Characterizing 1991 Gulf War women veterans from the Boston Biorepository and Integrative Network for Gulf War Illness: demographics, exposures, neuroimaging and cognitive outcomes

Abstract

Objective: Gulf War Illness (GWI) is a debilitating multisymptom condition that affects nearly a third of 1990–91 Gulf War (GW) veterans. Symptoms include pain, fatigue, gastrointestinal issues, and cognitive decrements. Our work has shown that GWI rates and potential causes for symptoms vary between men and women veterans. Studies have documented neuropsychological and neuroimaging findings mostly in men or combined sex datasets. Data are lacking for women veterans due to lack of power and repositories of women veteran samples. Methods: We characterized GW women veterans in terms of demographics, exposures, neuropsychological and neuroimaging outcomes from the newly collated Boston, Biorepository and Integrative Network (BBRAIN) for GWI. Results: BBRAIN women veterans are highly educated with an average age of 54 years. 81% met GWI criteria, 25% met criteria for current PTSD, 78% were white, and 81% served in the Army. Exposure to combined acetylcholinesterase inhibitors (AChEi) including skin pesticides, fogs/sprays and/or pyridostigmine bromide (PB) anti-nerve gas pill exposure resulted in slower processing speed on attentional tasks and a trend for executive impairment compared with non-exposed women. Brain imaging outcomes showed lower gray matter volumes and smaller caudate in exposed women. Conclusions: Although subtle and limited findings were present in this group of women veterans, it suggests that continued follow-up of GW women veterans is warranted. Future research should continue to evaluate differences between men and women in GW veteran samples. The BBRAIN women sub-repository is recruiting and these data are available to the research community for studies of women veterans.

Keywords:

• Gulf War Illness
• Neurotoxicant Exposures
• Women Veteran
• Acetylcholinesterase inhibitors

Introduction

The women who served in the 1990–1991 Gulf War (GW) are unique in that they were the first group of women military service members to have fewer restrictions in terms of military occupations and exposure to combat (United States, General Accounting Office and United States, Department of Defense, 1993). Of the nearly 41,000 women deployed to the GW, many were serving in more diverse military occupational specialties than ever before and were in many cases exposed to the same toxicants as their male counterparts (Coughlin et al., 2017). Additionally, toxicant exposures occurred regularly for GW veterans where they worked, slept and ate regardless of gender (Sullivan et al., 2018). Toxicant exposures, many of which were neurotoxic, include the chemical class of acetylcholinesterase inhibitors (AchEi) such as pesticide sprays and creams, chemical warfare (CW) agents and Pyridostigmine Bromide (PB) anti-nerve gas pills (Sullivan et al., 2018) in addition to oil well fires and diesel fuels. AchEi’s affect the neurotransmitter acetylcholine such that acute exposures can result in confusion and attentional difficulties (Figueiredo et al., 2018). Chronic health effects from these exposures can result in chronic neuroinflammation and oxidative stress which have been shown to affect multiple cognitive domains including memory, attention, executive functioning and mood (Figueiredo et al., 2018). Mood can specifically be affected by neurotoxicant exposures including AChEi’s and this can be considered in assessments and analytical models. Our prior research comparing PTSD and PB exposure in GW veterans showed that PTSD was not associated with cognitive outcomes but PB exposure was significantly correlated with executive system functioning outcomes (Sullivan et al., 2018).

Numerous studies have been conducted on the chronic health symptoms that have resulted from Gulf War exposures. These health impacts were characterized into symptom clusters commonly known as Gulf War Illness (GWI) case criteria. The two most commonly known criteria are the CDC case definition and the more stringent Kansas GWI case criteria (Fukuda et al., 1998; Steele, 2000). The symptom clusters of the CDC criteria are such that the veteran must endorse at least one symptom from two of the three domains of fatigue, mood and cognition, and musculoskeletal symptoms. Whereas the Kansas criteria adds respiratory, skin, and gastrointestinal symptoms. There has been a question of whether women GW veterans are at increased risk for GWI given the prior paucity of studies focused on the area. However, recent studies by our group have shown that in fact GWI affects a higher proportion of women GW veterans and they report more severe symptoms in some symptom domains relative to GW-era women (Sullivan et al., 2020). Specifically, GW women veterans reported higher rates of 7 symptoms related to cognitive, neurological, mood and respiratory problems compared with non-deployed women veterans; these symptoms largely encompassed the Kansas neurological/mood/cognition subdomain. In addition, GW women veterans with war-related pesticide, oil well fire and PB pill exposures were more likely to meet the Kansas GWI criteria (Krengel et al., 2021). Specifically, oil well fires were associated with fatigue, pain, neurological and skin related symptoms (Krengel et al., 2021).These results suggested that further assessment of CNS related outcomes was warranted in GW women veterans.

