The Implications of U.S. Department of Veterans Affairs’ Limited Scope of Gulf War Illness Research340 Cannon House Office Building
- Hon. Harry E. Mitchell, Chairman, Subcommittee on Oversight and Investigations
- Hon. John J. Hall, a Representative in Congress from the State of New York
- Lynn Goldman, M.D., MPH, Professor, Bloomberg School of PublicHealth, John Hopkins University, Baltimore, MD, and Member, Committee on Gulf War and Health, Institute of Medicine, The National Academies
- Accompanied By: Roberta Wedge, M.S., Senior Program Officer, Board on the Health of Select Populations, Institute of Medicine, The National Academies
- James H. Binn, Chairman, Research Advisory Committee on Gulf War Veterans’ Illinesses
Lea Steele, Ph.D., Adjunct Associate Professor, Kansas State University School of Human Ecology, Manhattan, KS, and Former Scientific Director, Research Advisory Committee on Gulf War Veterans’ Illnesses
Robert W. Haley, M.D., FACE, FACP, Professor of Internal Medicine – Epidemiology, University of Texas Southwestern Medical Center at Dallas, TX
Roberta F. White, Ph.D., Professor and Chair, Department of Environmental Health, and, Associate Dean for Research, Boston University School of Public Health, Boston, MA
Anthony Hardie, Madison, WI (Gulf War Veteran)
Submissions for the Record
Major Denise Nichols, RN, MSN, USAFR (Ret.), Vice Chair, National Vietnam and Gulf War Veterans Coalition
Hearings » The Implications of U.S. Department of Veterans Affairs’ Limited Scope of Gulf War Illness Research
Statement of Robert W. Haley, M.D., FACE, FACP
Professor of Internal Medicine – Epidemiology
University of Texas Southwestern Medical Center at Dallas, TX
Good morning, Mr. Chairman and distinguished Members of the Subcommittee. I want to thank you for inviting me to describe our research program on Gulf War illness at the University of Texas Southwestern Medical Center in Dallas and our collaborating universities and research organizations across the country. To introduce myself briefly, after training in Internal Medicine, I served for 10 years at the U.S. Centers for Disease Control and Prevention (CDC) where I received a US Public Health Service commendation medal for my research on controlling hospital-acquired infections. Since 1983 I have been on the faculty of the University of Texas Southwestern, doing clinical research, teaching research design to our young assistant professors, and supervising an Internal Medicine service at Parkland Hospital. During my first 10 years on the faculty, I volunteered as an attending physician at the Dallas VA Medical Center.
Initial Studies, 1994-1998
In 1994, three years after the first Gulf War, Ross Perot visited with our university president and me, as Director of Epidemiology. He acquainted us with the newly emerging problem of Gulf War syndrome, and asked if UT Southwestern would undertake a study of the problem with funding from the Perot Foundation. I put together a small research team and performed an epidemiologic survey and follow-up clinical study of the 24th Reserve Naval Construction Battalion (Seabees) that had served in the Gulf War. While the research world at the time was focused almost exclusively on stress and psychological explanations, our studies pointed clearly toward a physical illness. Our findings raised the following three provisional hypotheses to be explored in further research:
1.The Gulf War syndrome appeared to be a real physical brain illness—a chronic encephalopathy—with 3 subtypes, or variants.
2.The many symptoms appeared to be due to damage to cells in different deep brain structures.
3.The damage appeared to be caused by wartime exposure to combinations of neurotoxic chemicals, including low-level sarin, organophosphate (OP) pesticides and the pyridostigmine bromide (PB) anti-nerve agent medication.
These initial findings were published in January 1997 in three high profile peer-reviewed articles in the prestigious Journal of the American Medical Association.
The media reaction from that publication introduced me to several young Gulf War veterans dying of Lou Gehrig’s disease (amyotrophic lateral sclerosis, or ALS). My subsequent investigation documented a statistically significant three-fold increase in the rate of ALS in atypically young Gulf War veterans. When this was subsequently verified by a VA study, it led to service connection for all military veterans with ALS.
Second Round of Studies, 1998-2001
With leadership and staunch support from Texas Senator Kay Bailey Hutchison, the Department of Defense provided substantial research funding to follow up our findings. We began to explore new medical technologies that would probe directly for the nature and mechanisms of the hypothesized brain cell damage to give doctors a rational basis for diagnosing and treating it and give VA objective tests for determining service connection in these veterans. Here is a summary of the main findings from this work, which spanned 1998 to 2001 and was described in prominent peer-reviewed scientific publications.
