substernal distress that may coincide with decreases in pulmonary function, specifically, a reduction in forced vital capacity (FVC) and alveolar diffusion capacity (D L CO) [1,2].The toxic effects of oxygen are a concern for military and technical divers conducting prolonged multiday dives using oxygen rebreathers and for patients undergoing hyperbaric oxygen therapy or aggressive oxygen therapy for respiratory insufficiency at normobaric pressure.While there are theoretical models that predict the expected level of pulmonary function deficit because of prolonged exposure to raised PiO 2 that are based upon the expected decline in FVC, there is considerable individual variation in susceptibility to a uniform degree of pulmonary oxygen poisoning [3][4][5].Currently, there are no methods to predict individual susceptibility to PO 2 tox.Furthermore, a sensitive non-invasive biomarker that can detect changes in lung pathology at an early stage in the oxygen toxicity process has remained elusive.
Expired nitric oxide (F E NO) measurements have been studied as an exhaled marker of airway inflammation in a variety of lung diseases, including asthma, lung cancer, bacterial pneumonia, pulmonary fibrosis, and idiopathic pulmonary fibrosis [6,7].Nitric oxide (NO) The current study investigated individual differences to hyperbaric oxygen (HBO) stress using a small group of healthy, well-trained divers, rather than focusing on group mean changes in a larger, more diverse subject population.Consequently, the subject population was limited to qualified U.S. Navy-trained divers who were fit to dive and familiar with the signs and symptoms of both pulmonary and central nervous system (CNS) oxygen toxicity.Subjects were drawn from a convenience sample of U.S. Navy-trained divers (male or female) that were stationed at or worked at the Naval Submarine Medical Research Laboratory (NSMRL).During the informed consent process, all divers were reminded of the risks of pulmonary and CNS oxygen toxicity that could result from their participation in the study.Eight male U.S. Navy-trained divers aged 21-55 years (mean = 36.4years), weighing 74.1-113.2kg (mean = 91.8kg), and with body stature ranging from 165-180 cm (mean = 174 cm) participated as subjects after signing the informed consent.The subject population consisted of enlisted divers and undersea medical officers, as well as civilian U.S. Navy-trained divers.All had normal lung function, were non-smokers, and abstained from all other diving activities for the duration of the study.

Study Design and Hyperbaric Oxygen Exposure Profile
The study protocol was approved by the NSMRL Institutional Review Board in compliance with all applicable federal regulations governing the protection of human subjects.The HBO exposures consisted of six-hour and eight-hour dry resting trials, breathing 100% humidified O 2 at 202.65 kPa (2 ATA) in a hyperbaric chamber.The six-and eight-hour dives were performed on consecutive days.Pulmonary function and F E NO were measured immediately prior to each dive, between 10-and 60-min post-dive, and then daily for at least 3 days after the dive, until complete recovery of pulmonary function.Only subjects who showed no decrements in pulmonary function following the 6 h exposure conducted the 8 h dive.Based on previous studies, the level of pulmonary oxygen toxicity induced by 6 h of breathing 100% O 2 at 2 ATA was predicted to cause a temporary decrease in FVC of between 4 and 6 percent in 50% of the subjects [2][3][4]14].Extending the dive to 8 h increased the predicted decrement in FVC to 8% in 50% of the subjects [14].Both dives were below the CNS oxygen toxicity limit (previous studies have shown no evidence of CNS oxygen toxicity in divers exposed to 2 ATA of oxygen for up to 12 h [3]).Due to the level of uncertainty in these predictions, it was felt that the current approach of conducting two dives with increasing length of oxygen exposure would permit measurable but fully reversible levels of pulmonary oxygen toxicity in our subject population, without exposing particularly susceptible individuals to an overly long HBO exposure.
During each dive, an inside tender who was not on oxygen accompanied the diver (subject).The hyperbaric oxygen exposure profiles were carefully selected to elicit a mild but reversible level of pulmonary oxygen toxicity in the majority of subjects while also keeping the risk of a seizure from CNS oxygen toxicity to a minimum.A single 15 min air break was incorporated during the midpoint of each HBO exposure, during which the subjects ate a low-nitrate/-nitrite lunch.The total bottom time of the dive was adjusted for the 15 min air break to ensure that the total time breathing 100% oxygen at 2 ATA was either 6 or 8 h.The initial six-hour HBO exposures were conducted in two teams of three and one team of two divers.All the dives were conducted at the same time of day (initial press between 07:10 and 08:36).Compression and decompression rates were 18.3 msw/min (60 fsw/min) and 9.1 msw/min (30 fsw/min), respectively.
Control exposures were conducted on two of the subjects to determine if pulmonary function or F E NO were significantly affected by breathing air at surface pressure in the hyperbaric chamber on the built-in breathing system for six or eight hours.One subject completed a six-hour air exposure, and the other subject completed an eight-hour air exposure.Both subjects completed the control exposures approximately four months after their oxygen exposures.

