© RSNA, 2011
Subjects
Three years after the 1991 Gulf War, an epidemiologic study of 249 Gulf War veterans of the 24th Reserve Naval Mobile Construction Battalion was performed, and three primary and three minor Gulf War-related syndromes were classified (2). Subsequently, samples of the sick and well members of the battalion were studied in a nested case-control study to probe and understand the illness in that unit (3). In 1998, an initial SPECT cholinergic challenge experiment (17) was conducted in a sample of 42 battalion veterans (18 healthy control subjects, five veterans with syndrome 1, 12 veterans with syndrome 2, five veterans with syndrome 3, and two veterans with minor syndromes).
For the present study, 34 members of that sample (16 control subjects, four veterans with syndrome 1, 10 veterans with syndrome 2, three veterans with syndrome 3, and one veteran with one of the minor Gulf War-related syndromes) were brought back for reevaluation; this population was augmented to increase power with 23 additional Gulf War veterans (seven with syndrome 1, seven with syndrome 2, and nine with syndrome 3) who were also from the original epidemiologic study. Two subjects (one new veteran with syndrome 3 and one original control subject) did not participate because of claustrophobia. Three subjects (one control subject and two veterans with syndrome 2) did not complete the second session of imaging. One patient with syndrome 2 had unacceptably large movement during imaging. Two subjects (one control subject and one patient with syndrome 2) were not included in the analysis because of a slightly different infusion protocol. The one subject with the minor syndrome classification was also not included. This left 48 participants (13 control subjects [mean age ±standard deviation, 60 years ± 6]; 11 veterans [mean age, 51 years ± 6] with syndrome 1, which is associated with impaired cognition [distractibility, forgetfulness, depression, and fatigue]; 13 veterans [mean age, 63 years ± 6] with syndrome 2, which is associated with confusion-ataxia [reduced intellectual processing, confusion, vertigo, and disorientation]; and 11 veterans [mean age, 57 years ± 6] with syndrome 3, which is associated with central pain [joint pain in the extremities and in the neck and shoulders, muscle pain in the arms, and tingling numbness in the extremities]) in the hippocampal rCBF study with physostigmine challenge reported here.
Linear structural equation modeling (3) was used to express the dependence of major symptom scales on the three syndrome factors as regression coefficients and a residual error term. Compared with the prior syndrome classification by structural equation modeling in the 1994 study (3), this 2008-2009 study showed significantly lower syndrome 1 symptom scores (P = .007) and significantly higher syndrome 2 (P = .005) and syndrome 3 (P = .043) symptom scores.
Hippocampal Volumes
The hippocampus is a relatively small structure, and small systematic differences in hippocampal volume among the groups could impact the rCBF measurements. Victims of the 1995 Tokyo subway sarin attack had reduced hippocampal volume, as well as reduced volumes in the insular cortex and surrounding white matter (40). Significantly smaller hippocampi and trends toward reduced gray matter volumes in frontal, parietal, and occipital cortices were recently found in veterans of the 1991 Gulf War who were exposed to sarin or cyclosarin, independent of syndrome classification (41). However, voxel-based morphometry of the high-spatial-resolution anatomic images in the current study showed no significant group differences. Furthermore, the possible confounding of differential volume averaging due to individual or group hippocampal volume differences was mitigated by carefully choosing for analysis only those ASL voxels within the hippocampus, utilizing a hippocampus ROI mask derived from the high-resolution anatomic image data of each subject. For the 2 × 2 × 5 mm3 (+1-mm gap) ASL voxels, 3 or 4 voxels fit left to right into the hippocampus head, 2 or 3 voxels fit left to right into the hippocampus body, and 1 or 2 voxels fit left to right into the hippocampus tail; in the anteroposterior (head-to-tail) direction, about eight sections fit into the hippocampus. The means and standard deviations of the numbers of voxels that fit into the bilateral hippocampus were as follows: 316 ± 26, 321 ± 18, 305 ± 26, and 309 ± 31 for healthy veterans and those with syndromes 1, 2, and 3, respectively, corresponding to 7.6 cm3 ± 0.6, 7.7 cm3 ± 0.7, 7.3 cm3 ± 0.6, and 7.4 cm3 ± 0.8. Although these data indicate a slight trend for hippocampal atrophy of about 5% in veterans with syndrome 2 or syndrome 3 compared with control subjects and veterans with syndrome 1, one-way analyses of variance performed separately for the left hippocampus (P = .73), the right hippocampus (P = .41), and both hippocampi (P = .48) indicated no significant group differences in hippocampal volumes. Older subjects had smaller hippocampi (P = .052), but the effect of age on hippocampus rCBF did not depend on group (P = .46) or PTSD status (P = .28). The effect of hippocampus volume on hippocampus rCBF was similar in each group (P = .49) and also similar with respect to PTSD status (P = .51).