1.Chemical Evidence of Brain Cell Damage. We performed brain scanning with an MRI-based technique called Magnetic Resonance Spectroscopy (MRS), which measures chemical concentrations in small brain regions of living subjects. We found evidence of chemical alterations in the deep brain structures of ill Gulf War veterans compared to well veterans. This type of chemical change is characteristic of physical brain cell damage and is not found in stress and psychological reactions. The group of veterans with the syndrome 2 variant (“confusion-ataxia”) had evidence of more severe brain cell damage than the syndrome 1 (“impaired cognition”) and 3 (“central pain”) variants.
2.Abnormal Production of Brain Dopamine. We performed chemical assays of metabolites of the brain neurotransmitter dopamine (recall that reduced production of brain dopamine causes the symptoms of Parkinson’s disease). We found evidence that the brains of ill Gulf War veterans with the syndrome 2 variant were overproducing dopamine. Other research shows that dopamine excess can cause cognitive and emotional symptoms like those described by many Gulf War veterans.
3.Abnormality of Autonomic Nervous System. We used special computer modeling of 24 hour electrocardiogram (EKG) recordings to test for subtle abnormalities of the autonomic nervous system, which we suspected of causing symptoms like chronic diarrhea, sexual disturbance, excessive gallbladder disease, unrefreshing sleep and body temperature dysregulation. The results were consistent with loss of the normal day-night fluctuation in parasympathetic nervous system activity, which would indicate abnormal function of the autonomic nervous system. This abnormality was equally present in all three of the syndrome variants.
4.Locating Damaged Brain Areas. A fascinating experiment involved performing a SPECT scan of brain blood flow before and after infusion of a drug physostigmine that safely mimics the brain effects of sarin, OP pesticides and PB. Our prediction was that if the ill veterans’ brains had been damaged by these neurotoxic chemicals, their brain function would not respond like normal subjects to a repeat exposure. The findings were consistent with the prediction. In normal Gulf War veterans, the medication appeared to reduce blood flow, but it paradoxically increased blood flow in the ill Gulf War veterans with the syndrome 2 variant and showed other abnormal patterns in the syndrome 1 and 3 variants. A provisional diagnostic test modeled from these data discriminated each of the syndrome variant groups from each other and from the well veterans. Equally important, this experiment gave evidence that specific parts of the brain appeared to be damaged and not responding normally. These findings gave us a valuable starting place for designing the next round of studies probing specific brain areas found here to be abnormal.
5.Discovery of a Susceptibility Gene PON1. To try to explain why some Gulf War veterans developed this chronic encephalopathy while others working next to them did not, we studied the function of a susceptibility gene called PON1 that produces the blood enzyme paraoxonase that protects our brains from neurotoxic chemicals like sarin and OP pesticides. We found indeed that the ill Gulf War veterans were born with abnormally low levels of paraoxonase, making them highly susceptible to these neurotoxic chemicals; whereas, the well veterans were born with normal to high levels. We then developed a gene therapy product containing the PON1 gene, injected it into mice, and found that it protected their brains from damage caused by exposure to OP pesticides. The university has a patent application pending, and if awarded, a possible product to protect people from OP pesticide and nerve agent exposure might result.
Development of New Technology To Measure
Subtle Brain Abnormalities, 2001-2006
During the second round of studies it appeared that the brain cell damage in Gulf War illness is sufficiently subtle that the approach we had been using would not be sufficient and that we would have to develop new technology, or adapt existing cutting-edge technology, to thoroughly understand the condition and develop clinically useful diagnostic tests. Therefore over the next 5 years we concentrated on an intense technological development effort. To make this possible, we enlisted some of the top brain scientists and technology experts from the North Texas region and from universities throughout the country. These included the University of Texas at Arlington, the University of Texas at Dallas, Southern Methodist University, and the Johns Hopkins University, Emory University, and University of Florida schools of medicine
Again with stalwart support from Senator Kay Bailey Hutchison, additional funding was provided through the Department of Defense to accomplish the following technological development projects.
1.A New High Performance Brain Imaging Center Dedicated to the Problem. UT Southwestern Medical Center dedicated space in the imaging center for the most powerful FDA-approved MRI scanner (with 3 Tesla strength), with all the peripheral equipment configured for brain imaging studies of Gulf War illness, opened in 2004.