Pulmonary Function and Expired Nitric Oxide Measurements
Pulmonary function (FVC, forced inspiratory vital capacity [F I VC], forced expiratory volume in 1 s [F E V1]), the diffusion capacity for carbon monoxide (D L CO), and F E NO baseline measurements were collected from each diver twice per day (morning-am, and afternoon-pm) for five consecutive days before conducting the HBO exposures.During each measurement session, subjects conducted three repetitions for each pulmonary function test, meaning that the test met the American Thoracic Society (ATS) standards for repeatability [15][16][17][18].All pulmonary function tests were conducted on the VMAX Encore 22 Pulmonary Function Module (Viasys Healthcare Inc., Yorba Linda, CA, USA).F E NO was measured using a chemiluminescence NO analyzer (Sievers NOA 280i, GE analytical instruments, Boulder, CO, USA).During each measurement session, F E NO was measured at the following 5 expired flow rates: 50, 100, 150, 200, and 250 mL/s.These were used to determine alveolar NO concentration (C A NO) and maximum airway wall flux of NO (J'awNO) using a two-compartment model [19].Exhaled flow rates for on-line F E NO measurements were controlled by having the subject target the desired flow rate, presented on a computer screen, while expiring against a flow restrictor.Five different flow restrictors were used to achieve the different expired flow rates.At each expired flow rate, the mean value from at least three F E NO measurements that conformed to the standardized procedures recommended by the American Thoracic Society for online F E NO measurement [20] were taken during each measurement session and used in the analysis.The Sievers NO analyzer and VMAX Encore 22 Pulmonary Function Module were calibrated in accordance with the manufacturer's procedures at least twice daily (morning and afternoon) before each measurement session.During each measurement session, the subjects conducted the F E NO measurements before the pulmonary function tests to avoid the potential influence of the spirometry measurements on F E NO. Pre-dive measurements for NO and pulmonary function were taken during the two-hour period before the dive.Post-dive measurements of F E NO were initiated 10 min after the dive had reached the surface.As D L CO measurements were always conducted after the F E NO measurements, subjects breathed ambient air for at least 20 to 30 min following the dives before conducting their first D L CO measurement.Consequently, P A O 2 levels were expected to be at normal levels during the pulmonary function test and, thus, no corrections were made to D L CO for P A O 2 .

Data Analysis
A decrement in pulmonary function for an individual was defined as outside their normal variability if one or more of their pulmonary function tests (i.e., FVC, F I VC, F E V1, or D L CO) fell more than two standard deviations (SD) below their mean baseline value for that test.A change in F E NO was also considered outside normal variability if the change was greater than 2 SDs from the individual's mean baseline F E NO value.Intra-individual variability for F E NO and the various pulmonary function tests were expressed as 2 × the coefficient of variation (CV), where CV = (SD/mean) × 100, to facilitate the comparison among individuals and between variables with different units and different means.All statistical analyses were carried out using Statistica software (Statsoft Inc., Tulsa, OK, USA).Analysis on the effect of time of day (am or pm) on F E NO across all expired flow rates was performed using a repeated measures analysis of variance (ANOVA).Repeated measures ANOVA was also used to compare group mean changes in F E NO, C A NO, J'awNO , and pulmonary function at the different time points.When significant main effects of time were observed, the Dunnett post hoc test was used to explore differences between baseline and post-dive values.The relationships between F E NO levels and the percent changes in pulmonary function following the six-hour dive were evaluated using linear regression and the Pearson Product moment correlation coefficient.Significance was set at p < 0.05.

