More than 650,000 U.S. soldiers served in Operations Desert Shield and Desert Storm, the military campaign waged by a coalition force of 35 nations against Iraq in 1991 in response to the invasion and annexation of Kuwait. After the Gulf War (GW) ended, many soldiers returned home with a multitude of unexplained symptoms (e.g., fatigue, memory and concentration problems, muscle and joint pain, headaches, gastrointestinal and other health problems) (Freeman et al., 2020). Gulf War Illness (GWI) is the term commonly used to describe this condition, which is estimated to affect 25–32% of deployed GW veterans (White et al., 2016). Longitudinal research suggests few GW veterans have recovered or substantially improved from GWI in the decades since the Gulf War ended (Research Advisory Committee on Gulf War Veterans Illnesses (Rac-Gwvi), 2008, 2014; White et al., 2016).

Over the years, the cross-sectional studies that have examined GW veterans’ post-deployment health status have reported higher rates of chronic multisymptom illnesses such as fibromyalgia (Ismail et al., 2008; Kang et al., 2009), chronic fatigue syndrome (Unwin et al., 1999; Gray et al., 2002; Ismail et al., 2008; Kang et al., 2009; Li et al., 2011), irritable bowel syndrome (Gray et al., 2002; Kang et al., 2009), and chronic neurological disorders such as amyotrophic lateral sclerosis (Haley, 2003; Horner et al., 2008), brain cancer (Bullman et al., 2005; Barth et al., 2009; Young et al., 2010), seizures (Kang et al., 2000; McCauley et al., 2002; Kang et al., 2009), neuralgias and neuritis (Kang et al., 2000, 2009), and chronic migraine headaches (Unwin et al., 1999; Kang et al., 2000; Gray et al., 2002; Kelsall et al., 2004; Porter et al., 2018). There have also been reports of higher rates of arthritis (Unwin et al., 1999; Kang et al., 2000; Gray et al., 2002; Li et al., 2011; Porter et al., 2018), lung diseases (Unwin et al., 1999; Kang et al., 2000; Cowan et al., 2002; Gray et al., 2002; McCauley et al., 2002; Kang et al., 2009; Li et al., 2011), eye or vision problems (Kang et al., 2000) hypertension (Unwin et al., 1999; Kang et al., 2000; Gray et al., 2002; Ismail et al., 2008; Kang et al., 2009; Li et al., 2011), and heart disease (Kang et al., 2000; McCauley et al., 2002; Kang et al., 2009; Li et al., 2011; Porter et al., 2018) in GW veterans compared to other veteran control groups. Moreover, there is suggestive evidence that deployed GW veterans may be at significantly higher risks for age-related conditions (e.g., hypertension, hypercholesterolemia, myocardial infarct, diabetes, stroke, arthritis, and chronic bronchitis) compared to civilians (Zundel et al., 2019).

The first aim of this study was to examine the current health status of GW veterans who previously participated in research studies at the San Francisco VA Health Care System (SFVAHCS). We used the Veterans Affairs Frailty Index (VA-FI) (Orkaby et al., 2019; Cheng et al., 2021) as a proxy measure of aging and vulnerability to poor outcomes. Frailty is a clinical syndrome commonly described in older adults (Fried et al., 2001). Frailty can increase an individual’s vulnerability to adverse events such as mortality, morbidity, disability, and hospitalizations (Fried et al., 2001; Clegg et al., 2013), particularly after exposure to stressors (Morley et al., 2013). The VA-FI (Orkaby et al., 2019, Cheng et al., 2021) is an electronic index calculated from Veterans Health Administration (VHA) claims and electronic health records (EHR). The current study used the VA-FI as a proxy measure of the current health status of GW and non-GW veterans.

