Evaluation of a Gene–Environment Interaction of PON1 and Low-Level Nerve Agent Exposure with Gulf War Illness: A Prevalence Case–Control Study Drawn from the U.S. Military Health Survey’s National Population Sample

Background: Consensus on the etiology of 1991 Gulf War illness (GWI) has been limited by lack of objective individual-level environmental exposure information and assumed recall bias. Objectives: We investigated a prestated hypothesis of the association of GWI with a gene–environment (GxE) interaction of the paraoxonase-1 (PON1) Q192R polymorphism and low-level nerve agent exposure. Methods: A prevalence sample of 508 GWI cases and 508 nonpaired controls was drawn from the 8,020 participants in the U.S. Military Health Survey, a representative sample survey of military veterans who served during the Gulf War. The PON1 Q192R genotype was measured by real-time polymerase chain reaction (RT-PCR), and the serum Q and R isoenzyme activity levels were measured with PON1-specific substrates. Low-level nerve agent exposure was estimated by survey questions on having heard nerve agent alarms during deployment. Results: The GxE interaction of the Q192R genotype and hearing alarms was strongly associated with GWI on both the multiplicative [prevalence odds ratio (POR) of the interaction=3.41; 95% confidence interval (CI): 1.20, 9.72] and additive (synergy index=4.71; 95% CI: 1.82, 12.19) scales, adjusted for measured confounders. The Q192R genotype and the alarms variable were independent (adjusted POR in the controls=1.18; 95% CI: 0.81, 1.73; p=0.35), and the associations of GWI with the number of R alleles and quartiles of Q isoenzyme were monotonic. The adjusted relative excess risk due to interaction (aRERI) was 7.69 (95% CI: 2.71, 19.13). Substituting Q isoenzyme activity for the genotype in the analyses corroborated the findings. Sensitivity analyses suggested that recall bias had forced the estimate of the GxE interaction toward the null and that unmeasured confounding is unlikely to account for the findings. We found a GxE interaction involving the Q-correlated PON1 diazoxonase activity and a weak possible GxE involving the Khamisiyah plume model, but none involving the PON1 R isoenzyme activity, arylesterase activity, paraoxonase activity, butyrylcholinesterase genotypes or enzyme activity, or pyridostigmine. Discussion: Given gene–environment independence and monotonicity, the unconfounded aRERI>0 supports a mechanistic interaction. Together with the direct evidence of exposure to fallout from bombing of chemical weapon storage facilities and the extensive toxicologic evidence of biochemical protection from organophosphates by the Q isoenzyme, the findings provide strong evidence for an etiologic role of low-level nerve agent in GWI. https://doi.org/10.1289/EHP9009

. Agreement among the GWI Research, CDC, Kansas with no exclusions, and Kansas with exclusions case definitions in the prevalence case-control sample (n = 1,016). Kappa (95% confidence interval) measures the level of agreement beyond that expected by chance alone, where kappa ≤ 0 indicates no agreement, 0.01-0.20 none to slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.81-1.00 almost perfect agreement. The original Kansas case definition excluded veterans with certain comorbid conditions, but as the population aged, the exclusions became extreme, excluding many valid GWI cases. For this reason, in recent years the Kansas case definition has been applied either with no exclusions or with fewer comorbid conditions excluded. Cross-tables designated as "Cases only" are limited to studying agreement among case definitions in the 508 GWI cases used in the present prevalence case-control sample. Table S2. Comparison of unadjusted RERI's and correct asymmetrical 95% confidence intervals for the interaction of PON1 RR vs QQ genotype and having heard nerve agent alarms, computed from 1,016 Gulf War illness cases and controls with the SAS macro of Richardson & Kaufman and our adaptation of Zou's method. Table S3. Definition of the method used in our SAS macro to perform the sensitivity analysis to estimate the effect of misclassification of veterans' recall of having heard nerve gas alarms on our estimate of the strength of the GxE interaction. Table S4. Validation of our SAS macro for automating sensitivity testing to correct the odds ratio for misclassification of the environmental variable by reproducing Greenland and Lash's published resins-lung cancer example. 