key: cord-010578-uib9h1lb authors: Mawle, Alison C.; Nisenbaum, Rosane; Dobbins, James G.; Gary, Howard E.; Stewart, John A.; Reyes, Michele; Steele, Lea; Schmid, D. Scott; Reeves, William C. title: Seroepidemiology of Chronic Fatigue Syndrome: A Case-Control Study date: 1995-12-17 journal: Clin Infect Dis DOI: 10.1093/clinids/21.6.1386 sha: doc_id: 10578 cord_uid: uib9h1lb We performed serological testing for a large number of infectious agents in 26 patients from Atlanta who had chronic fatigue syndrome (CFS) and in 50 controls matched by age, race, and sex. We did not find any agent associated with CFS. In addition, we did not find elevated levels of antibody to any of a wide range of agents examined. In particular, we did not find elevated titers of antibody to any herpesvirus, nor did we find evidence of enteroviral exposure in this group of patients. We performed serological testing for a large number of infectious agents in 26 patients from Atlanta who had chronic fatigue syndrome (CFS) and in 50 controls matched by age, race, and sex. We did not find any agent associated with CFS. In addition, we did not find elevated levels of antibody to any of a wide range of agents examined. In particular, we did not find elevated titers of antibody to any herpesvirus, nor did we find evidence of enteroviral exposure in this group of patients. Chronic fatigue syndrome (CFS) is an illness of unknown etiology, characterized by debilitating fatigue lasting longer than 6 months and a variety ofnonspecific symptoms, including myalgia, arthralgia, lymphadenopathy, low-grade fever, sleep disorders, and inability to concentrate [1] . An infectious etiology has been suggested for CFS, although the evidence is not compelling [2] . Many patients report the sudden onset of a flulike illness that presaged their fatiguing illness, and a number of infectious agents are known to cause postinfection fatigue. Reports that viral antibody titers are elevated in CFS cases has led to the speculation that latent viruses may be reactivated in this illness as a result of an underlying perturbation of immune function, and that elevated titers of antibody to common agents may be a reflection of this disturbance. We conducted serological tests for a large number of infectious agents as part of a case-control study assessing risk factors for CFS. Our goals were (1) to determine whether we could detect a common exposure history among CFS patients who met a stringent research case definition and (2) to determine whether levels of antibody to any of the agents examined were elevated in CFS cases. Patients with CFS were recruited from the Atlanta component of the Centers for Disease Control and Prevention's (CDC's) surveillance study ofCFS [3] . Cases ofCFS in Atlanta Received who currently met the 1988 CFS research case definition [4] and who had been sick for~10 years were eligible for the study; 26 cases were recruited to participate in the study. We selected two controls for each case. They were matched for age (±5 years), race, and sex. Controls were selected by random-digit dialing in the five-county area of Atlanta covered by the surveillance system. All controls were screened to eliminate those with medical conditions that could bias the results (Reyes et al., unpublished data). We collected blood and stool samples from each participant. Serum and stool specimens were stored at -70°C. Each laboratory was requested to perform serological testing of samples as a single batch to minimize interassay variation. If that was not possible, samples from a case and two matched controls were tested in the same batch. All samples were coded so that the tests were performed in a blinded fashion. Retroviruses. Serological testing for HIV-1 and HIV-2 was performed with use of a U.S. Food and Drug Administration (FDA)-licensed ELISA kit (Genetic Systems, Redmond, WA). Confirmatory testing was performed with an FDA-licensed western blot test (Cambridge Biotech, Rockville, MD). Antibodies against human T-lymphotrophic virus types I and II were detected with an ELISA, and confirmatory testing was performed by western blotting [5] . To detect the presence of a retrovirus in peripheral blood lymphocytes (PBLs) from cases, we cultured 2 X 10 6 PBLs with allogeneic phytohemagglutinin-stimulated PBLs for 4 weeks. Cultures were fed weekly and culture supernatants were screened weekly for reverse transcriptase activity [6] . IgM with use of a monoclonal antibody-capture ELISA similar to that described previously [7] . To detect active infection, we prepared RNA from stool samples and performed a reverse transcriptase PCR to detect enteroviral sequences [8] . For four cases and three controls, materials were insufficient for enteroviral RNA testing. Arboviruses. Sera were screened with a hemagglutinationinhibition assay [9] against a panel of 10 arboviruses. Antibodies against Colorado tick fever virus and vesicular stomatitis virus were detected by means of a CF assay [10] . Titers of 1:10 were considered positive. Herpesviruses. Titers of antibody to cytomegalovirus, human herpesvirus 6, and varicella zoster virus were determined by means of an ELISA against the whole virus. Optical density readings were converted into antibody units with use of known standards [11] . Cytomegalovirus and varicella zoster virus values were considered positive at~100 antibody units, and human herpesvirus 6 values were considered positive at~200 antibody units. Epstein-Barr virus (EBV) early antigen antibody titers were determined with a commercial ELISA (Gull Laboratories, Salt Lake City), and a titer of~I: 10 was considered positive. Antibodies