In the absence of diagnostic tests and validated case definitions for GWI, researchers have relied upon different criteria to study the chronic multisymptom condition. In 2013, the VA contracted with the Institute of Medicine (IOM), now the National Academy of Medicine (NAM), to develop a single case definition for GWI. However, the IOM/NAM panel was unable to do this due to limitations with the data that was available at the time (Institute of Medicine, 2014). Instead, the panel recommended the Kansas GWI case definition (Steele, 2000) be used for research and the Center for Disease Control and Prevention (CDC) Chronic Multisymptom Illness (CMI) case definition (Fukuda et al., 1998) be used for clinical purposes. However, both of these case definitions were developed in the late 1990s and early 2000s and were based on the types and pattern of symptoms reported by GW veterans in the decade post-deployment (Steele et al., 2021). Furthermore, because both case definitions ask veterans whether they experienced symptoms over the past 6 months, neither case definition captures or accounts for potential changes in GWI symptoms over time (Maule et al., 2018; Steele et al., 2021). Therefore, a second aim of this study was to examine the current health status of GW veterans as a function of the different GWI case definitions ascertained between 2002 and 2018.

Many GW veterans were exposed to a myriad of chemicals during deployment (White et al., 2016). Some of chemicals present in the GW milieu were neurotoxins that have since been associated with chronic health symptoms that encompass multiple body systems (Steele, 2000). Therefore, a final aim of this study was to examine the relationship between GW veterans’ deployment-related experiences and exposures and their current VA-FI.

To our knowledge, this is the first study to use the VA Frailty Index to examine GW veterans’ vulnerability to poor health outcomes. It has been suggested that deployed GW veterans may be aging more rapidly than their civilian counterparts (Zundel et al., 2019). For this reason, it is important to characterize and assess this group of veterans for the development of age-related syndromes. The major objectives of the study are to: (1) examine the VA-FI of past GW veteran research participants as a function of different GWI case definitions, (2) compare VA-FI in deployed GW veterans with a matched group of non-GW veterans, and (3) examine the relationship between VA-FI and GW veterans’ deployment related experiences and exposures. Because exposures to GW related chemicals and toxins may have altered the trajectory of aging in these veterans, it is important to have a better understanding of how GWI affects GW veterans’ current health and functional status.

All study procedures were approved by institutional review boards (IRB) at the University of California, San Francisco and the San Francisco VA Health Care System (SFVAHCS). Informed consent was obtained from all GW veterans who participated in the original research studies at the SFVAHCS. A waiver for consent was obtained for the retrospective cohort analysis because some of the study participants had died or were no longer receiving care through VHA when the analyses were performed.

We performed a retrospective cohort analysis of 703 GW veterans who participated in research studies related to GWI and GW-related exposures at SFVAHCS between 2002 to 2018. Electronic health records (EHR) and VA health care encounter diagnostic codes were identified for 602 of the 703 past GW veteran study participants in the VA’s corporate data warehouse (CDW).

The VHA is an integrated health care system that captures and stores claims and health care data in a centralized national database (Price et al., 2015). The VA-FI is an electronic frailty index calculated from VHA administrative claims and EHR (Orkaby et al., 2019; Cheng et al., 2021). We extracted International Classification of Diseases (ICD) diagnosis and procedure codes, Current Procedural Terminology (CPT), and Healthcare Common Procedure Coding System (HCPCS) codes for the GW veteran study participants from tables in CDW on an index date (04/05/2023). Smoking status was determined from the most recent clinical assessments of tobacco use prior to the index date.

Structured data in CDW and from the Death Ascertainment File indicated that 44 of the former SFVAHCS GW veteran participants were deceased at the index date. We calculated the VA-FI for the 588 GW veteran participants who were alive at an index date. The VA-FI was calculated as the proportion out of 31 equally weighted deficits incurred by the patient (Orkaby et al., 2019). These deficits account for impairments across multiple physiologic domains and were ascertained by the presence of specific sets diagnostic and procedure codes in patients’ records. We further classified the veterans as non-frail (VA-FI 0–0.1), pre-frail (VA-FI 0.1–0.2), mildly frail (VA-FI 0.2–0.3), and moderately frail (VA-FI 0.3–0.4) and severely frail (VA-FI >0.4), using previously published frailty index cutoffs (Armstrong et al., 2015; Blodgett et al., 2015; Pajewski et al., 2016).

