sign in sign up
<<
Home , Chronic fatigue syndrome, Meningitis, Meningoencephalitis

50450ournal ofNeurology, , and Psychiatry 1996;60:504-509

Chronic and minor psychiatric morbidity J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 1996. Downloaded from after viral : a controlled study

Matthew Hotopf, Norman Noah,

Abstract is usually considered to be a Objective-To test the hypotheses that benign illness. It is most commonly caused patients exposed to viral meningitis would now by , as has be at an increased risk of developing decreased in .' Enteroviral and would have has been suggested as a cause of fatigue and an excess of neurological symptoms and chronic fatigue syndrome, sometimes called physical impairment. postviral fatigue syndrome or myalgic Methods-Eighty three patients were fol- (ME). A range of serological lowed up 6-24 months after viral meningi- and viral markers of infection in patients has tis and a postal questionnaire was used to been examined in case-control studies. Early compare outcome with 76 controls who results suggested a relation between these had had non-enteroviral, non-CNS viral markers of infection and current symptoms.2 7 . These studies have been criticised8 for their Results-For the 159 patients and con- use of inappropriate control groups, and selec- trols the prevalence of chronic fatigue tion and ascertainment biases. More recent syndrome was 12-6%, a rate higher than studies, which have overcome these method- previously reported from ological problems, suggest that the relation is attenders, suggesting that moderate to less compelling than previously thought,9 '3 severe viral infections may play a part in although one well conducted study found evi- the aetiology of some fatigue states. Those dence for a relation.'4 with a history of meningitis showed a One problem is the difficulty of defining slight, non-significant increase in preva- exposure given the frequency of enteroviral lence of chronic fatigue syndrome (OR infections in the community. We hypothesised 1-4; 95% CI 0.5-3.6) which disappeared that if enteroviral infection is a true risk factor when logistic regression analysis was used for chronic fatigue syndrome, then this effect to correct for age, sex, and duration of will be strongest in a sample of patients follow up (OR 1-0; 95% CI 0 3-2 8). exposed to enteroviral infection of the CNS. Controls showed marginally higher psy- Patients with viral meningitis represent such a chiatric morbidity measured on the gen- sample: in the enteroviruses

eral health questionnaire-12 (adjusted OR are the commonest cause of viral meningitis.'5 http://jnnp.bmj.com/ 0-6; 95% CI 0-3-1.3). Both groups had If enteroviruses cause chronic fatigue syn- similar rates of neurological symptoms drome we would expect patients recovering and physical impairment. The best pre- from viral meningitis to be at increased risk of dictor of chronic fatigue was a prolonged abnormal fatigue. Viral meningitis is also said duration of time off work after the illness to be benign and without neurological seque- (OR 4-93, 95% CI 1-3-18-8). The best pre- lae and this was also examined in this study. dictor of severe chronic fatigue syndrome There are two large studies that followed up diagnosed by Center for Control patients with viral meningitis. Lepow et all6 in on October 2, 2021 by guest. Protected copyright. criteria was past psychiatric illness (OR an uncontrolled study found that two thirds of 7-82, 95% CI 1-8-34.3). Duration of viral patients were at risk of fatigue, , and illness, as defined by days in hospital, did clumsiness three months after the illness, but not predict chronic fatigue syndrome. the impression was of recovery at one year, The of follow was Muller Institute of Psychiatry, Conclusions-(l) prevalence although up incomplete. et Denmark Hill, chronic fatigue syndrome is higher than al'7 traced 238 patients with London, UK expected for the range of viral illnesses and followed them up for two to 12 years. M Hotopf examined; (2) enteroviral infection is They could detect no difference between their S Wessely unlikely to be a specific risk factor for its and that of healthy controls. King's College School of Medicine and development; (3) onset of chronic fatigue These two studies did not aim to examine Dentistry, London, UK syndrome after a viral infection is pre- patients specifically for fatigue; nor did they N Noah dicted by psychiatric morbidity and pro- use standardised assessments of symptoms. S Wessely longed convalescence, rather than by the The aims of this study were: (a) to examine Correspondence to: Dr M Hotopf, Department severity of the viral illness itself. whether patients with viral meningitis are at of Psychological Medicine, greater risk of subsequent chronic fatigue syn- Institute of Psychiatry, 103 Denmark Hill, London, SE5 (J Neurol Neurosurg Psychiatry 1 996;60:504-509) drome, psychiatric morbidity, and neurologi- 8AZ, UK. cal complications than a control group of Received 31 January 1995 patients who had non-enteroviral, non-CNS, and in final revised from 1 September 1995 Keywords: chronic fatigue syndrome; viral meningitis; viral illnesses; (b) to attempt to find predictors Accepted 4 September 1995 psychiatric morbidity of fatigue and psychological symptoms. Chronic fatigue and minor psychiatric morbidity after viral meningitis: a controlled study 505

