Are muscle vasomotor function and physiology altered in Gulf War Illness?

What the study compared in 49 cases and 22 controls

Gulf War Illness is defined here by symptoms, not by a laboratory test: cases were classified using the symptom-based Kansas GWI questionnaire, and the comparison group was Veterans who did not meet that case definition. The authors asked whether four things differed between the groups — the ability of small muscle arteries to dilate, whole-body aerobic performance, muscle structure and mitochondrial content, and muscle gene expression.

The design matters for reading every result. This is a case-control comparison of peripheral muscle and vascular measures, so a null result speaks to those tissues and those measurements, not to every possible mechanism behind fatigue or muscle weakness. The authors also built in a check on their own method: they tested whether the same measurements could detect factors already known to impair vascular function.

Measuring vessel dilation with two different infusions

Vasomotor function was assessed in the profunda artery by infusing two substances directly into the artery and measuring the change in blood flow from baseline. Acetylcholine tests endothelium-dependent dilation, meaning the response requires the vessel's inner lining to act; adenosine tests endothelium-independent dilation, meaning the smooth muscle of the vessel responds without that lining step. Separating the two matters because they point to different mechanisms if a difference appears.

Neither response differed between groups. For acetylcholine at 10⁻⁶ M, the difference in percent change from baseline was −6.6% (95% CI −31.5% to 18.4%); for adenosine 50 µg it was −10% (95% CI −95% to 75%). The adenosine interval is wide, so this is a statement of no detected difference rather than proof of identical function. Conduit artery measures — cross-sectional area and its response to acetylcholine and nitroglycerin — were likewise similar.

Smoking was detected, which validates the method

The strongest evidence that the null result is not simply a blunt measurement comes from the covariates. In multivariable modeling, vasomotor function was significantly worse in subjects who smoked or had diabetes mellitus, and better in those reporting higher levels of vigorous exercise. Smoking was associated with a 42.5% lower endothelium-dependent response (95% CI −79.0% to −6.0%).

The authors describe these findings as internal controls that validate the methodology and the study's power to find important differences in vascular function. That framing is the authors' interpretation, but it rests on a concrete comparison inside the same dataset: a known risk factor produced a detectable signal while GWI status did not. Note the asymmetry, though — the smoking estimate comes from modeling, while the GWI estimates come from direct group comparisons.

Aerobic capacity and strength were similar, unlike heart failure

A symptom-limited cardiopulmonary exercise test measured oxygen uptake during exercise. Peak V̇O₂ differed by −1.5 mL/kg/min (95% CI −4.8 to 1.9), and V̇O₂ at the anabolic threshold — the point where ventilation begins to rise disproportionately — differed by −1.0 mL/kg/min (95% CI −4.1 to 2.1). Peak oxygen pulse, the V̇E/V̇CO₂ slope and the oxygen uptake efficiency slope were also similar.

Muscle strength and function pointed the same way: handgrip strength in each hand, one-repetition maximum leg strength, leg peak power and fatigability at 60% of one-repetition maximum showed confidence intervals spanning zero. The six-minute walk distance was 38 m shorter in cases (95% CI −80 to 5), an interval that includes no difference. The authors contrast this profile with chronic heart failure, where peak V̇O₂, threshold V̇O₂ and walk distance are significantly lower than in controls.

Muscle structure and mitochondrial content looked normal

Vastus lateralis biopsies were frozen and examined in several ways: hematoxylin and eosin staining for structure and centrally located nuclei as an index of muscle pathology or regeneration, immunohistochemistry and fluorescent labeling for fiber and mitochondrial distribution, and confocal microscopy to estimate sarcomere length. Mitochondrial volume and density and the content of glycolytic, anaerobic, tricarboxylic acid cycle and mitochondrial oxidative phosphorylation enzymes were measured by western blot.

The histology and protein comparisons used eight pairs of subjects with the highest and lowest total Kansas scores rather than the full cohort, and were tested with Wilcoxon rank-sum tests. Nuclei relative to total cell area and centrally located nuclei as a share of all nuclei did not differ (P = 0.16 for both), and sarcomere length was not significantly different. The authors report similar mitochondrial volume and density and similar enzyme content.

RNA sequencing found no metabolic pathway tied to GWI

Bulk RNA sequencing of muscle was used for unbiased gene expression, with differential expression analyzed in DESeq2 and pathway and gene ontology enrichment in DAVID, using a false discovery rate below 0.05. The authors report no metabolic pathway gene dysregulation associated with GWI.

The paper is explicit that this analysis may be underpowered to find small differences between cases and controls, and that the gene expression comparison used five pairs at the extremes of the Kansas score. The authors' conclusion is narrower than 'no gene differences exist': there were no consistent aberrations in biological pathways that could explain the symptoms, concordant with the absence of differences across the other vascular and exercise measures.

Self-reported exercise diverged from measured capacity

Self-reported vigorous and moderate exercise was significantly lower in GWI cases, yet objective aerobic performance and muscle strength were not significantly different. The authors read this as evidence that self-reported exercise is not a reliable measure of actual exercise capacity.

That divergence is one of the more practically useful observations in the paper, because it warns against using activity questionnaires as a stand-in for measured fitness in this population. It also sits alongside the authors' positive message: they found no intrinsic or irreversible difference in skeletal muscle structure or function, and conclude that GWI patients should obtain a similar training effect from physical activity as those without GWI.

Where the conclusion stops

The authors state they cannot comment on whether physical trauma or illness at the time of the Gulf War acted as a sensitizing event leading to a functional disorder, and they did not investigate other causes such as central nervous system or endocrine disorders. The measures here are peripheral: small-artery vasomotion, whole-body exercise performance, muscle histology and muscle gene expression.

The paper places GWI alongside other symptom-defined chronic illnesses such as long-COVID, myalgic encephalomyelitis/chronic fatigue syndrome and fibromyalgia, which it describes as distressing and disabling but lacking consistent pathophysiological abnormalities at current resolution. That comparison is the authors' framing, and it defines the boundary of this study: the negative findings concern muscle and peripheral vasculature, not a claim that the symptoms are unreal.

What to take from the comparison

The decision-relevant point is a contrast within one dataset: a known vascular risk factor, smoking, produced a 42.5% lower endothelium-dependent response, while GWI status produced a difference of −6.6% with a confidence interval spanning zero. When a study can register the former but not the latter, the absence of a GWI difference in these peripheral measures becomes more informative than a bare null result.

For readers weighing mechanisms, the priority is therefore to look beyond muscle and peripheral artery function — toward the central, sensitization or functional explanations the authors explicitly left untested — rather than to treat this paper as evidence about all proposed causes of GWI. Its practical implication is narrower and better supported: measured aerobic capacity and strength were similar, so exercise capacity in GWI should be judged from objective testing rather than self-report.

Sources used in this answer