2.High Resolution Imaging of Small Brain Structures. Our physicists developed new MRI techniques to obtain very high resolution images of small brain structures, such as the brain’s center for memory (the hippocampus) and sensation (the individual nuclei of the thalamus) not normally visualized adequately in standard MRI scans.
3.Rapid MRI-Based Tests to Replace SPECT. Although the prior SPECT study was very successful and offered a possible future diagnostic test for Gulf War illness, the protocol required two full afternoons and exposed the research subjects to radiation, both characteristics that reduce its usefulness in a diagnostic clinic setting. We therefore adapted an emerging MRI-based technique called Arterial Spin Labeling (ASL) that can obtain the same information as SPECT but in a 3-hour test on one day without radiation exposure. After validation, a provisional patent application was filed.
4.Functional MRI (fMRI) Tests to Probe Symptoms. The central problem in diagnosing and treating veterans with Gulf War illness is that the veterans’ complaints are all subjective symptoms with no objective signs. To provide medical understanding of the symptoms, we developed for each symptom an fMRI "probe" to demonstrate observably how the brain is functioning when a Gulf War veteran experiences a given symptom. We developed an fMRI test to probe each of the major symptoms, such as problems with fatigue, memory, attention and concentration, word-finding, rapid thinking and reaction (executive function), body pain, depressed feelings, and emotional lability.
5.MRI and EEG Tests of Functional Connections among Brain Structures. To assess the effects of brain cell damage on overall brain function, we adapted cutting-edge technology called Functional Connectivity to measure the amount of “electrical traffic” among brain areas. This measures how much different brain areas are “talking” with each other. Damage to a given brain area, or the “wires” between them, reduces or eliminates the electrical transmission between them and usually causes the brain to establish alternate "work around" pathways that bypass the deficit. Knowing the state of the functional connections could inform rehabilitation treatments.
6.MRI-Based Tests of the Brain’s “Wiring.” Some chemicals are known to damage the nerve bundles that connect different parts of the brain, and they can damage either the nerve bundles themselves or the insulation (myelin sheath) that covers the nerve bundles. To measure these we developed a test using the MRI-based technology Diffusion Tensor Imaging (DTI). This approach was used by Japanese researchers to demonstrate brain abnormalities in survivors of the terrorist sarin attack in the Tokyo subway, who were left with a chronic encephalopathy similar to that in Gulf War veterans.
7.High Resolution EEG. While MRI-based brain imaging shows spatially what is going wrong in the brain at a given point in time, we have developed a high-resolution electroencephalography (EEG) laboratory to measure the timing of sequential events in the brain pathways damaged by Gulf War chemical exposure. EEG testing is relatively quick and cheap and is likely to figure importantly in a clinical diagnostic strategy. Its high resolution implementation could also discover the order and timing of brain events amenable to rehabilitation treatment strategies.
8.Innovative Statistical Tests to Maximize the Power of Brain Imaging Tests. Since the cutting-edge brain imaging techniques being used are barely a decade old, few sophisticated statistical techniques have been developed for analyzing the complex data, and the relatively crude techniques available are typically not very powerful in detecting the types of subtle brain abnormalities that affect ill Gulf War veterans. We therefore developed a new body of statistical theory and applications that greatly increase the power of brain imaging tests and have incorporated them into a software package for which a patent is pending. This should have wide application beyond this program.
9.PON Laboratory. In the past the laboratory techniques used to measure the paraoxonase enzyme activity and the different forms of the PON1 gene that protect us from low-level nerve agents and OP pesticides have required time-consuming test tube chemistry not feasible to apply to large numbers of veterans in research studies. To overcome this we set up a special PON Laboratory headed by an expert on PON chemistry, and he has developed rapid, high throughput assays that can be used in large-scale studies.
10.National Survey. In the first two phases of our research, we performed our studies on Gulf War veterans from a single naval reserve Seabees Battalion. To determine whether the findings in this Battalion apply to the larger population of Gulf War veterans in general, we collaborated for a number of years with the well known research organization Research Triangle Institute International (RTI) in designing a computer-assisted telephone interview survey to apply in a randomly selected sample of all Gulf War veterans. Subsamples of the ill and well veterans selected from this nationally representative sample will be studied by the new brain imaging techniques.