Results
Baseline individual means, 2 × the coefficient of variation (2 × SD/mean × 100%) for F E NO (at 50 mL/s expired flow rate), and the different pulmonary function tests (i.e., FVC, F I VC, D L CO) that were derived from the twice daily measurements (am and pm) taken on five consecutive days (n = 20 data points per mean per subject for each pulmonary function test) are shown in the second column of Tables 1-4.The remaining columns in each table show the percent change in that variable from each individual's mean baseline level following the HBO exposures.In each of the tables, the subject's data are ordered from highest (top) to lowest (bottom) baseline F E NO.Additional tables showing changes in D L CO adjusted for Hb (D L CO adj ), D L CO adjusted for alveolar volume (D L CO/VA), alveolar volume (V A ), and F E V1 are presented in Appendix A (Tables A1-A4).
Table 1.Mean baseline levels, 2 × coefficient of variation (CV), and percent change in F E NO (expired flow rate = 50 mL/s) following the 202.65 kPa HBO exposures.The subject data (rows) are ordered from highest to lowest baseline F E NO.The baseline CV was determined from 10 measurements taken over 5 consecutive days before the dive (see Methods).Post-dive 1 and post-dive 2 measurements were taken between 15 min and 1 h post dive.Recovery measurements (Rec 1, 2, 3) were taken 1-, 2-, and 3-days post-dive.2-4).As shown in Table 1, there was a threefold range (19 to 59 ppb) in the baseline F E NO between subjects.Analysis of the F E NO baseline data using all expired flow rates indicated a significant time of day effect, with F E NO on average 10% lower in the afternoon compared to morning (p < 0.001).There was, however, no difference detected between the pre-dive F E NO taken on the morning before the six-hour dive and the baseline F E NO (p = 0.9994).Immediately following the six-hour oxygen exposure, all eight subjects had significant decreases in F E NO (i.e., values > 2 × CV less than their baseline), with the group mean change showing a 55% decrease (p < 0.0001).By the morning after the dive, F E NO levels had returned to normal in the majority of divers (six out of eight).