1   Table S5. Numerical values for Figure 4. The association of GWI with measures of low-level nerve agent exposure and genetic predisposition. Table S6. Interaction on the additive and multiplicative scales of having been located in the Khamisiyah plume and PON1 Q192R genotype on GWI. Table S7. Interaction on the additive and multiplicative scales of having been located in the Khamisiyah plume and PON1 Q192R type Q isoenzyme activity level on GWI. Table S8. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and butytylcholinesterase (BChE) genotype on GWI. Table S9. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and butytylcholinesterase (BChE) enzyme activity level on GWI. Table S10. Interaction on the additive and multiplicative scales of having taken pyridostigmine bromide and butytylcholinesterase (BChE) genotype on GWI. Table S11. Interaction on the additive and multiplicative scales of having taken pyridostigmine bromide and butytylcholinesterase (BChE) enzyme activity level on GWI. Table S12. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and PON1 Q192R type R isoenzyme activity level on GWI. Table S13. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and PON1 paraoxonase enzyme activity level on GWI. Table S14. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and PON1 arylesterase (phenylacetate) enzyme activity level on GWI. Table S15. Interaction on the additive and multiplicative scales of hearing nerve agent alarms and PON1 diazoxonase enzyme activity level on GWI. Table S16. Sensitivity analysis for correcting for unmeasured confounding the adjusted RERI for the effect of the GxE interaction of hearing alarms and PON1 RR vs QQ genotype on GWI on the additive scale. Table S17. Sensitivity analysis for correcting for unmeasured confounding the adjusted prevalence odds ratio for the effect of the GxE interaction of hearing alarms and PON1 RR vs QQ genotype on GWI on the multiplicative scale. Table S18. Methods and results of prior epidemiologic and clinical studies of the association of chemical weapons with GWI. Table S19. Prior studies identifying biochemical mechanisms by which low-level subclinical sarin exposure similar to that experienced in the 1991 Persian Gulf War causes chronic cellular pathology with behavioral changes resembling GWI. Table S20. Prior experimental evidence establishing that the PON1 Q192R type Q isoenzyme activity is the property of the PON1 gene that best protects the brain from the neurotoxic effects of low-level sarin nerve agent. Figure S1. Mean (SE) butyrylcholinesterase (BChE) serum activity by BChE genotype. All variant genotype groups were significantly (p<0.001) different from the U/U group except for the U/A group. The number at the base of each bar is the group sample size in the full USMHS genotyping sample (n=1,923). Figure S2. Prediction of irreversible or long-lasting health effects in soldiers exposed to airborne nerve agent above the U.S. Environmenal Protection Agency's Acute Exposure Guideline Level 2 (AEGL-2) but below the detection threshold of the widely deployed U.S. M8A1 nerve agent alarm device. On the basis of current science, the AEGL levels define the airborne concentrationtime thresholds (expressed as mg/m 3 per min) of a chemical as the level above which the general population, especially genetically susceptible individuals, could be expected to experience the following adverse effects: above AEGL-1 = notable discomfort, irritation, or certain asymptomatic nonsensory effects that are not disabling and are transient and reversible upon cessation of exposure; above AEGL-2 = irreversible or other serious, long-lasting adverse health effects or an impaired ability to escape; above AEGL-3 = life threatening health effects or death. 56 As tens of thousands of M8A1 detection alarms sounded repeatedly following U.S. and Colalition bombing of Iraqi chemical weapon storage sites throughout the conflict period, 57,58 most soldiers donned mission oriented protective posuture (MOPP) impervious suits, masks and gloves. Since, however, the AEGL-2 threshold (0.087 mg/m 3 for 10 min, 0.05 mg/m 3 for 30 min, etc.) was substantially below the M8A1 detection threshold (0.11 mg/m 3 ), there were undoubtedly periods of time before and after alarms sounded when solders were unprotected and exposed to AEGL-2 concentrations of nerve agent over the short times sufficient to cause irreversible or serious, longlasting health effects, particularly for more genetically susceptible individuals. Thus, the detection threshold of the M8A1 detector would have prevented incapacitating or fatal nerve agent exposures (above AEGL-3) if they had occurred, but not AEGL-2 exposures, which did occur but in 1991 were not yet known to have long-lasting effects. 