To compare GW with non-GW veterans, we randomly selected 558 veterans from CDW who were matched in age and gender to the living GW veteran participants of SFVAHCS research studies. Because Congress never repealed the Authorization for Use of Military Force after the Gulf War ended in February 1991 (Department of Veteran Affairs, 2021), any military personnel who served on active duty from August 2, 1990 to present is considered a Gulf War veteran. Therefore, to identify non-GW veterans, we used a list of social security numbers obtained from the Defense Manpower Data Center (DMDC) of GW-deployed and non-deployed GW era veterans. Veterans were presumed to be non-GW veterans if their SSN was not on the DMDC list. We extracted ICD-10, CPT, and HCPCS codes from the CDW for non-GW veterans on the same index date.

To our knowledge, this is the first study to evaluate vulnerability to poor health outcomes in GW veterans using the frailty index as a proxy measure. Frailty is generally considered to be a state characterized by reduced physiological reserve and loss of resistance to stressors caused by accumulated age-related deficits (Clegg et al., 2013). Over the years, researchers have operationalized frailty in a couple of different ways. The frailty phenotype describes frailty as a biological syndrome with specific phenotypic presentations (e.g., unintentional weight loss, self-reported exhaustion, weakness, slow walking speed and low physical activity) (Fried et al., 2001). The frailty index operationalizes frailty as a state caused by the accumulation of health deficits during the life course (Mitnitski et al., 2001). Because the frailty index is calculated as a ratio of the number of deficits present to the number of total deficits considered, the more deficits one has, the more likely one is to be frail, the higher the frailty index. The VA-FI is an electronic index developed to use VHA administrative claims and electronic health records data to measure frailty specifically in Veterans (Orkaby et al., 2019).

The first major finding of this study is that GW veterans had lower VA-FIs (i.e., were healthier) than age- and sex-matched non-GW veterans. This finding was somewhat unexpected given previous reports of higher rates of chronic multisymptom illness, neurological, and other disorders in GW veterans compared to other veteran control groups (e.g., Ismail et al., 2008; Kang et al., 2009). However, there were more smokers among non-GW veterans compared to our GW veteran cohort. When we stratified the analysis by current smoking status, non-smoking non-GW veterans had marginally, but not significantly, higher VA-FIs compared to non-smoking non-GW veterans. Among smoking veterans, there was no significant difference between GW and non-GW veterans.

Because the two veteran groups were matched on age, it is likely a majority of the “non-GW” veterans are veterans of the U.S. wars in Afghanistan (Operation Enduring Freedom; OEF) and Iraq (Operation Iraqi Freedom; OIF; Operation New Dawn; OND). However, we could not verify this in VA records because all military personnel who served on active duty from August 2, 1990 to present are considered Gulf War veterans. This is because Congress never repealed the Authorization for Use of Military Force after the Gulf War ended in February 1991 (Affairs, 2021).

If, in fact, most of the “non-GW” veterans were OEF/OIF/OND veterans, one reason why this group of veterans had higher VA-FI than GW veterans may be due to traumatic brain injuries (TBI), the “signature injury” of the U.S. wars in Iraq and Afghanistan (Menon et al., 2010; Arriola and Rozelle, 2016). It is well documented that TBI can produce both acute and chronic consequences that lead to permanent disabilities that increase long-term mortality (Masel and DeWitt, 2010; Corrigan and Hammond, 2013) TBIs can also result in various secondary pathological conditions (Institute of Medicine, 2009). Because the changes initiated by TBI can persist for months to years after the initial injury (Bramlett and Dietrich, 2015), When we examined the individual VA-FI deficits, non-GW veterans had higher rates of anxiety, coronary artery disease, cerebrovascular disease, diabetes, hypertension, peripheral vascular disease, and weight loss than deployed GW veterans. It is noteworthy that a number of studies have reported increased risk for anxiety (Mallya et al., 2015), cardiovascular (Eric Nyam et al., 2019) and metabolic disorders (Kumar et al., 2018), including diabetes (Capatina et al., 2015) in the chronic phase of TBI recovery.