In samples of both women and men it has been found that exposures during the GW have resulted in increased rates of criteria-defined GWI in deployed veterans and that cognitive impairments are higher in individuals with toxicant exposures. Specifically, White et al found that pesticide exposures were correlated with attention/executive functioning deficits in neuropsychological testing and Sullivan et al found that those with high exposures to pesticides and PB pills showed deficits in the areas of attention, processing speed and visual memory (Sullivan et al., 2018; White et al., 2001). When looking at exposures to chemical weapons, it was found that exposures were related to impairments in motor function and visuospatial functioning (Proctor et al., 2006). In addition, sarin nerve gas exposure was correlated with poorer functioning in visuospatial skills, motor speed and verbal memory. These exposures were not only found to be the cause of initial health concerns but have also been found to be associated with declines in cognitive status in addition to increased health symptoms over time (Krengel et al., 2022; Yee et al., 2020; Zundel et al., 2020). Neuroimaging studies, also most commonly conducted in combined samples of both men and women, have reported that exposed GW veterans have lower gray and white matter volume, which was related to omission errors on a continuous performance task and lower hippocampal volume that was related to verbal comprehension, and lower white matter volume which was related to verbal fluency and visuospatial dysfunction (Chao et al., 2010; 2011). Gender differences in neuroimaging findings and neuropsychological testing are suggested in studies comparing men and women by post hoc analyses, but small female group sizes have limited confidence in these findings. Unfortunately, there is a scarcity of studies reporting on women veteran outcomes alone which remains a problem for researchers and clinicians because when men and women GW veterans have been compared, differences between the groups are found in both neuropsychological and imaging outcomes. It has been recommended that women and men be compared separately in terms of imaging outcomes (Addiego et al., 2021).

As stated above, research in neuropsychological function and health symptoms in GW veterans has not always adequately represented women veterans due to the smaller pool of available women veterans and a lack of oversampling in recruitment efforts. This is an important limitation in the literature on military health as occupational samples have documented differences in exposure related outcomes by sex and when cohorts are divided into men and women, there do appear to be differences in outcomes (Bekele et al., 2021; Proctor et al., 2006). This is particularly relevant because women made up 7% of troops who were deployed to the Gulf War. Specifically, women were found to be different from men in symptoms relating to cardiac, gastrointestinal, and neurologic health when comparing global exposures versus unexposed in a longitudinally followed cohort of GW veterans (Bekele et al., 2021; Proctor et al., 2006). Our research team has also reported on health symptoms in a relatively large cohort of several hundred women, known as the GW women veteran’s cohort, and have found correlations between these exposures with health symptom outcomes (Heboyan et al., 2019; Krengel et al., 2021). In the GW women veterans’ cohort, we have found that GWI is more common in women and that those with exposure to smoke from oil well fires, PB pills, skin pesticides and insect baits were associated with higher rates of Kansas GWI case criteria, while smoke from oil well fires and exposure to chemical warfare agents were associated with higher rates of the CDC criteria (Krengel et al., 2021). When examining specific exposures with specific symptom domains, it was found that women who were exposed to smoke from oil well fires were over 5 times more likely to report neurologic issues such as cognitive impairment. In this large cohort of women veterans, neuropsychological testing, neuroimaging and biomarker data were not collected. Our studies showing differences in self-reports of exposures and increased neurologic complaints in women veterans from the GW resulted in the development of a data repository specific to women GW veterans. This expanded sample of women GW veterans was established to allow evaluation of the extent to which women with neurotoxicant exposures experienced measured deficits in neuropsychological function and/or structural differences when examining neuroimaging findings.