specific for EBV viral capsid antigen (VCA) and nuclear antigen were titrated by indirect immunofluorescence assay (IFA) [12] . A titer of~1:10 was considered positive. Type-specific herpes simplex virus serology was performed with a glycoprotein G-based western blot assay [13] . Respiratory viruses. Sera were tested for the presence of antibodies to adenovirus; parainfluenza viruses I, 2, and 3; and respiratory syncytial virus. Testing was performed by ELISA with a 1:100 serum dilution [14] . Antibodies to coronavirus subtypes OC43 and SN229E were detected with a microneutralization assay [15] . Hepatitis viruses. Sera were screened for antibody to hepatitis B or hepatitis C by ELISA, with use of commercially available kits (Abbott Laboratories, Abbot Park, IL). ELISApositive sera were subjected to confirmatory testing by western blotting (Abbott Laboratories). Other viruses. Antibody to measles nucleoprotein [16] , rubella (Rubestat G, Bio Whittaker, Walkersville, MD), and parvovirus B19 [17] was detected by ELISA. Measles neutralization titers were also determined [18] . Rickettsiaceae. Serological testing for antibodies to Rickettsia typhi, Rickettsia ricketts ii, Coxiella burnetii, and Ehrlichia chaffeensis was performed with use of IFA against fixed whole organisms [19] . A titer of~1:32 was considered positive. Bartonella (formerly Rochalimaea species). Sera were screened for antibodies to Bartonella henselae, B. quintana, and B. elizabethae by means ofIFA with fixed whole organisms [20] . A titer of~1:32 was considered positive. Borrelia burgdorferi. Sera were screened by ELISA for antibodies against flagellin. A positive result was confirmed by western blotting (Robbins et aI., manuscript in preparation). Candida albicans. Precipitating antibodies were detected by both latex agglutination and immunodiffusion. The result was considered positive if either the immunodiffusion assay was positive or the latex agglutination titer was~1:4. We also screened for the circulating Candida antigens mannan A and mannan B [21] . Chlamydia. Sera were screened by IFA at 1:32 dilution against Chlamydia trachomatis, serotype L2. This contains the lipopolysaccharide antigen specific for the genus Chlamydia and thus detects antibodies to all Chlamydia species [22] . We analyzed our data using matched analysis procedures. We used a nonparametric test (Cochran-Mantel-Haenzel statistic [23] with modified ridit scores [24] ) to assess differences in continuous variables between cases and controls. The logarithm in base 2 of the titers of antibody to EBV-VCA, EBVearly antigen, EBV -nuclear antigen, and coronavirus OC43 as well as the measles neutralization titer were used in the analysis, since they have a natural interpretation when titers are considered (each dilution represents 1 unit in the log base 2 scale). In all statistical testing we carried out, a P value of~.05 was considered significant. Since the design of the study was exploratory and not hypothesis-testing, we did not correct for multiple comparisons [25] . Statistical analyses were performed with the statistical software SAS (SAS Institute, Cary, NC). We recruited 26 patients with CFS (23 female and 3 male) and 52 matched controls. Two controls withdrew, leaving 2 cases with only 1 matched control. All participants were white (this finding is consistent with those of other studies). The median duration of illness for the cases was 5.3 years (range, 2.2-10.5 years), and the median age at onset of illness was 33.5 years (range, 16-49 years). Occupation, income, and education were comparable between cases and controls. Agents infrequently detected in or absent from CFS cases. Reverse transcriptase PCR analysis of stool specimens revealed the presence of enteroviral RNA in three cases (13.6%) and four controls (8.5%). There was no overlap between those who had enterovirus-specific IgM and those who had detectable viral RNA in their stool. There were no statistically significant differences between cases and controls for any of these agents. Agents frequently detected in CFS cases. All other agents tested were detected in ;;:::25% of CFS cases, and antibody levels were compared between cases and controls. These agents included the herpesviruses cytomegalovirus, varicella zoster virus, human herpesvirus 6, EBV (early antigen, nuclear antigen, and VCA), and herpes simplex virus 1; the respiratory viruses adenovirus, respiratory syncytial virus, coronavirus, and parainfluenza viruses 1,2, and 3; measles virus; parvovirus B19; and rubella (tables 1 and 2). The seropositivity rates in both cases and controls were ;;:::50%. There were no differences between the two groups in terms of seropositivity rates or virusspecific antibody levels, as measured by titer, by antibody units, or by optical density in an ELISA. We have screened a well-characterized group of patients with CFS for evidence of infection with a wide range of agents and have compared percentages of infection with age-, race-, and sex-matched controls. We found no evidence of a common infectious agent in the patients studied. There were no differ-ences in prevalence of current enteroviral infection between cases and controls, as determined by levels of circulating enterovirus-specific IgM, nor in the frequency of detectable enterovirus in stool samples. Thus, in the group of patients from Atlanta, enteroviral infection was not more common in CFS cases. We were unable to detect evidence of infection with any known human retrovirus, and our inability to detect reverse transcriptase activity in cocultures suggests that there is no common, unknown retrovirus in