Subject
The four subjects with the lowest baseline F E NO and lowest post-dive F E NO (subjects 2, 3, 4, and 9) had clinical symptoms of pulmonary O 2 toxicity and showed significant decreases in pulmonary function on one or more of the pulmonary function tests immediately following the six-hour exposure (see Tables 2-4).The clinical symptoms reported included chest fullness/tightness, congestion, mild substernal burning, and tickling or cough on deep inhalation.Subjects 1, 5, 7, and 8, who had baseline F E NO levels greater than the group mean of 34 ppb, showed no pulmonary function deficits or symptoms of pulmonary O 2 toxicity following the six-hour HBO exposure and, thus, conducted the eight-hour HBO exposure the following day.Immediately following the eight-hour dive, three of these subjects had pulmonary function deficits (see Tables 2-4) and all four subjects showed greater decreases in F E NO than following their six-hour dive (mean ± SD F E NO post-dive 1 vs. post-dive 2 = 22.2 ± 3.4 ppb vs. 16.6 ± 2.7 ppb, respectively, n = 4, p < 0.01).Subject 5 had the highest baseline F E NO and was the only subject who did not show symptoms of PO 2 tox or a pulmonary function deficit following the HBO exposures.
During the three days following the dives, five subjects showed significant increases in F E NO (see Table 1).However, the timing of these increases and the duration of the elevated F E NO was variable among the subjects.Consequently, the group analysis did not reveal any statistically significant change in the mean F E NO from the pre-dive baseline during recovery days one (p = 0.8642), two (p = 0.0579), or three (p = 0.3358).
The pulmonary function test that demonstrated the greatest number of significant decrements following the oxygen exposures was D L CO (see Table 4).The three subjects with the lowest baseline F E NO (subjects 4, 3, and 9) had the greatest relative decrements in D L CO, which persisted for one to three days post-exposure.Subjects 1 and 8 also showed significant decreases in D L CO during the recovery period.When D L CO was corrected for V A , all subjects except subject 5 showed significant decrements at some point during the recovery period (see Table A2 in Appendix A).Both D L CO and D L CO/VA showed a significant main effect of time (p < 0.05 and p < 0.01, respectively) that was predominantly due to lower values during the second day of recovery compared to baseline (see Tables 4 and A2).
The relationship between the relative change in D L CO immediately following the six-hour dive and the immediate pre-and post-dive levels of F E NO is shown in Figure 1.Regression analysis of these data found that the relative change in D L CO immediately post-dive was significantly related to the immediate post-dive F E NO (r = 0.948, p < 0.001), as well as to the pre-dive F E NO (r = 0.902, p < 0.01).Using the mean baseline F E NO in the regression analysis instead of the pre-dive F E NO slightly improved the relationship (r = 0.931, p < 0.001).
Regression analysis of these data found that the relative change in DLCO immediately post-dive was significantly related to the immediate post-dive FENO (r = 0.948, p < 0.001), as well as to the pre-dive FENO (r = 0.902, p < 0.01).Using the mean baseline FENO in the regression analysis instead of the pre-dive FENO slightly improved the relationship (r = 0.931, p < 0.001).While some subjects had significant decrements in the spirometry tests following the dives, the group mean relative changes in the spirometry tests were on average smaller than those found for DLCO.Immediately following the six-hour dive, FIVC appeared to be more affected than FVC; however, neither FIVC nor FVC showed a significant main effect of time following group analysis (p = 0.0658 and p = 0.2176, respectively).
The two-compartment model analysis of the FENO data showed a significant 58% decrease in J'awNO (mean ± SD, baseline vs. post-dive 1 = 1681 ± 747 pl/s vs. 709 ± 465 pl/s, p < 0.001), with no change in CANO (mean ± SD baseline vs. post-dive 1 = 2.9 ± 1.2 ppb vs. 2.6 ± 0.6 ppb; p = 0.995) immediately following dive 1.A comparison of pre-and postexposure measurements for FENO and pulmonary function for the two subjects who conducted the surface control trials showed that all the dependent variables following the control exposure were within each individual's normal daily variability.While some subjects had significant decrements in the spirometry tests following the dives, the group mean relative changes in the spirometry tests were on average smaller than those found for D L CO. Immediately following the six-hour dive, F I VC appeared to be more affected than FVC; however, neither F I VC nor FVC showed a significant main effect of time following group analysis (p = 0.0658 and p = 0.2176, respectively).
The two-compartment model analysis of the F E NO data showed a significant 58% decrease in J'awNO (mean ± SD, baseline vs. post-dive 1 = 1681 ± 747 pl/s vs. 709 ± 465 pl/s, p < 0.001), with no change in C A NO (mean ± SD baseline vs. post-dive 1 = 2.9 ± 1.2 ppb vs. 2.6 ± 0.6 ppb; p = 0.995) immediately following dive 1.A comparison of pre-and post-exposure measurements for F E NO and pulmonary function for the two subjects who conducted the surface control trials showed that all the dependent variables following the control exposure were within each individual's normal daily variability.