To examine these outcomes in a large group of women, it was necessary to combine samples from multiple studies to increase power. This was made possible by using a recently developed data-repository, the Boston Biorepository, Recruitment, and Integrative Network for Gulf War Illness (BBRAIN) (Keating et al., 2021). This repository utilizes the GWI Common Data Elements (CDEs) for neuropsychological testing and neuroimaging protocols that were created by experts from the United States Department of Veteran Affairs and the Department of Defense Congressionally Directed Medical Research Program (CDMRP) (Systems Assessment Working Group, 2022) by group consensus. The purpose of these CDEs is to allow for the aggregation of data to increase statistical power in GWI research. The BBRAIN repository was created to allow for data mining from prior studies and to improve power for studies by combining smaller sub-cohorts from prior studies that included women veterans. This paper characterizes the newly established cohort of 1990–91 GW women veterans in the existing BBRAIN repository and examines neuropsychological and neuroimaging findings specifically in GW women veterans.

Participants and methods

The current study is an analysis of data from a national bio/data-repository (Boston Biorepository Recruitment and Integrative Network for Gulf War Illness; BBRAIN) which includes key data samples from previously conducted GW studies that have been merged and combined into a single multicohort dataset (Keating et al., 2021). Data include general health and chronic symptom questionnaires, demographics, deployment and self-reported exposure histories as well as key neuropsychological test variables. Two separate datasets were combined: data from the multisite Gulf War Illness Consortium (GWIC) case-control study and data from three University of California San Francisco and San Francisco VA Gulf War veteran cohort studies. These cohorts have been previously described (Chao et al., 2011; Keating et al., 2021; Steele et al., 2021). All participants signed informed consents to share data for future studies (CDMRP/GWIRP GW170055, IRB #H-37828).

War-related exposures

Environmental exposures were collected from the Kansas Gulf War and Health Questionnaire, which is part of the current CDE for GWI exposures (Steele, 2000; Systems Assessment Working Group, 2022) This self-reported survey includes questions about a broad range of exposures during the 1991 Gulf War, including chemical weapons, pesticides, and anti-nerve gas pills. Chemical warfare (CW) agent exposure was determined by reporting hearing chemical alarms, while seeing smoke from oil well fires was used to determine particulate matter exposure. Pesticide exposure to repellents (DEET) was determined if participants reported using pesticide cream or spray on their skin or insecticides by reporting seeing the area in which they lived sprayed or fogged with pesticides (organophosphate, carbamate pesticides). Anti-nerve gas pill exposure was determined if the veteran reported taking the prophylactic Pyridostigmine Bromide (PB) pills. For analysis purposes, GW veterans were categorized as having a respective exposure if they reported seven or more days of experiencing the area in which they lived fogged with pesticides, hearing chemical alarms and/or taking PB pills. They were categorized as unexposed if they reported less than seven days of the exposure. These exposure duration periods were based on previous research. Wolfe et al., showed that at least 7 days of exposure to 3 blister packs of PB pills per day were associated with worsened health symptoms compared to those who took less than 21 PB pills (Wolfe et al., 2002). Participants were considered exposed to pesticide cream or liquid on their skin if they reported 31 or more days of use and unexposed if they reported less than 31 days of use. This exposure definition was based on work by Sullivan et al which found pesticide applicators from the GW who reported 31 or more days of exposure to pesticide cream on skin to have decrements in cognitive functioning (Sullivan et al., 2018). Participants were considered exposed to AchEis if exposed to one or more of the following environmental exposures per the above definitions: chemical warfare agents, repellents, insecticides and/or PB pills.

Post-Traumatic stress disorder and mood disorder protocol

Post-traumatic stress disorder (PTSD) was determined by clinical interviews conducted by trained research staff. The Clinician Administered PTSD Scale for DSM-5 (CAPS) was used to document case criteria (Weathers et al., 2018). The CAPS for DSM-5 is a structured clinical interview which assesses each of the diagnostic criteria for PTSD querying for symptoms in the past month related to intrusive memories, avoidance, negative changes in cognition and mood outcomes, alterations in arousal and reactivity (Weathers et al., 2018). From these questions, the DSM-5 criteria provided current and lifetime PTSD diagnosis. This diagnosis was made in relation to deployment related and other life traumatic events. Because the current study was interested in controlling for the potential confounding effects of PTSD in relation to current health status, analytic models were adjusted for the presence/absence of current PTSD. The structured clinical interview for DSM-5 (SCID) was also administered for diagnosis of major depressive disorder and to identify any exclusionary diagnosis of bipolar disorder and schizophrenia (Steele et al., 2021).