11.Mouse Model of Chemical Brain Damage. To develop effective treatments for a new disease it is often necessary to understand how the disease-producing process works at the cellular and molecular levels so that new drugs or rehabilitation strategies can be directed precisely at the offending element. To make this kind of research possible, our Neurotoxicology Laboratory has developed a mouse model in which we can administer the neurotoxic chemicals to laboratory mice by a carefully tested recipe that results in a chronic behavioral disturbance comparable to Gulf War illness in humans. The brains of these mice can be studied by neuroscientists to discover exactly if and how neurotoxic chemicals damage brain cells.
Third Round of Studies, 2007-Present
After more than a decade since the initial research results on ill Gulf War veterans and with the new testing technology for studying subtle brain damage now in hand, Senator Hutchison championed and spearheaded a substantial new Gulf War illness research program through the VA hopefully to gain a higher level of understanding of the disease, a practical diagnostic approach, and ideas for treatment to be tested in clinical trials. The research program, designed to implement the 2004 Research Recommendations of the VA Research Advisory Committee on Gulf War Veterans’ Illnesses (RAC), has three basic components:
1.A National Survey and Serum/DNA Bank of Gulf War-Era Veterans
2.A Series of Neuroimaging and Biomarker Studies
3.Pre-Clinical Studies of the Mouse Model
These components were designed on an “industrial model” much like a defense contract program to develop a major new weapons system. All of the components were developed to interact with each other so that the findings of each would inform the progress of the others. At present we are approximately 2 years into the National Survey, 18 months into the Neuroimaging studies and 9 months into the Pre-Clinical studies. The following are summaries of progress to date.
National Survey and Serum/DNA Bank of Gulf War-Era Veterans. To date we have completed the extensive standardized telephone interview with 8,018 Gulf War-era veterans to measure the manifestations of the illness, risk factors, and family impact of the illness. By the end of August, we should have completed the field work for collecting and banking serum, plasma, RNA, and DNA from all of the ill veterans and a random sample of well veterans, comprising a total of approximately 2,100 survey participants. The following are provisional findings from the initial analysis of the survey data.
•Provisional finding: Regardless of the case definition used, chronic Gulf War illness appears to be 3- to 4-times more common in the deployed than the non-deployed Gulf War-era populations, and this difference is statistically significant in the studies thus far.
•Provisional finding: The findings support the conclusion of the 2004 and 2008 RAC reports that, from subtracting the prevalence of Gulf War multisymptom illness (CDC definition) in the non-deployed population from that in the deployed population, in 2007-2008 approximately 23% of the deployed force still had the chronic multisymptom illness from deployment-associated exposures.
•Provisional finding: The three clinical variants of the Gulf War illness, described in our prior studies, were identified again and appear to be strongly validated in the data from the national sample. This suggests that the chronic multisymptom illness identified in the Seabees unit by our prior studies is the same as that affecting the larger population of Gulf War veterans.
•Provisional finding: In the naval reserve Seabees Battalion surveyed first in 1995, the prevalence rate of Gulf War illness appears to have remained relatively unchanged over the intervening 12-13 years, except that the milder syndrome 1 variant initially affecting younger Gulf War veterans tended to have evolved toward the more severe symptoms of the syndrome 2 variant as these individuals aged.
•Further analyses of the survey data are proceeding, assays of paraoxonase enzymes and PON1 genes are nearly complete, and we have selected subsamples of ill and well veterans to participate in the next phase of the Neuroimaging and Biomarker Study.
Neuroimaging and Biomarker Study. Because of the complexity of studying a new brain disease, this component was designed in at least three sequential phases: a) conducting developmental pilot studies to validate the new neuroimaging techniques in normal volunteers, b) performing the complete battery of new tests on the members of the Seabees Battalion studied previously to see whether the disease had changed in the decade since the prior studies and to confirm whether the new tests detect the targeted abnormal brain function, and c) final verification of the findings in random subsamples of the participants in the National Survey of Gulf War Veterans. Provisional findings are as follows.
•Provisional finding: Findings of the prior SPECT experiment were reproduced, and we verified that MRI-Based ASL provides comparable results as the more involved and invasive SPECT, providing a far more efficient and safer diagnostic test.
•Provisional finding: The prior MRS findings of chemical abnormalities in deep brain structures (basal ganglia) were reproduced, and the findings were extended to abnormalities in hippocampus.
•Provisional finding: DTI identified a mild abnormality of myelin in white matter in ill Gulf War veterans.
