The Physical and Mental Health of Post-9/11 Female and Male Veterans: Findings from the Comparative Health Assessment, 2025, Dursa et al

[forestglip]

The Physical and Mental Health of Post-9/11 Female and Male Veterans: Findings from the Comparative Health Assessment

Erin K. Dursa, Yasmin S. Cypel, William J. Culpepper, Paul A. Bernhard, and Aaron I. Schneiderman

Background
Females are the fastest-growing group in the veteran population, yet there is a paucity in the literature of sex-specific results from studies of chronic disease in veterans that limit our understanding of their health issues. This study provides nationally representative estimates of the physical and mental health of females and males from the Operation Enduring Freedom, Operation Iraqi Freedom, and Operation New Dawn (OEF/OIF/OND) veteran population.

Methods
Data from the 2018 Comparative Health Assessment Interview Research Study (CHAI), a cross-sectional nationwide survey of the health and well-being of OEF/OIF/OND veterans and a comparison sample of U.S. nonveterans, were analyzed to provide sex-stratified and deployment-stratified lifetime prevalence estimates and adjusted relative odds of physical and mental health conditions in a large population-based study of OEF/OIF/OND veterans.

Results
Overall, female veterans were significantly more likely to report cancer, respiratory disease, irritable bowel syndrome/colitis, bladder infections, vision loss, arthritis, back/neck pain, chronic fatigue syndrome, migraine, posttraumatic stress disorder, and depression. Male veterans were significantly more likely to report obesity, diabetes, heart conditions, hypertension, high cholesterol, hearing loss, fractures, spinal cord injury, sleep apnea, and traumatic brain injury. Both males and females who deployed were significantly more likely to report adverse health outcomes than those who did not deploy.

Conclusion
This article reports sex-stratified and deployment-stratified lifetime prevalence estimates and adjusted relative odds of physical and mental health conditions in a large population-based study of OEF/OIF/OND veterans. This study demonstrates the value of epidemiological research on female veterans and its importance in understanding the burden of disease in the female veteran population.

Link | PDF (Journal of Women's Health) [Open Access]

 

[forestglip]

Some data on lifetime prevalence of "chronic fatigue syndrome" as determined by:

measured using the 2018 National Health Interview Survey question: “Has a doctor or other health care provider ever told you that you had any of the following conditions?”. Respondents reported “yes” or “no”

Female nondeployed: 4.6% (3.6–5.7)
Female deployed: 6.9% (6.0–7.8)

Male nondeployed: 2.0% (1.3–2.7)
Male deployed: 4.6% (4.1–5.2)

 

[Utsikt]

Didn’t some other study find that only half ot the people that have been told they have CFS, actually had it according to the criteria?

 

[Hutan]

That's a really high lifetime prevalence, even at half the rate. I wonder how they determined lifetime prevalence.

Sampling procedure:

The 2018 CHAI is a cross-sectional nationwide survey of the health and well-being of OEF/OIF/OND veterans and a comparison sample of U.S. nonveterans. A stratified random sampling of the U.S. Veterans Eligibility Trends and Statistics (USVETS) file23generated the veteran sample (n = 67,500) of adults 18 years of age and over. Females were oversampled to account for 30% of the total veteran sample. One third of the veteran sample was set aside to be surveyed if response rates fell below expected levels. Nonveterans (n = 16,843) with no prior military experience were drawn via address-based sampling from a nationally representative panel of the noninstitutionalized U.S. adult population (KnowledgePanel®).24 Further details about the nonveteran comparison sample have been published elsewhere.25–27

In total, 38,633 veterans were asked to report about their health by completing a web-based questionnaire or a computer-assisted telephone interview via recommended methodologies.28 Initial mail contacts to veterans included a $1 preincentive and an additional $50 postincentive for those who completed the survey. All participants provided informed consent prior to survey administration. Study procedures were approved by the VA Central Institutional Review Board.

The veteran response rate was 39.5% (n = 15,170 eligible, returned surveys) in the range of earlier large-scale studies of post-9/11 veterans.18,29 Four veterans were removed because their deployment status could not be ascertained. The total analytic sample comprised 15,166 veterans (5,642 female veterans and 9,524 male veterans).

So, they asked 38633 veterans to report about their health with a $50 incentive for completion. The response rate was 39.5%. Within the veterans were deployed and non-deployed, and male (9,524) and female (5,642).

They also asked 16,843 people with no military experience to complete the survey - these were the non-veterans (male and female).

 

[Hutan]

There might be an error in Table 1:

Table 1 shows descriptive statistics for baseline characteristics stratified by sex. The veteran sample comprised 17.3% female and 82.7% male.

But, here's the top part of Table 1


5,642 is not 17.3% of 15,166. It's more like 37%. I don't know what else they could be reporting. 17.3% +82.7% = 100%

If they can't get as something as basic as that right, it makes me wonder about all of the adjustments they made to the data, to account for confounding.

 

[forestglip]

5,642 is not 17.3% of 15,166. It's more like 37%. I don't know what else they could be reporting. 17.3% +82.7% = 100%

Interesting, good catch. I'll email the author.

 

[Nightsong]

Interesting, good catch. I'll email the author.

The authors state that all statistics were weighted, except for the counts:

All statistics, except raw counts, were weighted to account for CHAI’s complex sampling design and nonresponse.
Weights were further calibrated to frame totals on veteran sex and service characteristics.

Each data table also contains the footnote "Weighted statistics except for count".

 

[Hutan]

The authors state that all statistics were weighted, except for the counts:

Ah, that makes good sense. I was finding it hard to believe that the authors could get that wrong.

I think they could have presented things better. I remain a bit concerned about the adjustments that were made. (Disclaimer: I didn't finish reading the paper before I had to stop last night.)

With a response rate of 39.5%, perhaps the people with ill health to report were more likely to participate in the survey. But, even so, the rates of CFS are incredibly high. Are doctors in the US handing out diagnoses of CFS like lollies?

 

[forestglip]

With a response rate of 39.5%, perhaps the people with ill health to report were more likely to participate in the survey. But, even so, the rates of CFS are incredibly high. Are doctors in the US handing out diagnoses of CFS like lollies?

I haven't read this in detail, but I think this only gives prevalence for deployed and non-deployed veterans. The doctors these groups see might mainly treat veterans and thus are more familiar with GWI and related syndromes than regular doctors, so might be more likely to make that diagnosis.