Neuropsychological testing protocol

The neuropsychological test variables used for these analyses were selected based on commonality from multiple studies within the repository. Overlapping neurocognitive tests from the neuropsychological batteries across studies included the California Verbal Learning Test Second Edition (CVLT-II), Delis Kaplan Executive Function System Color-Word Interference Test (D-KEFS), Trail Making Test (TMT), and the Conners Continuous Performance Test Third Edition (CPT3) (Conners, 2014; Delis et al., 1987–2000; 2001; Reitan & Wolfson, 1985). All tests were administered by a trained neuropsychological test administrator. All cognitive measures compared are included in the current list of CDEs for GWI as reported by Cohen et al and Jeffrey et al (Systems Assessment Working Group, 2022; Jeffrey et al., 2019). These tests were specially chosen by GWI researchers to increase the sensitivity and comparability of neuropsychological testing for GWI. The CDE recommendations included tests of attention, executive function, and memory (Systems Assessment Working Group, 2022).

Neuroimaging protocols

Magnetic resonance imaging (MRI) neuroimaging data were also collected on a subset of participants from the total sample. For the Boston GWIC images, a 3 Tesla Philips scanner (Philips Healthcare, Best, The Netherlands) was used. For the SFVA images, a Siemens Skyra with a 32-channel receiver head coil scanner was used. For the Boston data, a three plane TFSE scout scan, sense reference scan and accelerated high resolution MPRAGE scan on the sagittal plane was acquired (TR = 6.8 msec, TE = 3.1 msec, flip angle = 9°, slice thickness = 1.2 mm, 170 slices, FOV = 250 mm, matrix = 256 × 256). For the SFVA site, the imaging parameters included: 3D whole brain gradient echo MRI TR/TE/TI = 2500/2.98/1100 ms, 1.0 × 1.0 × 1.0 mm3 resolution. All scans were quality checked and if approved were post-processed in Freesurfer software. The Boston GWIC study used Freesurfer version 6 and the SFVA site used Freesurfer version 7.1 for post-processing. Brain regions of interest in this study included total white matter, total grey matter, hippocampus, putamen, and caudate nucleus volumes measured in mm³.

Statistical methods

Descriptive statistics were produced overall and by exposure status for yes versus no AchEi exposure (pesticides, PB, chemical weapons), using means and standard deviations (SD) for continuous variables and frequencies and percentages for categorical variables. Differences in demographic characteristics by exposure were tested using two-sample t-tests for continuous variables and chi-squared tests or Fisher’s Exact tests for categorical variables. Neuropsychological test outcomes by exposure were analyzed using multivariable linear regression models, which controlled for age, education and current PTSD in the analyses. Standardized effect sizes for adjusted mean differences between exposed and unexposed veterans are described through Cohen’s d, which was calculated as the difference in adjusted means divided by the pooled standard deviation of the outcome.

Separately, brain imaging outcomes by exposure status were investigated using multivariable linear regression models, controlling for age, education, intracranial volume, current PTSD and other exposures. Considering the known negligible impact of Freesurfer version differences on brain volumetric outcomes we did not include the software version as a controlling variable (Filip et al., 2022). pvalue equal to or less than 0.05 were considered statistically significant.

Results

Demographics and exposures

Characteristics of the 74 women included in the study sample are presented in Table 1. The average age of these veterans was 55 years old with a standard deviation of 6.5 years. The majority were white (78%), had bachelor’s degree or higher (69%). Twenty-five percent had a current diagnosis of PTSD. The study population consisted of 22% non-white women. Most women veterans (81%) met the GWI Kansas case criteria, which contains a symptom domain for cognitive complaints. The proportion of veterans exposed to chemical weapons, skin pesticides, pesticide sprays/fogs, PB pills and total AChEi’s did not vary by age, education or race. A higher proportion of veterans who served in the Army reported exposure to chemical weapons, skin pesticides, area fogging and total AChEi’s, compared to veterans in other military branches (Table 1).

Table 1. Characteristics of Gulf War Women Veterans by War-related Toxicant Exposure (n = 74).