these patients. This is consistent with other studies that have been unable to detect retroviruses in CFS [26, 27] . We found no evidence of infection withB. burgdorferi, the causative agent ofLyme disease. Infection with B. burgdorferi can lead to a CFS-like illness, and in areas where Lyme disease is endemic, it can contribute to CFSrelated estimates [28] . In this group of patients from Atlanta, there was no evidence that Lyme disease is a factor in CFS. Other agents that had <25% seroprevalence in these patients included arboviruses, hepatitis B and C, Rickettsiaceae, Bartonella species, Chlamydia species, and precipitating antibodies to C. albicans. We also looked for circulating Candida antigens, since these are frequently found in severely immunodeficient patients. None of our patients with CFS had detectable mannan A or mannan B, which suggests that these patients were not immunodeficient. The lack of precipitating antibodies to C. albicans argues against the occurrence of a chronic candidal infection in patients with CFS. Herpesviruses-in particular, EBV and human herpesvirus 6-have been associated with CFS. Although elevated titers to EBV-VCA and EBV early antigen were initially described as occurring in patients with CFS [29, 30] , subsequent studies made it clear that these were not specific to CFS [31] [32] [33] . In this study, we found no differences in titers to EBV-VCA or EBV nuclear antigen between cases and controls. However, the titer of early antigen was slightly elevated in the cases compared with that in controls, although the difference was not statistically significant. In conclusion, we were unable to find evidence of a single infectious agent associated with CFS in this patient population, nor did we find evidence of elevated antibody titers consistent either with reactivation of a latent virus or with generalized immune activation. The chronic fatigue syndrome: a comprehensive approach to its definition and study Is chronic fatigue syndrome an infectious disease? Epidemiology of chronic fatigue syndrome: the Centers for Disease Control Study Chronic fatigue syndrome: a working case definition Simultaneous confirmation and differentiation of human T-lymphotropic virus types I and II infection by modified Western blot containing recombinant envelope glycoproteins Isolation, characterization, and transmission of human T-lymphotrophic virus types I and II in culture Echovirus 30 infection and aseptic meningitis in parents of children attending a child care center Enteroviral RNA sequences detected by polymerase chain reaction in muscle of patients with postviral fatigue syndrome Techniques for hemagglutination and hemagglutination-inhibition with arthropod-borne viruses Adaptation ofLECF microtechnique in standardized diagnostic complement fixation method and adaptation to micro test Prior herpes simplex type 2 infection as a risk factor for HIV infection Immunofluorescence in cells derived from Burkitt's lymphoma Evaluation of a test based on baculovirus-expressed glycoprotein G for detection of herpes simplex virus type-specific antibodies Viral antibody screening system that uses a standardized single dilution immunoglobulin G enzyme immunoassay with multiple antigens Antigen detection in human respiratory coronavirus infections by monoclonal time resolved fluoroimmunoassay Baculovirus expression of the nucleoprotein gene of measles virus and utility of the recombinant protein in diagnostic enzyme immunoassays Human parvovirus BI9-specific IgG, IgA, and IgM antibodies and DNA in serum specimens from patients with erythema infectiosum Evaluation of monoclonal antibody-based capture enzyme immunoassays for detection of specific antibodies to measles virus A comparison of the complement fixation, indirect immunofluorescent antibody, and microagglutination tests for the serological diagnosis of rickettsial diseases Serological response to Rochalimaea henselae antigen in suspected cat-scratch disease Serodiagnosis of fungal diseases Single antigen fluorescence test for Chlamydia antibodies Average partial association in threeway contingency tables: a review and discussion of alternative tests On the combination of independent two-sample tests of Wilcoxon No adjustments are needed for multiple comparisons Search for retrovirus in the chronic fatigue syndrome Lack of evidence for infection with known human and animal retroviruses in patients with chronic fatigue syndrome Jennekens FGL Chronic muscle weakness caused by Borrelia burgdorferi meningoradiculitis Evidence for active Epstein-Barr virus infection in patients with persistent unexplained illness: elevated anti-early antigen antibodies Persisting illness and fatigue in adults with evidence of Epstein-Barr virus infection Frequency of chronic active Epstein-Barr virus infection A cluster of patients with a chronic mononucleosis-like syndrome: is Epstein-Barr virus the cause? Antibodies to Epstein-Barr virus in patients with chronic fatigue The authors gratefully acknowledge Drs. L. Schonberger, W. Gunn, A. Khan, and G. Holmes (CDC), whose initial work on CFS led to this study. They are indebted to Drs. W. Bellini, C. Black, R. Craven, D. Erdman, T. Folks, 1. Hierholzer, N. Karabatsos, R. Lal, S. Lambert, M. Pallansch, R. Regnery, E. Reiss, and T. Tzianabos (CDC), in whose laboratories many of the reported assays were performed, and to Dr. 1. Gow (Institute of Neurological Science, Scotland) for performing the enteroviral RTPCR. They also acknowledge D. Connell, B. Randall, and R. Jacobs (Abt Associates) as well as E. Parris (CDC), who were responsible for study logistics and data processing.