Observational and genetically-informed characterization of Gulf War Illness symptoms in 1990-1991 Gulf War deployed Veterans and non-deployed Gulf War Era Veterans.

Gulf War Illness (GWI) has been identified in more than 20% of Veterans following military service in the 1990-1991 Gulf War (GW). Unfortunately, after several decades, GWI pathogenesis remains unclear and the patterns of GWI symptom co-occurrence are still poorly understood.This population-based case-control study included 34,179 (16% females; 34% GW deployed; mean age 61 years) Veterans enrolled in the Million Veteran Program (MVP) who were active-duty military personnel in 1990-1991. We analyzed 14 GWI symptoms in GW-deployed and non-deployed GW-Era Veterans. Multivariable generalized linear models were used to estimate GWI symptom associations with demographic factors and deployment status. Correlation, factor, and network analyses were performed to examine the underlying structure of GWI symptoms. In addition, polygenic risk score and one-sample Mendelian randomization analyses were conducted to evaluate the genetically inferred effects of health-related traits on symptom factors.Among 14 GWI symptoms, joint pain had the highest prevalence in both GW deployed (86%) and non-deployed GW-Era Veterans (79%). The median age of symptom onset ranged from 38 to 46 years among deployed Veterans and from 45 to 54 years among non-deployed Veterans. Factor analysis identified a five-factor latent structure as the best-fit model for GWI symptoms, and latent class analysis classified Veterans into seven symptom classes. The network of GWI symptoms showed that fatigue had the largest closeness and betweenness and had more edges in deployed Veterans. Younger age, female sex, non-European descent, lower educational attainment, enlisted rank, and GW deployment, as well as multiple GW-related exposures in deployed Veterans, were associated with increased GWI symptom burden. Genetic liability to multiple health-related outcomes, including type 2 diabetes (T2D) and posttraumatic stress disorder (PTSD), showed putative causal effects on GWI-symptom factors, with the T2D effect being larger in deployed Veterans and the PTSD effect being larger in non-deployed Veterans.The present findings demonstrate differences in symptom onset between deployed and non-deployed GW Veterans. They also reinforce the importance of symptom clusters of relevance to GWI, associations between deployment-specific exposures and symptom burden, and different genetic linkages underlying symptoms for deployed and non-deployed groups.Not applicable.

Duke Scholars

Author Elizabeth Joanne Gifford Sanford School of Public Policy