			Chemical Weapons		Skin Pesticides		Pesticide Sprays/ Fog		PB pills		AChEi
		Overall	+	–	+	–	+	–	+	–	+
N(%)		74	41(55)	33(45)	33(45)	40(55)	18(25)	54(75)	34(46)	40(54)	60 (81)
Age mean(SD)		55(6.5)	55(7.4)	54(5.3)	54(7.6)	55(5.6)	54(8.5)	55(5.9)	55(7.8)	54(5.2)	54.7 (6.6)
Race N(%)	Black/African American	11(15)	5(50)	5(50)	4(40)	6(60)	2(22)	7(78)	3(30)	7(70)	6 (10)
	White/Caucasian	58(78)	29(52)	27(48)	24(43)	32(57)	12(21)	44(79)	27(48)	29(52)	49 (83)
	Other	5(7)	4(100)	0	4(100)	0	3(75)	1(25)	3(67)	1(33)	5 (7)
Education N(%)	High school/GED	3(4)	3(100)	0	0	3(100)	0	3(100)	2(67)	1(33)	3 (5)
	Some school beyond high school	20(27)	6(32)	13(68)	6(33)	13(67)	7(37)	12(63)	7(37)	12(63)	14 (23)
	Bachelor’s degree	22(30)	14(67)	7(33)	12(57)	9(43)	7(35)	13(65)	12(57)	9(43)	19 (32)
	Advanced degree	29(39)	15(54)	13(46)	15(54)	13(46)	3(11)	25(89)	13(46)	15(54)	24 (40)
Service Branch N(%)	Army	60(81)	33(55)	24(42)	31(54)	28(46)	16(29)	40(69)	30(53)	27(47)	52 (86)
	Air Force	7(9)	4(6)	3(4)	0	7(18)	0	7(100)	2(29)	5(71)	4 (6)
	Marines	3(4)	1(33)	2(67)	1(33)	2(67)	0	3(100)	2(67)	1(33)	2 (4)
	Navy	4(5)	0	4(100)	1(25)	3(75)	1(25)	3(75)	0	4(100)	2 (4)
Current PTSD N(%)	Yes	16(25)	28(60)	19(40)	25(53)	22(47)	13(28)	33(72)	21(45)	26(55)	11 (22)
	No	49(75)	10(42)	14(58)	8(33)	16(67)	4(17)	20(83)	13(54)	11(46)	40 (78)
GWI N(%)	Case	60(81)	35(61)	22(39)	30(53)	27(47)	17(30)	39(70)	29(51)	28(49)	51 (85)
	Control	14(23)	3(21)	11(79)	3(21)	11(79)	0	14(100)	5(36)	9(64)	9 (15)

Note. AChEi: acetylcholinesterase inhibitors; GWI: Gulf War Illness; PTSD: Post-traumatic stress disorder.

‘+’: Exposed and ‘-‘: Unexposed.

Neuropsychological outcomes

After adjusting for age, education, and current PTSD, women who reported exposure to combined AChEi’s (pesticides, PB, chemical weapons) had significantly worse cognitive scores in the domain of attention and processing speed and showed a moderate to large effect size for these outcomes (Table 2). Participants who reported exposure to AChEi’s had significantly slower time to completion on Trial 1 and 2 and more self-corrected errors on the DKEFs Color-Word Interference trial 2 (p = 0.0041). There were no significant differences when comparing AChEi exposed vs unexposed groups on the CVLT-II test outcomes associated with verbal learning and memory. While there was a moderate effect size for AchEi-exposed women compared to AChEi-unexposed women on D-KEFS Color-Word Interferences time to completion on Trial 3, their times did not differ significantly (p = 0.06).

Table 2. Association of neuropsychological outcomes with war-related acetylcholinesterase inhibitors exposure Among Women Gulf War Veterans.

