Europace Advance Access originally published online on August 18, 2006
Europace 2006 8(10):846-851; doi:10.1093/europace/eul092
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
SYNCOPE
Familial aggregation of fainting in a case–control study of neurally mediated hypotension patients who present with unexplained chronic fatigue
1 Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, 615 N. Wolfe Street, E6148, Baltimore, MD 21205, USA; 2 Division of Epidemiology, Department of Health Sciences Research, Mayo Clinic College of Medicine, 200 First Street, SW, Rochester, MN 55905, USA; 3 Department of Pediatrics, Johns Hopkins School of Medicine, Brady 212, 600 N. Wolfe Street, Baltimore, MD 21287, USA
Manuscript submitted 22 February 2006. Accepted after revision 11 May 2006.
* Corresponding author. Tel: +1 410 955 8720; fax: +1 410 614 8883. E-mail address: kalucas{at}jhmi.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Aims We hypothesized that family history of fainting is a risk factor for adult-onset neurally mediated hypotension (NMH) in patients who present with chronic fatigue rather than fainting.
Methods and results A nested case–control study of Gulf War veterans gathered fainting histories directly from 197 first-degree relatives of 16 fatigued NMH cases, 26 fatigued controls, and 17 healthy controls. NMH status was determined by tilt testing. Female relatives of fatigued controls reported more fainting than female relatives of cases (P<0.01); there were no significant differences in the frequency of fainting in male relatives by case or control status, or in either gender by age at first faint or by familial relationship to the proband. The odds of NMH patients giving any family history of fainting were 0.56 (95% CI 0.15, 2.07). Recurrence risks calculated using lifetime prevalences of fainting in male military personnel of similar age to our participants were low (1.5–1.7) and did not differ by case or control status.
Conclusion Family history of fainting is not a risk factor for adult-onset NMH in fatigued veterans. Our findings may differ from other studies of familial aggregation in NMH because of study methods or because NMH-fatiguers may differ from NMH-fainters.
Key Words: Fatigue, Heredity, Gulf War syndrome, Vasovagal syncope
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Having a family history of fainting has been described as a risk factor for vasovagal syncope in uncontrolled studies or case reports of adults with recurrent fainting.1
–5 Patients with an increased tendency to vasovagal or neurally mediated hypotension (NMH) during tilt testing can also present with chronic fatigue (NMH-fatiguers), but may not experience recurrent fainting.6–12 This was noted by Rowe et al. in a study of adolescents with unexplained fatigue but no histories of syncope. Later studies of patients meeting criteria for chronic fatigue syndrome (CFS) have found the prevalence of NMH in stage 1 of a tilt test to be 14–93%,6–11,13–18 and yet the majority of CFS patients do not report a history of syncope.6 Because CFS and Gulf War illnesses have many symptoms in common,19–21 and because of the high prevalence of NMH in some studies of CFS, the primary aim of the parent study was to investigate the prevalence of NMH and postural tachycardia syndrome (POTS) in chronically fatigued Gulf War veterans.22 A secondary aim of the study was to evaluate whether having a family history of fainting was a risk factor for NMH in those veterans who were NMH-fatiguers. Although studies suggest that familial aggregation occurs in persons with NMH who present with recurrent fainting (NMH-fainters), no one has ever investigated whether a positive family history of fainting is a risk factor for those who have NMH provoked during tilt but who present clinically with fatigue rather than with recurrent syncope. Knowing if familial aggregation exists could lead to more specific genetic studies, which, theoretically, could enable screening for susceptibility to idiopathic chronic fatigue. The presence of a family history of fainting in patients with unexplained fatigue might also suggest to physicians a need for a tilt table test, and may lead to more rapid identification of NMH in these patients. Therefore, we posed the research question ‘Is familial aggregation of fainting, a risk factor for NMH in those with NMH who present with chronic fatigue rather than with fainting?'. We conducted a nested case–control study of NMH cases and non-NMH controls to answer this question, and hypothesized that NMH cases would have a significantly greater prevalence of positive family histories than non-NMH controls. |