Another unexpected finding of this study is that rates of fatigue and chronic pain, two hallmark symptoms of GWI, were not higher in GW veterans compared to non-GW veterans at index date. However, not all of the GW veterans in the cohort had CMI/GWI at the time of the SFVHACS research studies. When we examined subgroups of GW veterans, we found differences in rates of chronic pain between controls and CDC CMI cases and between controls and Kansas GWI. Because the GW veterans were originally queried about their CMI/GWI symptoms 5–21 years prior to the index date, this suggests the veterans’ CMI/GWI symptoms of pain have not changed much over time. In contrast, rates of fatigue were not significantly different between controls and CMI cases or between controls and Kansas GWI cases. This suggests that, unlike pain, the veterans’ CMI/GWI symptoms of fatigue have likely changed or evolved over time.

Not all of the GW veterans in the retrospective cohort actually participated in research studies at the SFVAHCS. Some veterans were excluded because they did not meet the original studies’ inclusion criteria. Other veterans changed their minds and withdrew from the studies. Results indicate that veterans who were excluded from the original research studies had higher VA-FIs than veterans who were included and veterans who withdrew. This is not unexpected considering that the original SFVAHCS research studies’ exclusionary criteria included contraindications for MRI (e.g., pacemakers), severe physical impairment, psychiatric disorders with psychotic features, history of neurologic or systemic illnesses that affect central nervous system function, and/or moderate to severe traumatic brain injury. In fact, significantly more veterans who had been excluded from the original SFVAHCS research study were deceased at the index date compared to veterans who were included or who withdrew from SFVAHCS research studies.

There was a significant effect of CMI status on VA-FI: Veterans who met CDC CMI criteria at the time of the SFVAHCS studies had higher VA-FIs (i.e., were frailer) at the index date than controls (i.e., veterans who had insufficient symptoms to meet CMI criteria). Examination of individual VA-FI deficits revealed higher rates of anxiety, chronic pain, depression, lung disease, peripheral neuropathy, and dementia in CMI cases compared to controls. Considering that the mean age of the GW veteran cohort was 62 years at the index date, it is concerning that there was a higher rate of dementia (9.8% vs. 0.7%) among CMI cases compared to controls. We have previously reported a higher-than-expected rate of mild cognitive impairment (Chao, 2020), considered a prodromal phase of dementia (Petersen, 2011), in a subset of the GW veterans who are part of the current retrospective sample. Therefore, the current finding of a higher rates of dementia in CMI cases relative to controls lends further support to the idea that veterans with CMI are at increased risk for neurodegenerative diseases.

As with CDC CMI, there was a significant effect of Kansas GWI case status on VA-FI. As one might expect, GWI cases had higher VA-FIs than controls. However, veterans who were excluded from being considered as GWI cases because of medical and/or psychiatric conditions (i.e., GWI excludes) had even higher VA-FIs than GWI cases. When we examined individual VA-FI deficits, veterans who were excluded from being considered as GWI cases had higher rates of anemia, arthritis, coronary artery disease, depression, diabetes, hypertension, kidney disease, liver disease, peripheral neuropathy, and perivascular disease. This is not surprising because many of these conditions were likely what prevented these veterans from being considered as Kansas GWI cases at the time of the SFVAHCS studies. However, many of these conditions are also related to aging or may be accelerated by aging. Therefore, it is possible that not all of these medical conditions were pre-existing at the time of the SFVAHCS studies.