		Exposed N = 60	Unexposed N = 14	Effect size
		Adjusted Mean		Cohen’s d	p-value
Attention/ processing speed	CPT 3 Omissions	49.01	47.45	0.19	0.5483
	CPT 3 Commissions	49.87	47.66	0.27	0.3953
	CPT 3 Hit Reaction Time Raw Score	431.03	407.42	0.33	0.2961
	D-KEFS Color-Word Interference trial 1 time (sec)*	31.19	27.60	0.64	0.0429
	D-KEFS Color-Word Interference trial 1 self-corrected errors	0.41	0.23	0.26	0.4117
	D-KEFS Color-Word Interference trial 1 uncorrected errors	0.02	−0.03	0.31	0.3235
	D-KEFS Color-Word Interference trial 2 time (sec)*	23.37	20.29	0.92	0.0041
	D-KEFS Color-Word Interference trial 2 self-corrected errors*	0.05	0.61	−1.28	0.0001
	D-KEFS Color-Word Interference trial 2 uncorrected errors	0.07	−0.01	0.30	0.3382
Verbal Memory	CVLT-II Correct in Trials 1–5	49.39	47.23	0.25	0.4286
	CVLT-II Correct in short delay free recall	10.24	9.93	0.13	0.6720
	CVLT-II Correct in long delay free recall	10.80	10.66	0.05	0.8673
Executive Functioning	D-KEFS Color-Word Interference trial 3 time (sec)	62.39	54.38	0.58	0.0654
	D-KEFS Color-Word Interference trial 3 self-corrected errors	0.97	1.25	−0.28	0.3619
	D-KEFS Color-Word Interference trial 3 uncorrected errors	0.68	0.53	0.19	0.5370
	D-KEFS Color-Word Interference trial 4 time (sec)	70.38	62.24	0.51	0.1067
	D-KEFS Color-Word Interference trial 4 self-corrected errors	1.04	0.75	0.28	0.3694
	D-KEFS Color-Word Interference trial 4 uncorrected errors	0.70	0.47	0.17	0.5769
	TMT Trail A: Time (sec)	31.97	30.17	0.17	0.5776
	TMT Trail B: Time (sec)	78.34	72.45	0.25	0.4314

Note. Acetylcholinesterase inhibitors (AchEi) represent Pesticide and/or PB pills and/or Chemical Alarms.

CPT 3: Conners Continuous Performance Test; CVLT-II: California Verbal Learning Test version 2; DKEFS: Delis Kaplan Executive Function System; TMT: Trail Making Test.

Neuroimaging structural differences

MRI data were available for 52 women in the sample. When adjusting for age, current PTSD, intracranial volume, and education level women reporting exposure to AChEi’s had significantly smaller total gray matter volumes than women without AChEi combined exposures with a large effect size (p = 0.03; d=-0.85); they also had lower caudate volumes (p = 0.053; d=-0.77) as reported in Table 3.

Table 3. Association of brain imaging volumetric measures with deployment acetylcholinesterase inhibitor exposure among women gulf war veterans.

	AChEi Exposed	Unexposed
	Adjusted Mean		Effect size
Volume (mm^3)	N = 60	N = 14	Cohen’s d	p-value
Total white matter	449710.6	451668.0	−0.04	0.9166
Total grey matter*	572807.9	595072.6	−0.85	0.0330
Hippocampus	7831.13	8204.12	−0.59	0.1355
Caudate	6270.40	6699.26	−0.77	0.0533
Putamen	8672.73	8627.46	0.05	0.8941

Note. AChEi: acetylcholinesterase inhibitor *p < 0.05.

Discussion

This paper describes results from a newly combined multicohort sample of 1990–91 Gulf War women veterans from the BBRAIN biorepository that evaluates associations of and war-related exposures with measures of cognitive function and neuroimaging outcomes. This combined dataset of women GW veterans is a geographically representative sample encompassing U.S. women veterans from many different states and is available to share for other relevant studies. This group of women veterans was highly educated, served in all military branches but primarily in the Army and were predominantly white. BBRAIN women veterans also reported environmental exposures comparable to many other prior studies of mixed sex cohorts (Proctor et al., 2006; Steele et al., 2021; Sullivan et al., 2018). A higher proportion of Army veterans reported use of skin pesticides compared to veterans in other cohorts. Skin pesticides were potentially problematic for GW veterans, since the repellent DEET that was used in the GW was up to 75% active ingredient making it much more potent than current formulations (Fricker et al., 2000; Sullivan et al., 2018). In fact, nearly all exposures were reported most frequently, although not significantly, by Army veterans with the exception of total AChEi’s. In this first descriptive study of this GW women cohort, we further examined the relationships between combination exposures to AChEi toxicants and neurocognitive and neuroimaging outcomes in women veterans.