59 The greater sensitivity of the Czech GSP-11 (0.05 mg/m 3 ) and the French AP2C/APACC (0.01 mg/m 3 ) detection devices explains the repeated Czech and French detections not detected by the ubiquitous U.S. M8A1 detectors. Figure reproduced from Neuroepidemiology 2013; 40: 160-177 by permission of S. Karger AG, Basel. Table S1. Agreement among the GWI Research, CDC, Kansas with no exclusions, and Kansas with exclusions case definitions in the prevalence case-control sample (n = 1,016). Kappa (95% confidence interval) measures the level of agreement beyond that expected by chance alone, where kappa ≤ 0 indicates no agreement, 0.01-0.20 none to slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.81-1.00 almost perfect agreement. The original Kansas case definition excluded veterans with certain comorbid conditions, but as the population aged, the exclusions became extreme, excluding many valid GWI cases. For this reason, in recent years the Kansas case definition has been applied either with no exclusions or with fewer comorbid conditions excluded. Cross-tables designated as "Cases only" are limited to studying agreement among case definitions in the 508 GWI cases used in the present prevalence casecontrol sample.
Abbreviations: G, has the risk genotype; E, exposed to the environmental risk; N, number of cases or controls with (1) or without (0) G; POR, prevalence odds ratio; Se and Sp, sensitivity and specificity of recall of E in case or control groups stipulated to model levels of recall bias; subscripted numbers indicate having the risk genotype (1) or not (0).
Note: This approach extends to GxE interaction the standard method for correcting the strength of association of a risk factor from the stipulated Se and Sp of its measurement in cases and controls, 1     Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the POR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. Because of missing values on Khamisiyah plume exposure due to classified or missing unit location information for 34 cases and 66 controls, the analyses included 474 cases and 442 controls. Abbreviations: CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction.
a Adjusted RERI = 1.20 (95% CI -0.93 -6.08) Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. Because of missing values on Khamisiyah plume exposure due to classified or missing unit location information for 34 cases and 66 controls, the analyses included 474 cases and 442 controls. Abbreviations: CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the POR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 504 cases and 502 controls; 4 cases and 6 controls were excluded for missing genotype data. Abbreviations: BChE, butyrylcholinesterase; CI, confidence interval; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. a Rare AChE genotypes include K/K, K/UK, U/A, and U/AK; other rare genotypes were not represented in the study sample. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: BChE, butyrylcholinesterase; CI, confidence interval; POR, prevalence odds ratio; RERI, relative excess risk due to interaction.  Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: BChE, butyrylcholinesterase; CI, confidence interval; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Serum paraoxonase catalytic activity for sarin is mediated mostly by the R isoenzyme. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Serum arylesterase catalytic activity for sarin is mediated by both Q and R isoenzymes. Note: The synergy index is a measure of interaction on the additive scale; it has the same distribution as the OR, viz., 0 to plus infinity with 1.0 as the equivalency point indicating no association. The ratio of the PORs, obtained from the interaction term in a logistic regression analysis, is a measure of interaction on the multiplicative scale. The potential confounders controlled for in the adjusted models include: age (years), sex (M, F), service branch (Army [referent], Navy, Air Force, Marines), rank (officer, enlisted), active duty vs Guard/Reserve, special strata (yes, no), Combat Exposure Scale (0=missing, 1=light [referent], 2=light to moderate, 3=moderate to heavy and heavy). One subject's missing age was imputed to the mean age of the sample. The analyses included 508 cases and 508 controls. Abbreviations: CI, confidence interval; CI, confidence interval; PON1, paraoxonase-1; POR, prevalence odds ratio; RERI, relative excess risk due to interaction. Serum diazoxonase catalytic activity for sarin is mediated mostly by the Q isoenzyme. This aRERI agrees exactly with the RERI adjusted for measured confounders in Table 2 calculated by Zou's SAS macro; whereas, its asymmetrical 95% CI from bootstrapping is slightly less conservative than that from Zou's method. Abbreviations: PRRUD, stipulated prevalence rate ratio in the underlying population for the association of the unmeasured confounder (U) with GWI; P0, stipulated probability of U in those in the underlying population who did not hear alarms; P1, stipulated probability of U in those in the underlying population who heard alarms; PRREU, the association of U with hearing alarms, assumed equal to PRRUD; k, adjustment factor calculated by the first equation below; aOR, the odds ratio from a logistic regression for the gene-environment interaction adjusted for the measured confounders; aRERIc, relative excess risk due to interaction on the additive scale, adjusted for measured confounders and corrected for unmeasured confounding, calculated by the second equation below; 95% CI, asymmetrical 95% confidence limits of aRERIc calculated by bootstrapping with 5,000 repetitions; plausible values of P0 and P1 are >0 to <1 and of PRRU, >1 to <10.
Assumption: PRRUD = PRREU a Equations for calculating aRERIc adapted from Corollary 3B in section 5 and the second example in section 6 of VanderWeele et al. 5    23 Hypothesisraising clinical description Description of a previously unsuspected chronic encephalopathic symptoms similar to GWI in workers who had repetitive subclinical sarin exposures in U.S. nerve agent factories during the Cold War, associated with unusual EEG changes. Burchfiel et al. 1976Burchfiel et al. , 1982 Laboratory experiments Administration of subclinical doses of sarin to Rhesus monkeys (1 μg/kg i.m. weekly x 10) produced chronic electroencephalographic (EEG) changes similar to those reported in the Duffy et al. study. Henderson et al. 2001Henderson et al. , 2002 Laboratory experiments Inhalation administration of subclinical doses of sarin to rats (0, 0.2, or 0.4 mg/m 3 of sarin for 1 h/day for 1, 5, or 10 days; follow-up at 30 d) produced persistent alteration in the numbers of muscarinic cholinergic M1 and M3 receptors in cortical and hippocampal brain regions, compatible with cognitive dysfunction.  Laboratory experiments  [40][41][42][43] Laboratory experiments Administration of corticosterone in drinking water daily x 5 or 7 d followed by sarin surrogate DFP (diisopropyl fluorophosphate, 1.5 mg/kg s.c.) initiated chronic neuroinflammation in the brains of mice with adverse effects on oligodendrocytes and epigenetic modification of genes related to the brain's immunologic and cognitive systems. Alshelh et al. 2020 44 Clinical study Neuroinflammation was recently demonstrated in veterans with GWI by in vivo positron-emission-tomography (PET) imaging of the brain. Deshpande et al. 2010Deshpande et al. , 2016Deshpande et al. , 2018Deshpande et al. , 2020 [45][46][47][48] Laboratory experiments Administration of a subclinical dose of DFP to rats (0.5 mg/kg daily s.c. x 5d; follow-up at 3-6 mo) was followed by behavioral abnormalities analogous to chronic depression, anxiety and memory impairment as well as hippocampal neuronal damage leading to a chronic elevation of intracellular calcium concentration, all largely corrected by 2 previously FDA-approved drugs.
Conclusion: The above experimental research has established that the PON1 Q192R is a gene that biologically modifies the pathological effects of organophosphate exposure and is not merely serving as a proxy marker. 49 Table S20. Prior experimental evidence establishing that the PON1 Q192R type Q isoenzyme activity is the property of the PON1 gene that best protects the brain from the neurotoxic effects of low-level sarin nerve agent.