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Fatigued veterans were defined as US military personnel who were on the ground in the Persian Gulf at some time between August 2, 1990 and July 31, 1991, and who reported persistent, unexplained fatigue that began during or after the Gulf War, but not later than July 31, 1992, plus two other unexplained symptoms including cognitive difficulties, headaches, joint aches without swelling, muscle aches, unrefreshing sleep, or post-exertional fatigue. Cases were defined as fatigued veterans with NMH, and were compared with fatigued and healthy veterans without NMH. The symptom fatigue must be chronic in nature, and not just occurring during tilt testing for NMH. Seven non-fatigued veterans who tested positive for NMH were used for a sub-study to determine if our findings changed with the fatigue status of cases, but these seven cases were not part of the main analysis. Persons who were ineligible were those 51 or older, those who could not walk without assistance, who had an illness which could be worsened by tilt testing, who took medications which could interfere with the tilt test, who had a recent history of drug or alcohol abuse, or who were lactating or pregnant. Veterans were recruited nation-wide using a variety of means. Details about recruitment, questionnaires, examinations, and other study methods are given elsewhere.22
We have used the term NMH to refer to the changes in blood pressure and heart rate that are otherwise known as vasovagal syncope, neurocardiogenic syncope, vasodepressor syncope, or neurally mediated syncope. The term NMH emphasizes the potential for affected patients to present with light-headedness and fatigue, often on a daily basis, without necessarily having experienced syncope.6,12
All participants underwent tilt table testing, with NMH defined as a fall from baseline blood pressure 
25 mmHg in systolic blood pressure without an increase in heart rate and accompanied by syncope or presyncope.6 Although we also noted whether POTS occurred during the tilt test (defined as an increase in heart rate of 30 or more beats per minute over the baseline rate, and occurring within 10 min of upright tilt),23,24 POTS status was not included in our analyses. Baseline blood pressure and heart rate were measured while the participant was supine on the tilt table for 15 min prior to tilting upright. The average of the 10 and 15 min readings defined baseline values.
A 70° angle was used throughout the upright tilt test. Blood pressure was measured at 1 min and at 5-min intervals thereafter. Heart rate was recorded at the same time, except in the first 10 min when it was recorded every minute. If the participant began to exhibit signs of presyncope, blood pressure was measured each minute. If the participant had not experienced NMH within 45 min, the table was returned to the horizontal position, 1–3 µg/min of isoprenaline were infused to increase baseline heart rate by 20%, and the participant was tilted for up to 15 min longer.
Tilt testing occurred at 11 a.m., after a 4-h fast. The temperature in the tilt laboratory was 22.2°C, and visual and auditory stimulation were kept to a minimum. Participants were asked to remain motionless throughout the test.
The study application requested contact information for all first-degree relatives 18 or older, with ‘first-degree relative’ defined to avoid inclusion of half- or step-relatives. A trained interviewer telephoned family members and asked about lifetime history of fainting and age at first faint. Fainting was defined as ‘a temporary loss of consciousness associated with pallor, limpness, symptoms such as nausea, light-headedness, or sweating, and a prompt recovery of consciousness’. To avoid confusion with medical events that might be mistaken for a faint, relatives whose first faint occurred after diagnosis of diabetes, heart disease, stroke, or epilepsy were excluded from the analysis.
Data were analysed using Stata 8.0 (Stata Corporation, College Station, TX, USA). Family history data were used to stratify cases and controls, and odds ratios (OR) were estimated using one relative per kindred to investigate for any association between family history and NMH. A positive family history was recorded if any of the proband's family members had a fainting history. Generalized estimating equations allowed determination of an overall OR for the association between having a positive family history of fainting and having NMH, using all the family history data.25 All P-values were two-sided.