Discussion
Traditionally, the "gold standard" for assessing PO 2 tox has been to measure changes in pulmonary function using spirometry (i.e., FVC) or D L CO.However, the sensitivity of these pulmonary function tests to assess PO 2 tox susceptibility has been questioned [21,22], and more recent research has explored other components in exhaled breath as potential biomarkers of PO 2 tox [22][23][24][25][26][27].Since the initial discovery that NO was present in expired air [28], F E NO has been one of the most widely studied exhaled breath biomarkers of pulmonary health.F E NO increases significantly in a variety of inflammatory airway diseases and is now commonly used to diagnose and phenotype asthmatics [7].It was thus originally hypothesized that F E NO would increase following HBO exposure due to free oxygen radical initiation of inflammatory reactions in the lungs.
One of the first published papers on the effect of hyperoxia on F E NO reported that F E NO increased with exposure to normobaric hyperoxic gas mixtures [29].However, the 10 min normobaric oxygen exposures in this early study were unlikely to result in inflammation of the lungs.The results of the study by Schmetterer et al. [29] are in direct contrast with our finding of a marked acute reduction in F E NO following prolonged HBO exposures.One potential reason for the disparate results is that, at the time that Schmetterer et al. [29] performed their study, there was no standard method for measuring F E NO.Since that time, it has become clear that F E NO is highly dependent on the expired flow rate and, thus, the recommended guidelines for F E NO measurements published by the ATS in 2005 [20] have since standardized expired flow rates at 50 mL/s using a flow resistor that also prevents contamination of the F E NO measurement from the high levels of NO found in the nasal cavity.Although we did observe significant increases in F E NO during the recovery days in five subjects, which may be reflective of a delayed inflammatory reaction in the lungs, only one of these subjects (subject 9) exhibited consistent decrements in pulmonary function during all three recovery days that was concomitant with abnormally elevated F E NO levels.
A second main finding from our study is that the duration of the HBO exposure affected the relative magnitude of the post-dive decrease in F E NO, with the eight-hour HBO dive resulting in significantly lower post-dive F E NO levels than the six-hour HBO exposure.This finding implies that the magnitude of the temporary F E NO decrease following the HBO exposures may be dose dependent.Since conducting these pilot HBO dives in 2007, we have conducted a wide variety of dry human hyperoxic exposures with varying inspired oxygen partial pressures and exposure durations to determine if the F E NO decreases found in the current study follow a predictable dose-response relationship.Findings from these studies have been presented to the undersea and hyperbaric medical and research community at various scientific forums [30][31][32] and were summarized in preliminary form by Fothergill and Weathersby [33].This study showed that the relative change in F E NO following dry resting hyperoxic exposure follows an exponential decline that is tightly related to the hyperoxic dose of the preceding exposure [33].In the statistical model of the changes in F E NO with varying HBO exposures, Fothergill and Weathersby [33] used the following expression to define the hyperoxic dose of the HBO dives based upon the inspired partial pressure of the oxygen breathed (PiO 2 ) and the duration of the exposure: Other investigators have also reported acute decreases in F E NO levels following HBO exposures [34][35][36][37][38][39][40]; however, the oxygen dose involved in these studies has rarely been great enough to induce changes in lung function or PO 2 tox symptoms noticeable enough to determine if the F E NO changes were related to PO 2 tox susceptibility.Our study is, therefore, somewhat unique, in that we were able to observe symptoms of PO 2 tox and measure significant decreases in lung function in some of our subjects, and then relate them to the observed changes in F E NO.Based upon our observations, we found that those individuals who had the lowest pre-dive F E NO levels exhibited the lowest post-dive F E NO levels and were most susceptible to PO 2 tox.This significant linear relationship between the pre-dive baseline levels of F E NO and the relative decrease in D L CO measured immediately post-dive should be taken with caution when interpreting the effects of HBO exposure on PO 2 tox susceptibility.In a more recent study, in which healthy U.S. Navy-trained divers were exposed to 6.5 h of 100% O 2 at 2.0 ATA [27], one subject, who aborted the dive early due to severe PO 2 tox symptoms, was found to have a 15% increase in D L CO immediately post-dive compared to his predive base line [41].Concomitant with the increase in D L CO was a 125% increase in total airway resistance and a 35% increase in proximal airway resistance (as measured using an impulse oscillometry methodology) [41].We surmise that the elevated D L CO post-dive for this subject was an artifact caused by the increase in pulmonary resistance that resulted in a large negative interpulmonary pressure being generated during the fast inspiratory maneuver required to perform the D L CO measurement.The negative interpulmonary pressure could result in increased blood volume entering the lung before the D L CO breathhold maneuver, raising the potential sink for the inhaled carbon monoxide gas mixture and artifactually raising the D L CO level.Therefore, we hypothesize that subjects who are particularly susceptible to PO 2 tox might experience a narrowing of the airways, possibly due to loss of normal airway tone.
Acute changes in airway diameter can be evoked by increases in cholinergic nerve activity or withdrawal of nitrergic neural activity [42].Interestingly, noncholinergic neurotransmitters such as NO are thought to control human airway smooth muscle and normal airway tone via nitrergic parasympathetic nerves [42].Thus, factors that compromise normal nitrergic parasympathetic control of airway tone, such as reduced levels of NO, would cause narrowing of the airways.The acute post-dive increase in airway resistance seen in the above PO 2 tox case was concomitant with an extremely low post-dive F E NO of 3.5 ppb [41].This is consistent with a neurogenic PO 2 tox response rather than an inflammatory reaction to the HBO exposure.
Although our study was not designed to elucidate the underlying mechanisms responsible for the reduction in F E NO with HBO exposure, our results are consistent with the observation from previous animal work [10,43] that suggests that endogenous levels of NO may serve to protect the lung from hyperoxic lung injury [43].Based upon our current findings, we suspect that, once F E NO levels fall below a critical level, the antioxidant defense and other processes in the lung that depend on NO become overwhelmed by the hyperoxic stress, resulting in changes in lung function and symptoms of PO 2 tox.However, as discussed in a review paper by Lui et al. [44], the role of the various NOS isoforms in the generation of NO in the face of hyperoxic stress and the impact of NO in the pathogenesis of acute lung injury is still under debate.
Several studies have attempted to ascertain the underlying mechanisms responsible for the decrease in F E NO with HBO exposures [45][46][47].The common thesis of these studies centers around the hypothesis that the decrease in F E NO with hyperoxic exposures is due to decreased enzymatic generation of NO due to oxidation of tetrahydrobiopterin (BH 4 ), which is an essential cofactor required for NO production by NOS [48].Fismen et al. [45] found that increased O 2 concentrations reduced BH 4 levels in human endothelial cells in a dose-dependent manner, without directly affecting the NOS enzyme.Similarly, Hesthammer et al. [46] reported that BH 4 levels in human umbilical vein endothelial cells (HUVEC) decreased in a dose-dependent manner.Although the latter study found that HUVEC NO production was also decreased following a 40 kPa O 2 exposure, a further decrease in HUVEC NO production was not observed when the oxygen exposure was increased to 60 kPa.In a follow up study by Hesthammer et al. [47], in which BH 4 was measured in venous blood samples of subjects exposed to 100% oxygen for 90 min at atmospheric pressure, both F E NO and BH 4 significantly decreased when measured 10 min after the exposure.Although oxidation of BH 4 levels and its subsequent uncoupling/inhibitory effects of NOS on NO production appear to be a plausible reason for the reduced F E NO with hyperoxic exposures; other mechanisms include the reaction of oxygen or superoxide radicals with NO to form peroxynitrite, which likely also contribute to the reduced F E NO.