Experimental model
Finding Davies et al. 1996 50 In vitro assays From assays of the rate of hydrolysis of sarin by the plasma from 93 human volunteers, plasma from PON1 QQ homozygotes had a mean hydrolysis rate of sarin 9.3 times that of RR homozygotes.
La Du et al. 2001 51 In vitro assays Sera from 25 veterans with GWI and 20 well control veterans were assayed for rate of hydrolysis of sarin (sarinase activity) as well as serum hydrolytic activity of the PON1 Q and R isoenzymes. Sarinase activity was correlated with Q isoenzyme activity but not with R isoenzyme activity. The catalytic efficiency of the purified Q isoenzyme with sarin was over 4-fold greater than with the R isoenzymes. This study is particularly relevant because it shows that the Q isoenzyme can effectively hydrolyze sarin in blood at the low physiologic concentrations expected with low-level sub-symptomatic sarin exposure.

Kanamori-Kataoka and Seto 2009 52
In vitro assays The maximum rate of hydrolysis of sarin with purified PON1 Q and R isoenzymes from plasma of 63 civilian volunteers was 3.5 times greater with the Q isoenzyme than with the R isoenzyme, confirming the finding of Davies et al.
Valiyaveetti et al. 2010 53 In vitro assays Acetylcholinesterase (AChE) is exceptionally sensitive to inhibition by sarin nerve agent and considered its primary target. In a series of vitro assays, purified human PON1 type Q isoenzyme, at physiological concentrations present in blood, was shown to potently prevent inhibition of AChE by sub-micromolar concentrations of sarin. Valiyaveetti et al. 2011,2011 54,55 In vivo experiments Intravenous treatment of guinea pigs with purified human PON1 type Q isoenzyme significantly increased survival, reduced physiologic signs of nerve agent exposure, and attenuated brain AChE inhibition after microinstillation inhalation exposure to 1.2 x LC50 of sarin. Figure S1. Mean (SE) butyrylcholinesterase (BChE) serum activity by BChE genotype. All variant genotype groups were significantly (p<0.001) different from the U/U group except for the U/A group. The number at the base of each bar is the group sample size in the full USMHS genotyping sample (n=1,923). Figure S2. Prediction of irreversible or long-lasting health effects in soldiers exposed to airborne nerve agent above the U.S. Environmenal Protection Agency's Acute Exposure Guideline Level 2 (AEGL-2) but below the detection threshold of the widely deployed U.S. M8A1 nerve agent alarm device. On the basis of current science, the AEGL levels define the airborne concentration-time thresholds (expressed as mg/m 3 per min) of a chemical as the level above which the general population, especially genetically susceptible individuals, could be expected to experience the following adverse effects: above AEGL-1 = notable discomfort, irritation, or certain asymptomatic nonsensory effects that are not disabling and are transient and reversible upon cessation of exposure; above AEGL-2 = irreversible or other serious, long-lasting adverse health effects or an impaired ability to escape; above AEGL-3 = life threatening health effects or death. 56 As tens of thousands of M8A1 detection alarms sounded repeatedly following U.S. and Colalition bombing of Iraqi chemical weapon storage sites throughout the conflict period, 57,58 most soldiers donned mission oriented protective posuture (MOPP) impervious suits, masks and gloves. Since, however, the AEGL-2 threshold (0.087 mg/m 3 for 10 min, 0.05 mg/m 3 for 30 min, etc.) was substantially below the M8A1 detection threshold (0.11 mg/m 3 ), there were undoubtedly periods of time before and after alarms sounded when solders were unprotected and exposed to AEGL-2 concentrations of nerve agent over the short times sufficient to cause irreversible or serious, long-lasting health effects, particularly for more genetically susceptible individuals. Thus, the detection threshold of the M8A1 detector would have prevented incapacitating or fatal nerve agent exposures (above AEGL-3) if they had occurred, but not AEGL-2 exposures, which did occur but in 1991 were not yet known to have long-lasting effects. 59