•Provisional finding: EEG found an increase in slow brain waves in ill Gulf War veterans consistent with neurotoxic brain injury.
•Provisional finding: Functional Connectivity tests identified abnormal increase in brain communication in ill Gulf War veterans, indicative of generalized brain hyperarousal.
•Provisional finding: fMRI tests identified abnormal brain patterns underlying the major symptoms in ill Gulf War veterans.
◦fMRI test of Learning and Remembering identified abnormal function in the brain’s memory center (hippocampus).
◦fMRI test of Working Memory found that ill veterans do not use the normal rapid memory pathways but, instead, an inefficient slower work-around pathway.
◦fMRI test of Attention and Concentration identified abnormal function in deep brain structures that normally direct attention and concentration (basal ganglia).
◦fMRI test of Word Generation identified abnormal function in the basal ganglia.
◦fMRI test of Pain Processing found exaggerated response to pain sensation in the cerebral cortex.
◦fMRI test of Emotional Control found activation of abnormal pathways for managing emotionally evocative stimuli.
•High Resolution MRI images have identified abnormal cavities in the brain’s memory center (hippocampus) in ill veterans, suggesting chronic effects of brain cell damage.
In our latest study of the Seabees battalion, virtually every neuroimaging test showed evidence of substantial differences between sick and well groups of Gulf War veterans. This suggests that our unique brain imaging program might explain most symptoms and provide powerful objective diagnostic tests for clinical use and determination of service-connected status. It also provides a rich mosaic of evidence to suggest mechanisms of the brain dysfunction to be further tested in our pre-clinical mechanistic studies the third component of the ongoing program).
The uniform success of the tests is due to our strategy for developing the imaging tests by targeting veterans’ symptoms and the brain regions known to perform the implicated functions and the clinical classification of veterans into the three syndrome variants identified in our initial studies. Since the findings differ somewhat among these clinical variants, failing to test and analyze the groups separately would have resulted in less powerful, or even negative, findings.
As far determining which neural mechanisms are in play, we have not analyzed the data sufficiently yet to favor one mechanism over others. We can state the following general working hypotheses about mechanisms:
a.Although we find evidence of abnormalities in both deep gray matter and white matter, primacy of the deep gray matter involvement seems likely (e.g., pain is not a symptom of primary diseases of white matter).
b.Deep gray matter abnormalities identified appear bilaterally asymmetrical.
c.White matter abnormality appears to involve myelin, rather than axonal, degeneration. If correct, this is optimistic for treatment; myelin may be more amenable than axonal damage.
Besides explaining the specific deficits, the mosaic of evidence points to certain general findings:
a.Structures activating during a task in well veterans often do not activate in sick veterans, but other structures do. This abnormal activation probably indicates the brain’s attempts to compensate for, or work around, damaged areas.
b.The brain in sick veterans appears to be hyper-aroused and hyper-responsive to stimuli.
1.The brain appears to be working overtime to overcome the many deficits.
2. Chronic fatigue may be due to the brain’s exhaustion from this overwork.
3.The emotional lability and hyper-reactivity may also be due to this overwork.
The symptoms of veterans suffering from Gulf War illnesses are subjective, and the causes, diagnoses, and treatments are elusive. Therefore, a guiding principle for this research program has always been that objective studies—verified by researchers at different institutions and replicated in representative and increasingly-large samples of veterans—are required to arrive at conclusions on which action can be based.
With this rigorous approach, the findings to date make us optimistic that this multi-perspective testing protocol might lead to objective diagnosis. If it continues to progress along these lines, the testing approach should prove useful for future clinical and research work in the following ways:
a.Developing an objective diagnostic testing protocol for clinical work and service connection.
b.Providing pathogenetically homogeneous groups for clinical trials so that promising treatments can be tested with far fewer participants, and thus with less time and cost.
In the next phase of the Neuroimaging and Biomarker Study beginning shortly, we are preparing to process through the successful brain imaging protocol at least 80 Gulf War veterans selected randomly from the National Survey of Gulf War-Era Veterans representing the three syndrome variants and well control veterans. The findings in this sample will examine the previously raised hypotheses about the nature of Gulf War illness in the larger population of Gulf War veterans—a vital step that is required before any of the prior findings can be considered strongly supported.
Selected Scientific Papers Published from the Program
1.Haley RW, Kurt TL, Hom J. Is there a Gulf War syndrome? Searching for syndromes by factor analysis of symptoms. Journal of the American Medical Association 1997;277:215-222.