Study Strengths and Limitations
To our knowledge, this is the first study that combined measurements of F E NO with traditional measures of pulmonary function in healthy divers to assess PO 2 tox susceptibil-ity following provocative HBO exposures that resulted in significant decrements in lung function.While the current study involved a small number of subjects, the study design incorporated multiple baseline and recovery measurements of F E NO and pulmonary function to provide a robust indication of daily inter-individual variation and accurately define when these dependent variables fell significantly outside of the individual's normal range, following the HBO exposure.The results clearly showed a wide individual variability in pulmonary function changes resulting from the HBO exposures, with half of our subject population showing minimal changes in lung function following the six-hour dive and the other half showing significant decreases that were more than two standard deviations below their normal day-to-day range.While this experimental design allowed us to analyze individual susceptibility to PO 2 tox, the small n approach leaves group statistical analysis susceptible to type II errors from the large variability in individual responses to the HBO stress.However, given our primary aim, we felt the small n approach was ethically more defensible as a pilot study on individual PO 2 tox susceptibility than a larger n study with limited individual pre-dive data but a higher power to detect group-level changes in pulmonary function post-dive.
An additional limitation of the current study is that only two subjects completed a control (normobaric air) condition, and that the study design was unblinded.This may have led to experimenter and subject bias regarding the expectation of pulmonary function decrements and PO 2 tox symptoms following the HBO exposures.While performing the pulmonary function measurements in accordance with the ATS recommendations [15][16][17][18] helps to reduce this potential bias, most spirometry measurements are dependent upon the individual performing a maximal inspiratory and/or expiratory effort to determine if pulmonary function is affected by the HBO exposure.In contrast, measurements of F E NO are conducted at a fixed expired flow rate and do not require maximum effort by the subject.F E NO may thus offer an alternative or complementary assessment of pulmonary hyperoxic stress that is less prone to the subject's effort than traditional spirometry measurements.While we acknowledged that there are many sources of NO in the lungs that can contribute to F E NO, and that the underlying mechanistic role of NO in hyperoxic acute lung injury is still controversial, F E NO may provide a useful noninvasive marker of the hyperbaric oxidative stress response of the lungs and lead to new insights into individual susceptibility to PO 2 tox.
Author Contributions: D.M.F. was the principal investigator on the study, responsible for the conceptualization and study original design, funding acquisition, data collection, formal analysis, interpretation of the results, and manuscript preparation.J.W.G. was a co-investigator on the study who assisted in data collection, data analysis, and interpretation of the results.He also helped critically revise and approve the final content in the article.All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Naval Submarine Medical Research Laboratory Institutional Review Board (protocol # NSMRL 2007.0003,approved 5 January 2007) and is in compliance with all applicable Federal regulations governing the protection of human subjects.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data from the current study are not publicly available due to government restrictions regarding data sharing, but are available from the corresponding author on reasonable request and when requirements are met.38