Several deployment-related experiences and exposures were significantly and positively correlated with VA-FI. These include hearing chemical alarms sound (a proxy measure for potential exposure to low levels of chemical nerve agent), using powdered pesticides, wearing flea collars, seeing ones’ living area sprayed or fogged with pesticides, and coming into contact with chemical agent resistant coating paint. Notably, most of these deployment-related experiences and exposures contain neurotoxins, particularly organophosphates. Previous studies have reported that many of the exposures associated with GW deployment are highly intercorrelated (Fricker et al., 2000; Cherry et al., 2001; Steele et al., 2012). In line with this, we also observed a high degree of correlation among deployment exposures in the GW veterans who completed the Kansas Gulf War and Health Questionnaire. Because this suggests a potential for confounding error when evaluating associations individually, in post-hoc analyses, we used a backward-step linear regression to examine the associations of deployment experiences and current health, as measured by VA-FI. After the least significant variables were removed, only Kansas GWI case status, wearing flea collars, and witnessing living are being sprayed or fogged with pesticides were significantly associated with VA-FI.

It is noteworthy that (Haley and Kurt, 1997) previously reported that wearing flea collars during the GW was associated with increased risk of developing Haley Syndrome 1 (impaired cognition). The Research Advisory Committee on Gulf War Veterans’ Illnesses (Research Advisory Committee on Gulf War Veterans Illnesses (Rac-Gwvi), 2008) has also cited evidence of an association between exposure to pesticides and multisymptom illness consistent with GWI. The current finding suggests that exposures to pesticides, particularly those present in flea collars, may be associated with frailty and current ill health independent of GWI.

This study had some limitations that warrant discussion. First, the cohort included only GW veterans who enrolled in research studies at the SFVAHCS and whose EHR were located in the VA’s CDW. This may limit the generalizability of these findings to the larger population of GW veterans. Presumably, the veterans whose EHR we were unable to locate in CDW were not utilizers of VA healthcare services. This may further limit the generalizability of these findings because it has been reported that VA users are more likely to have multiple medical conditions, endorse a disability, and score lower on measures of functioning than non-VA users (Meffert et al., 2019). Another limitation is the study’s reliance on self-reports, which is subject to recall bias, for health symptoms and deployment-related experiences and exposures. However, this limitation is common to other Gulf War research studies because there is no diagnostic test for GWI and self-report is currently the only method available for assessing GWI. Furthermore, because there was an absence of records or measurements quantifying chemical exposures during deployment, self-report is one of the only methods of investigating the effects of GW exposures on veteran health. It is possible that there were deployment-related exposures not assessed by the Kansas Military History and Health questionnaire that had an effect on the VA-FI. A final limitation is that the study was not adequately powered to examine potential sex differences in VA-FI and how biological sex may affect the relationship between exposures to deployment-related chemicals and current health status. Future, better powered studies will be necessary to address this question.

We evaluated GW veterans’ vulnerability to poor health outcomes using the frailty index as a proxy measure conclusion. Our results suggest that GW veterans, as a group, are not frailer (i.e., in worse health) than non-GW veterans who use VA healthcare. However, GW veterans who met criteria for severe CMI and who had Kansas GWI exclusionary conditions 5–21 years ago were significantly more frail than other GW veterans and non-GW veterans. We also found that GW veterans who met CDC CMI criteria had higher rates of dementia than control GW veterans. Thus, it may be prudent to counsel GW veterans with CMI to adopt lifestyle changes that have been associated with lowering dementia risk. The veteran subgroup with the highest VA-FI were GW veterans who had Kansas GWI exclusionary conditions. This suggests that research and treatment studies that employ the Kansas GWI case definition may be overlooking the subgroup of GW veterans who are most vulnerable to poor health outcomes.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by University of California, San Francisco Institutional Review Board. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

LC designed the experiment, analyzed the data, drafted the manuscript, contributed to the article, and approved the submitted version.

This study was supported by grant from the U.S. Department of Veterans Affairs (No. CX000798-05 and IK6CX002522).

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.