Recurrence risk (
) is the probability that a type of relative (i.e. sibling) of an affected individual manifests a disease, compared with the general population.26 Recurrence risk of syncope in siblings was calculated by dividing the prevalence of syncope in siblings by the population prevalence of syncope.26 The Framingham study reported a 3.0–3.5% syncope prevalence,27 which has been challenged as being too low.28 Therefore, we also used prevalence data from male military recruits of similar age as our comparison population to calculate recurrence risk.29
The study was reviewed and approved by the Johns Hopkins Joint Committee on Clinical Investigation institutional review board and by the Army Surgeon General's Human Subjects Research Review Board. All participants gave written informed consent.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
We recruited 49 chronically fatigued Gulf War veterans and 44 non-fatigued, healthy Gulf War veterans. As shown in Table 1, there were 20 veterans with NMH and 29 without NMH, which represent 20 NMH cases and 29 non-NMH fatigued controls. The Vasovagal Syncope International Study Investigators (VASIS) classification of tilt outcome is included in the table.30
,31 We were able to contact 135 relatives from 43 of these 49 veterans' families. One family member who did not know if she had ever fainted was excluded, leaving a family sample of 134 persons representing 42 fatigued veterans' families. Of the 44 non-fatigued Gulf War veterans, 27 did not have NMH; their families formed the second control group. We contacted 64 relatives from 17 non-fatigued veterans' families, but were unable to collect data on one relative. Because our research question focuses on those with NMH who present with chronic fatigue, we did not use data on 18 relatives from seven families of 44 non-fatigued Gulf War veterans for the main analysis. However, we used these data for a sub-study to see if the results differed based on fatigue status of NMH cases. Figure 1 details the derivation of the sample.
|
|
Mean family size for cases, 3.3 persons; for fatigued controls, 3.1 persons; and for non-fatigued controls, 3.7 persons. A prior history of fainting was reported in 7/16 NMH cases with family data, 7/26 fatigued non-NMH controls with family data, and 2/17 non-fatigued non-NMH controls with family data. The average age at first faint for these 16 participants was 21.1 years old, whereas the average age of these participants at the time of the study was 40.1 years old. Only one participant reported fainting as a current symptom, and that person tested negative for NMH.
As shown in Table 2, the frequency of any family history of fainting was either the same or greater for controls than for cases; the difference reached significance in women when comparing cases with fatigued controls (P<0.01). There were no differences in age at first faint or in family relationship to the proband (data not shown).
|
The odds of having any family history of fainting were 0.56 (95%CI 0.15, 2.07) when cases were compared with all controls (Table 3). Though not statistically significant, the data suggest that having a family history of fainting was protective rather than a risk factor for NMH. The two control groups were pooled for generalized estimating equations analysis, which considered all the family data and accounted for family size. Again, the results were not significant, with the OR of having a family history of fainting of 0.55 (95%CI 0.22, 1.42; n=197).
|
Data were too sparse to calculate recurrence risks for offspring. Sibling recurrence risks (
s) did not differ by case/control status; s ranged from 1.5 to 1.7, when the prevalences in male military personnel of similar age to our participants were used (Table 4).
|
Data from the seven non-fatigued veterans with NMH and family history data were used for a sub-study to see how they compared with the non-NMH controls. Their average family size was 2.6 persons, and 2/7 (28.6%) had a history of fainting. Average age at first faint was 13.7 years, but did not differ significantly from the non-NMH controls. Three of the seven had a family member with a history of fainting, but the prevalence of family history was not significantly different when compared with the non-NMH controls, nor did recurrence risks differ significantly (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
This is the first study to investigate familial aggregation in adults with NMH who present with fatigue rather than fainting, and the first study of familial aggregation in adults with NMH to use controls. Unexpectedly, we did not find family history of fainting to be a risk factor for NMH in chronically fatigued Gulf War veterans. Moreover, the range of recurrence risks did not suggest a strong hereditary influence on NMH in this group.
Our use of family history of fainting as a cost-effective proxy for NMH is in keeping with work published by others. Among investigations of familial aggregation in adults with NMH who present with recurrent fainting, Mathias et al. examined the familial tendency to faint among relatives of 132 consecutive adult and paediatric patients with vasovagal syncope referred to their specialty clinic. Overall, 44/119 (37%) of patients had a positive family history, but only those whose first faint occurred before age 20 were most likely to have a family history.2 This finding is in agreement with studies of familial aggregation in NMH in children1,5,32,33 and raises the possibility of impact upon age at onset. In addition, Mathias et al. investigated family histories of 641 adults and children referred to their tertiary autonomic centre for recurrent syncope and presyncope between 1992 and 1998. Vasovagal syncope was diagnosed in 227/641 (35.4%), and a positive family history was found in 81/227 (35.7%); patients ranged from 11 to 83 years old.3 Newton et al. found that 84/441 (19.0%) of those with vasovagal syncope had a positive family history of fainting. On the basis of sibling and offspring recurrence risks, which exceeded 1000, this group concluded that vasovagal syncope has a strong genetic component.4 However, their risk calculation used the prevalence of vasovagal syncope requiring hospitalization (0.03%) rather than a general population prevalence (conservatively estimated in the Framingham study as 3.0–3.5%),27 which is likely to have overstated the risk a 100-fold. Participant ages were not specified, so it is unclear if the study included children.