2.Haley RW, Hom J, Roland PS, Bryan WW, Van Ness PC, Bonte FJ, Devous MD, Mathews D, Fleckenstein JL, Wians FH, Wolfe GI, Kurt TL. Evaluation of neurologic function in Gulf War veterans: a blinded case-control study. Journal of the American Medical Association 1997;277:223-230.
3.Haley RW, Kurt TL. Self-reported exposure to neurotoxic chemical combinations in the Gulf War: a cross-sectional epidemiologic study. Journal of the American Medical Association 1997;277:231-237.
4.Hom J, Haley RW, Kurt TL. Neuropsychological correlates of Gulf War syndrome. Archives of Clinical Neuropsychology 1997;12:531-544.
5.Haley RW. Is Gulf War syndrome due to stress? The evidence reexamined. American Journal of Epidemiology 1997;146:693-703.
6.Haley RW. Point: Bias from the "healthy-warrior effect" and unequal follow-up in three government studies of health effects of the Gulf War. American Journal of Epidemiology 1998; 148:315-323.
7. Haley RW, Billecke S, La Du BN. Association of low PON1 type Q (type A) arylesterase activity with neurologic symptom complexes in Gulf War veterans. Toxicology and Applied Pharmacology 1999; 157:227-233.
8.Roland PS, Haley RW,Yellin W, Owens K, Shoup AG. Vestibular dysfunction in Gulf War syndrome. Otolaryngology–Head and Neck Surgery 2000;122:319-329.
9.Haley RW, Marshall WW, McDonald GG, Daugherty M, Petty F, Fleckenstein JL. Brain abnormalities in Gulf War syndrome: evaluation by 1H magnetic resonance spectroscopy. Radiology 2000;215:807-817.
10.Sinton CM, Fitch TE, Petty F, Haley RW. Stressful manipulations that elevate corticosterone reduce blood-brain barrier permeability to pyridostigmine in the rat. Toxicology and Applied Pharmacology 2000;165:99-105
11.Haley RW, Fleckenstein JL, Marshall WW, McDonald GG, Kramer GL, Petty F. Effect of basal ganglia injury on central dopamine activity in Gulf War syndrome. Archives of Neurology 2000;57:1280-1285.
12.La Du BN, Billecke S, Haley RW, Broomfield CA. Serum paraoxonase (PON1) isozymes: the quantitative analysis of isozymes affecting individual sensitivity to environmental chemicals. Drug Metabolism and Disposition. 2001;29:566-569.
13.Haley RW, Luk GE, Petty F. Use of structural equation modeling to test the construct validity of a case definition of Gulf War syndrome: invariance over developmental and validation samples, service branches and publicity. Psychiatry Research 2001;102:175-200.
14.Cowan J, Sinton CM, Varley AW, Wians FH, Haley RW, Munford RS. Gene therapy to prevent organophosphate intoxication. Toxicology and Applied Pharmacol. 2001;173:1-6.
15.Haley RW, Maddrey AM, Gershenfeld HK. Severely reduced functional status in veterans fitting a case definition of Gulf War syndrome. American Journal of Public Health 2002; 92:46-47.
16.Haley RW. Excess incidence of ALS in young Gulf War veterans. Neurology 2003;61(6): 750-756.
17.Haley RW. Gulf War syndrome: narrowing the possibilities. Lancet Neurol. 2003;2:272-3.
18.Haley RW, Vongpatanasin W, Wolfe GI, Bryan WW, Armitage R, Hoffmann RF, Callahan TS, Charuvastra E, Shell WE, Marshall WW, Victor RG. Blunted circadian variation in autonomic regulation of sinus node function in veterans with Gulf War syndrome. American Journal of Medicine 2004; 117(7): 469-478.
19.Spence JS, Carmack PS, Gunst RF, Schucany WR, Woodward WA, Haley RW. Increasing the power of group comparisons in SPECT brain imaging through spatial modeling of intervoxel correlations. JASA Journal of the American Statistical Association 2007; 478:464-473.
20.Haley RW, Spence JS, Carmack PS, Gunst RF, Schucany WR, Petty F, Devous MD Sr, Bonte FJ, Trivedi MH. Abnormal brain response to cholinergic challenge in chronic encephalopathy from the 1991 Gulf War. Psychiatry Research Neuroimaging 2009; 171: 207-220.