The few findings from our initial descriptive analyses suggest that exposure to AChEi toxicant chemicals during the 1990–91 GW are associated with subtle differences in attention and processing speed and potentially with executive functioning skills in women GW veterans. These results are similar to previous combined sex studies of cognitive outcomes in GW veterans in identifying attention and executive system functioning decrements and with those of other organophosphate pesticide exposed groups (Sullivan et al., 2018). Although quite subtle, the current study results suggest that performing analyses separately or controlling for sex when comparing cognitive outcomes in Gulf War veterans could be important. This is particularly relevant due to prior studies that have reported differences in health symptom reporting outcomes related to AChEi exposures in men and women GW veterans (Zundel et al., 2019). To our knowledge, this is the first study to report cognitive outcomes and their association with exposures separately in GW women veterans and therefore provides new information and a new women veteran cohort to the field of GWI research. Although few differences were noted in this likely underpowered study, we are continuing to add to our cohort.

There were also a few significant exposure-related differences observed on MRI brain imaging. Women veterans who reported AchEi exposure during the GW had significantly smaller gray matter volumes than unexposed women veterans. Total gray matter volume was previously reported to be lower in a chemical weapon exposed sample of both men and women GW veterans. However, ours is the first study to report this in women veterans only (Chao et al., 2011). In addition, exposed women had smaller caudate nuclei compared with the unexposed group. The caudate nucleus has been shown to be associated with executive functioning and this potentially validates the association between AChEi exposures and a trend for decreased executive functioning in this sample of women veterans. This is consistent with other findings of caudate/putamen differences in relation to exposures to organophosphate/nerve agent exposure (Figueiredo et al., 2018).

We are aware of only one prior study that evaluated brain imaging outcomes in GW women and men compared separately and with women and men with chronic fatigue syndrome (ME/CFS). It was found that women and men GW veterans differed on brain imaging measures compared with their respective sex matched ME/CFS participants. However, there was no direct comparison of GW women and GW men on brain imaging outcomes. It was recommended that brain imaging analyses should be performed separately by sex and our results concur with this recommendation (Addiego et al., 2021).

Collectively, the findings presented here provide an initial evaluation of a newly combined cohort of GW women veterans. This study reported on neuropsychological and neuroimaging outcomes in women veterans; however, the BBRAIN repository also includes blood and saliva biomarkers, extensive physical and mental health symptom survey data and additional neuroimaging sequences in subgroups of women veterans that is available for future sharing and analysis.

This study has several limitations. Despite inclusion of a relatively larger sample than previously reported of 74 GW women veterans, these findings are still limited due to small sample size and was potentially underpowered to detect subtle differences. In addition, this was an exploratory study that could have increased the risk of type 1 error. Larger samples of GW women veterans are essential to validate these findings and provide further insights concerning potential associations of GW toxic exposures with neuropsychological decrements and brain structure differences. Although GW veterans are a highly educated deployment group, this cohort of women was slightly more educated than most, which may affect the generalizability of the findings. Another limitation that these methods present is the self-reported nature of toxic exposure data during the Gulf War. However, this remains the primary method of Gulf War exposure assessment as toxic exposures were not well documented at the time. In this study, we provide the first comparisons of cognitive and neuroimaging outcomes in this combined multisite sample of women GW veterans from the BBRAIN repository. As the BBRAIN repository continues to grow, additional data will provide larger samples and additional power for more definitive assessments, expanding opportunities for research investigators interested in studying women’s military health outcomes.

Conclusions

This study describes a newly established cohort of women Gulf War veterans from the BBRAIN repository. Findings suggest subtle but potentially important cognitive and neuroimaging differences in women exposed to AChEi neurotoxicants during the war that differ from prior reports of combined men and women cohorts. Specifically, the finding of lower gray matter volumes in exposed groups suggests that this group may need further follow-up as they age. These results also suggest that separate evaluation of brain structure and function in women veterans may be warranted in future studies. These unique outcomes may lead to individualized assessment protocols and treatments for women GW veterans. This may also benefit women with similar exposures from other deployments in addition to women with chronic health effects from occupational exposures to similar types of neurotoxicants.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by U.S. Department of Defense.