Our study had several methodological strengths, which may account for some of the differences between our findings and those reported by others.2–4 First, all cases and controls were ascertained based upon two-stage tilt table testing using a protocol consistent with that of the American College of Cardiology.34
Secondly, to increase the accuracy of reports of an episodic event like syncope and to limit recall bias,35,36 we collected fainting data from family members directly rather than relying on reports from cases or controls. Using surrogates to provide data can also lead to misclassification, as reported in a study of renal cancer where surrogates ‘were notably less reliable than self-respondents in identifying a positive family history of cancer’.37
A third and major strength of our study was our use of controls. The frequency of a family fainting history in our cases was similar to that reported by others,2,3 but was greater in controls than in cases (44% of NMH cases having family histories versus 65% of fatigued controls and 47% of healthy controls). The presence of controls led us to conclude that family history of fainting was not a risk factor for adult-onset NMH, which presents with fatigue.
If different mechanisms are at play in NMH-fainters than in NMH-fatiguers, then family fainting histories may not be an appropriate proxy for NMH in those presenting with fatigue. It could also be that adult-onset NMH in recurrent fainters has a stronger genetic component, whereas adult-onset NMH in fatiguers has a stronger environmental component. There are questions about whether environmental factors to which Gulf War veterans were exposed are associated with ill health.38–41 An environmental factor associated with the expression of NMH in this group could overwhelm any influence of genetic factors in tilt-induced hypotension.
The association between NMH and having a family history of fainting may be underestimated because of incomplete penetrance of any involved putative gene or genes. In addition, if an environmental or other trigger is required, the family member may not have had an opportunity for such an exposure. In either case, genetically susceptible persons may not have fainted, thus underestimating the ORs for the association between family history of fainting and NMH. Finally, because we limited entry to symptomatic veterans whose fatigue began within a year of the end of the conflict period, these findings may not be generalizable to all fatigued Gulf War veterans with NMH.
In conclusion, this controlled study suggests that fatigued veterans with NMH are not more likely to have a family history of fainting than those without NMH.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
This study was conducted at the Johns Hopkins Hospital, 600 N. Wolfe Street, Baltimore, MD 21287, USA. We are grateful to Shirlene Waller for interviewing family members and to Scott Seggerman of the Defense Manpower Data Center for his assistance in recruiting participants. This work was supported by grant DAMD17-00-2-0072 from the United States Department of Defense. The Department of Defense did not participate in the design or conduct of the study; in the collection, analysis, or interpretation of the data; or in the preparation, review, or approval of the manuscript.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
[1] Marquez M, Urias K, Hermosillo A, Jardon J, Iturralde P, Colin L, et al. Familial vasovagal syncope. Europace 2005; 7: 472–4.
[2] Mathias C, Deguchi K, Bleasdale-Barr K, Kimber J. Frequency of family history in vasovagal syncope. Lancet 1998; 352: 33–4.[ISI][Medline]
[3] Mathias C, Deguchi K, Schatz I. Observations on recurrent syncope and presyncope in 641 patients. Lancet 2001; 357: 348–53.[CrossRef][ISI][Medline]
[4] Newton J, Kenny R, Lawson J, Frearson R, Donaldson P. Prevalence of family history in vasovagal syncope and haemodynamic response to head up tilt in first degree relatives. Clin Auton Res 2003; 13: 22–6.[CrossRef][ISI][Medline]
[5] Newton J, Kerr S, Pairman J, McLaren A, Norton M, Kenny R, et al. Familial neurocardiogenic (vasovagal) syncope. Am J Med Genet 2005; 133A: 176–9.