Methods ral infections, and J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 May 1996. Downloaded from CASE SELECTION (which has been shown to be a risk factor for Cases were identified from four virology labo- fatigue'8) were also excluded. Controls with ratories and from the records of a hospital spe- any evidence of "" (nuchal rigidity cialising in infectious . These and ) or a history of lumbar punc- represented a mix of urban and rural popula- ture during the illness were excluded. tions. In the virology laboratories records of samples of CSF in the routine day books over CASE NOTE INFORMATION the previous two years (July 1 99 1-June 1993) For all subjects, case notes or hospital dis- were used to identify patients, whose hospital charge summaries were examined to gather case notes were then obtained and examined. baseline sociodemographic information and For cases at the infectious diseases hospital the details regarding current and past physical and disease cards (which document by disease psychiatric illness. Information on the nature each admission to the hospital) were used to of the illness leading to admission to hospital identify cases of viral meningitis and the rele- was also collected on a checklist of symptoms vant case notes were then checked. and signs. Investigations performed in hospital In all cases the following inclusion criteria were recorded as well as any information on were met: age 16-65; a clinical diagnosis of the clinical state of the patient at discharge. acute onset meningitis (characterised by pyrexia, , photophobia, and neck POSTAL SURVEY stiffness) leading to admission to hospital; A questionnaire was designed to elicit sociode- spontaneous resolution of symptoms; evidence mographic details and information regarding of lymphocytic meningitis as defined by lym- medical and psychiatric history and service phocytes in the CSF (> 8 per cm3). Cases use. Current psychiatric health was measured were excluded if there was any evidence of a using the general health questionnaire- 12 bacterial or other non-viral cause of the (GHQ-12)19 and the Beck inven- meningitis from microscopy and culture of the tory.20 Fatigue was measured on a validated CSF. Further exclusion criteria included seri- questionnaire used in several studies of ous medical illnesses, which would be a poten- chronic fatigue and chronic fatigue syn- tial cause of fatigue, and patients with a history drome.21 Somatic symptoms were identified of psychosis, substance misuse, or eating dis- on a checklist of 32 symptoms. Finally, func- order. Other psychiatric disorders were not tional impairment and were excluded. Cases with illness that had appeared assessed on the medical outcome survey short in the context of recent travel outside Europe form (SF-36) scored according to predeter- and North America were also excluded. mined thresholds.22 Table 1 gives the defini- Finally, cases were excluded if there was a tions of the different fatigue syndromes. clinical suspicion or definite confirmation of Chronic fatigue syndrome is defined according HIV infection. to the Green College criteria23 or to the Center for Disease Control (CDC) criteria (chronic CONTROL SELECTION fatigue syndrome +).24 Controls were identified from one virology

laboratory and the infectious diseases hospital. http://jnnp.bmj.com/ Those identified from the laboratory were Results proved cases of self limiting viral infections Of 255 eligible patients, 159 (62%) which the laboratory had identified from sam- responded. Of the remainder, 49 had changed ples other than chronic fatigue syndrome. address and could not be traced. If these were From this information hospital casenotes or removed from the sample the response rate general practice records were examined. was 77%. The remaining non-responders Controls were identified from the infectious either failed to return a questionnaire despite diseases hospital using the disease cards and reminders or retumed the questionnaire on October 2, 2021 by guest. Protected copyright. casenotes. Inclusion criteria were of self limit- uncompleted. There were no differences ing viral illness which had led to admission to between responders and non-responders in hospital or consultation with a general practi- terms of age and sex. A similar proportion of tioner in the past two years in patients aged persons in each group responded. Of the 16-65 years. Exclusion criteria were as for the responders, there were 76 controls and 83 patients. Viral illnesses characterised by persis- patients. Table 2 gives a description of the tent pathology-such as B, enterovi- diagnoses of control subjects.

Table 1 Definition offatigue states Chronic fatigue syndrome Chronicfatigue syndrome + Table 2 Diagnosesfor the 76 controls Chronic fatigue (Green College) (CDC) Diagnosis No (%) * Score > 4 on fatigue * As for chronic fatigue * As for chronic fatigue questionnaire syndrome Hepatitis A 25 (32-8) Plus Plus Epstein-Barr hepatitis 2 (2 6%) * Present > 6 months *Present 50% of the time *Four of the following Adenovirus 6 (7-9) * No medical illness or major *Causes functional impairment symptoms: new headache; 1 (13) mental illness to explain fatigue *Both mental and physical post exertional malaise; 20 (26 3) symptoms of fatigue memory/concentration 4 (5-3) difficulties ; sore Unknown: URTI 13 (17 1) throat; ; Unknown: 2 (2-6) polyarthralgia, and Unknown: 2 (2 6) unrefreshing sleep URTI = upper respiratory tract infection. 506 Hotopf, Noah, Wessely