Figure 1 .
Figure 1.Relationship between the change in D L CO immediately following 6 h of breathing 100% oxygen at 202.65 kPa, and the immediate pre-dive F E NO (circles) and post-dive F E NO (squares).F E NO was measured at 50 mL/s expired flow rate.The numbers next to the data points are subject number identifiers.Subjects 3, 4, and 9 all showed significant decrements in D L CO immediately post-dive (see Table 4).Regression equations for the solid line and dashed line are as follows: Percent change in D L CO = −0.2265+ 0.0043 × pre-dive F E NO (r = 0.9018, p = 0.0022; r 2 = 0.8132); Percent change in D L CO = −0.2286+ 0.0096 × post-dive F E NO (r = 0.9485, p = 0.0003; r 2 = 0.8996).

Funding:
Sponsored by the Office of Naval Research through the In-house (Dept of Navy) Laboratory Independent Research (ILIR) Program, work unit number 50707.

Table 2 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in FVC following the 202.65 kPa HBO exposures.
* = Symptoms of PO 2 tox reported.NA = Not applicable.Cells highlighted in grey indicate significant decrements in FVC from the individual's mean baseline.

Table 3 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in F I VC following the 202.65 kPa HBO exposures.
* Symptoms of PO 2 tox reported.Cells highlighted in grey indicate significant decrements in F I VC from baseline.NA = Not applicable.

Table 4 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in D L CO following the 202.65 kPa HBO exposures.
* Symptoms of PO 2 tox reported.Cells highlighted in grey indicate significant decrements in D L CO from baseline.NA = Not applicable.
Symptoms of pulmonary O 2 toxicity reported.Cells highlighted in grey indicate significant decrements in D L CO adj from baseline.NA = Not applicable. *

Table A2 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in D L CO adjusted for alveolar volume (D L CO/V A ) following the 202.65 kPa HBO exposures.

Subject Baseline D L CO/V A Post-Dive 1 6 h O 2 Post-Dive 2 8 h O 2 Rec 1 Rec 2 Rec 3 Mean (mL/mmHg/min)
Symptoms of pulmonary O 2 toxicity reported.Cells highlighted in grey indicate significant decrements in D L CO/V A from baseline.NA = Not applicable. *

Table A3 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in alveolar volume (V A ) following the 202.65 kPa HBO exposures.Symptoms of pulmonary O 2 toxicity reported.Cells highlighted in grey indicate significant decrements in V A from baseline.NA = Not applicable. *

Table A4 .
Mean baseline levels, 2 × coefficient of variation (CV), and percent change in F E V1 following the 202.65 kPa HBO exposures.