[6] Bou-Holaigah I, Rowe P, Kan J, Calkins H. The relationship between neurally mediated hypotension and the chronic fatigue syndrome. JAMA 1995; 274: 961–7.[Abstract]
[7] DeLorenzo F, Hargreaves J, Kakkar V. Pathogenesis and management of delayed orthostatic intolerance in patients with chronic fatigue syndrome. Clin Auton Res 1997; 7: 185–90.[CrossRef][ISI][Medline]
[8] Schondorf R, Benoit J, Wein T, Phaneuf D. Orthostatic intolerance in the chronic fatigue syndrome. J Auton Nerv Syst 1999; 75: 192–201.[CrossRef][ISI][Medline]
[9] Stewart J, Gewitz M, Weldon A, Arlievsky N, Li K, Munoz J. Orthostatic intolerance in adolescent chronic fatigue syndrome. Pediatrics 1999; 103: 116–21.
[10] Streeten D, Thomas D, Bell D. The roles of orthostatic hypotension, orthostatic tachycardia, and subnormal erythrocyte volume in the pathogenesis of chronic fatigue syndrome. Am J Med Sci 2000; 320: 1–8.[ISI][Medline]
[11] Tanaka H, Matsushima R, Tamai H, Kajimoto Y. Impaired postural cerebral hemodynamics in young patients with chronic fatigue with and without orthostatic intolerance. J Pediatr 2002; 140: 412–7.[CrossRef][ISI][Medline]
[12] Rowe P, Bou-Holaigah I, Kan J, Calkins H. Is neurally mediated hypotension an unrecognised cause of chronic fatigue? Lancet 1995; 345: 623–4.[CrossRef][ISI][Medline]
[13] Freeman R and Komaroff A. Does the chronic fatigue syndrome involve the autonomic nervous system? Am J Med 1997; 102: 357–64.[CrossRef][ISI][Medline]
[14] LaManca J, Peckerman A, Walker J, Kesil W, Cook S, Taylor A, et al. Cardiovascular response during head-up tilt in chronic fatigue syndrome. Clin Physiol 1999; 19: 111–20.[CrossRef][ISI][Medline]
[15] Poole J, Herrell R, Ashton S, Goldberg J, Buchwald D. Results of isoproterenol tilt table testing in monozygotic twins discordant for chronic fatigue syndrome. Arch Intern Med 2000; 160: 3461–8.
[16] Rowe P, Calkins H, DeBusk K, McKenzie R, Anand R, Sharma G, et al. Fludrocortisone acetate to treat neurally mediated hypotension in chronic fatigue syndrome: a randomized controlled trial. JAMA 2001; 285: 52–9.
[17] Streeten D and Anderson G Jr. The role of delayed orthostatic hypotension in the pathogenesis of chronic fatigue. Clin Auton Res 1998; 8: 119–24.[CrossRef][ISI][Medline]
[18] Timmers H, Weiling W, Soetekouw P, Bleijenberg G, Van Der Meer J, Lenders J. Hemodynamic and neurohumoral responses to head-up tilt in patients with chronic fatigue syndrome. Clin Auton Res 2002; 12: 273–80.[CrossRef][ISI][Medline]
[19] Kang HK, Natelson BH, Mahan CM, Lee KY, Murphy FM. Post-traumatic stress disorder and chronic fatigue syndrome-like illness among Gulf War veterans: a population-based survey of 30 000 veterans. Am J Epidemiol 2003; 157: 141–8.
[20] McCauley LA, Joos SK, Barkhuizen A, Shuell T, Tyree W, Bourdette DN. Chronic fatigue in a population-based study of Gulf War veterans. Arch Environ Health 2002; 57: 340–8.[ISI][Medline]
[21] Steele L. Prevalence and patterns of Gulf War illness in Kansas veterans: association of symptoms with characteristics of person, place, and time of military service. Am J Epidemiol 2000; 152: 992–1002.
[22] Lucas K, Armenian H, DeBusk K, Calkins H, Rowe P. Characterizing Gulf War illnesses: neurally mediated hypotension and postural tachycardia syndrome. Am J Med 2005; 118: 1421–7.[CrossRef][ISI][Medline]
[23] Grubb B, Kosinski D, Boehm K, Kip K. Postural orthostatic tachycardia syndrome: a neurocardiogenic variant identified during head-up tilt table testing. Pacing Clin Electrophysiol 1997; 20: 2205–12.[CrossRef][Medline]
[24] Sandroni P, Opfer-Gehrking T, Benarroch E, Shen W-K, Low P. Certain cardiovascular indices predict syncope in the postural tachycardia syndrome. Clin Auton Res 1996; 6: 225–31.[CrossRef][ISI][Medline]
[25] Liang K. Extended Mantel–Haenszel estimating procedure for multivariate logistic regression models. Biometrics 1987; 43: 289–99.[CrossRef][ISI][Medline]
[26] Farrer L and Cupples L. Determining the genetic component of a disease. In Haines J and Pericak-Vance M (Eds.). Approaches to Gene Mapping in Complex Diseases 1998; New York Wiley-Liss pp. p93–101.
[27] Savage D, Corwin L, McGee D, Kannel W, Wolf P. Epidemiologic features of isolated syncope: the Framingham Study. Stroke 2004; 16: 626–9.
[28] Colman N, Nahm K, Ganzeboom K, Shen W, Reitsma J, Linzer M, et al. Epidemiology of reflex syncope. Clin Auton Res 2004; 14:(Suppl. 1), 9–17.[CrossRef][ISI][Medline]
[29] Lamb L, Green H, Combs J, Cheeseman S, Hammond J. Incidence of loss of consciousness in 1980 Air Force personnel. Aerospace Med 1960; 31: 973–88.
[30] Benditt D, Petersen M, Lurie K, Grubb B, Sutton R. Cardiac pacing for prevention of recurrent vasovagal syncope. Ann Intern Med 2006; 122: 204–9.
[31] Sutton R, Petersen M, Brignole M, Raviele A, Menozzi C, Giani P. Proposed classification for tilt induced vasovagal syncope. Eur J Cardiac Pacing Electrophysiol 1992; 2: 180–3.
[32] Camfield P and Camfield C. Syncope in childhood: a case–control clinical study of the familial tendency to faint. Can J Neurol Sci 1990; 17: 306–8.[ISI][Medline]
[33] Mathias C, Deguchi K, Bleasdale-Barr K, Smith S. Familial vasovagal syncope and pseudosyncope: observations in a case with both natural and adopted siblings. Clin Auton Res 2000; 10: 43–5.[CrossRef][ISI][Medline]
[34] Benditt D, Ferguson D, Grubb B, Kapoor W, Kugler J, Lerman B, et al. Tilt table testing for assessing syncope. J Am Coll Cardiol 1996; 28: 263–75.[CrossRef][ISI][Medline]
[35] Aitken J, Green A, MacLennan R, Jackman L, Martin N. Comparability of surrogate and self-reported information on melanoma risk factors. Br J Cancer 1993; 67: 1036–41.[ISI][Medline]
[36] Weiss K, Chakraborty R, Majumder P. Problems in the assessment of relative risk of chronic disease among biological relatives of affected individuals. J Chronic Dis 1982; 35: 539–51.[CrossRef][ISI][Medline]
[37] Lipworth L, Fryzek J, Fored C, Blot W, McLaughlin J. Comparison of surrogate with self-respondents regarding medical history and prior medication use. Int J Epidemiol 2001; 30: 303–8.
[38] Institute of Medicine. Health consequences of service during the Persian Gulf War: initial findings and recommendations for immediate action. Washington, DC National Academy Press.
[39] NIH Technology Assessment Workshop. The Persian Gulf experience and health. JAMA 1994; 272: 238–45.[CrossRef][ISI][Medline]
[40] Persian Gulf Veterans Coordinating. Unexplained illnesses among Desert Storm veterans: a search for causes, treatment, and cooperation. Arch Intern Med 1995; 155: 262–8.[Abstract]
[41] Presidential Advisory Committee on Gulf War Veterans' Illnesses. Final report. Washington, DC U.S. Government Printing Office.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||