DEMOGRAPHIC CHARACTERISTICS on the checklist of 32 possible symptoms J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 May 1996. Downloaded from There were no significant differences between (mean number for cases 7 4 and 9 0 for con- those exposed to viral meningitis and controls trols, t = 1 10, 154 df, P = 0 27). To assess in terms of age (32 v 31 years), home owner- whether the cases were at greater risk of neu- ship (56% v 51% own home), duration of stay rological symptoms total scores on the eight in hospital (5 9 days v 4 9 days), and marital neurological symptoms (headaches, , status (55% v 45% married or cohabiting). tingling in the arms and tingling in the legs, Cases of viral meningitis were more likely to double vision, light headedness, increased sen- be female (64% v 46%, x = 5-08, ldf, P = sitivity to noise, and increased sensitivity to 0 02), and to have taken more time off work light) were compared. Again no excess of after the illness (49 days v 30 days, U = 2115, symptomatology was found (1 *8 v 2- 1 neuro- z = -2-13, P = 0 03 (Mann-Whitney U test)). logical symptoms for cases and controls The duration of follow up was also longer in respectively: t = 1-41, 155 df, P = 0-16). patients who were followed up for a mean of Impairment measured on the six subscales of 18 months compared with 13 for the controls the SF-36 showed no differences between (U= 1965, z = -4 00, P = 0.0001). The dura- cases and controls in any area of functioning tion of follow up and sex differences are likely (table 4). This finding was not changed when to confound the results of other outcome vari- odds ratios were adjusted for age, sex, and ables. Nine of the control subjects were not duration of follow up using logistic regression admitted to hospital but were treated by their analysis. . FACTORS PREDICTIVE OF CHRONIC FATIGUE, OUTCOME VARIABLES CHRONIC FATIGUE SYNDROME, AND CHRONIC Fatigue and psychiatric outcome FATIGUE SYNDROME + Table 3 shows the outcome for cases and con- We went on to examine the features of those trols. Cases of viral meningitis were not at we identified with chronic fatigue, chronic increased risk of chronic fatigue, chronic fatigue syndrome, and chronic fatigue syn- fatigue syndrome, or chronic fatigue syndrome drome +. These three conditions represent +. There was no increase in risk of depression gradations of severity along the range of as measured by the BDI, or emotional distress fatigue so it is not surprising that there were on the GHQ-12. To control for the potential more cases of chronic fatigue (36) than of confounding variables of age, sex, and dura- chronic fatigue syndrome (20) and chronic tion of follow up, a forced entry logistic regres- fatigue syndrome + (11). For all three diag- sion analysis was performed. Again no noses there were no differences between cases differences could be detected in outcome and healthy controls in terms of age, sex distri- between cases and controls. These outcomes bution, duration of follow up, or duration of were no different if the controls who had not stay in hospital. Table 5 summarises the main been admitted to hospital were excluded from differences between those with chronic fatigue the analysis. syndrome and those who were well at follow up. These results are similar for chronic Physical symptoms and impairments fatigue and chronic fatigue syndrome +.

Cases and controls did not differ in their score More cases had used services for emotional http://jnnp.bmj.com/

Table 3 Outcome for cases of viral meningitis and other viral illness controls Cases Controls 95% CIfor (n = 83) (n = 76) X2 Pvalue OR 95% CI adj OR adj OR Chronic fatigue 21 15 0-62 0-43 1-35 0-6-2-9 1 13 0-5-2-6 (25%) (20%) Chronic fatigue syndrome 12 8 0 47 0-49 1-39 0-5-3-6 0-96 0-3-2-8

(14%) (10%) on October 2, 2021 by guest. Protected copyright. Chronic fatigue syndrome + 8 3 1 93 0-16 2-55 0-6-10-0 1-78 0-4-8-0 (10%) (4%) Case on GHQ 39 37 0-15 0-15 0-88 0-4-1-6 0-64 0-31-1-34 (3/4) (48%) (51%) Case on BDI 11 17 2-82 0 09 0-49 0-4-2 3 0-78 0-14-1 01 (12/13) (13%) (24%) OR = Odds ratio; adj OR = adjusted odds ratio.

Table 4 Impairment on SF-36 atfollow up Cases Control 95% CIfor (n = 83) (n = 76) X2 P value OR 95% CI adj OR adj OR Psychological 31 30 0-17 0-67 0 87 0-5-1-7 0 79 0 4-1 7 impairment (38%) (41%) 24 20 0-04 0-83 1-08 0-5-2-2 1-02 0-5-2-2 (29%) (28%) Perception of illness 43 39 0-02 0-88 0 95 0-5-1-8 1 20 0-6-2-5 (54%) (55%) Physical impairment 45 41 0-06 0 79 0-92 0-1-1-7 1-14 0-5-2-4 (55%) (57%) Role 21 15 0-48 049 1-31 0-6-2-8 1-68 0-7-40 fumction (26%) (21%) impairment Social impairment 4 9 2-7 0-1 0 37 0-1-1-3 0-28 0 1-1 1 (5%) (11%) OR = Odds ratio; adj OR = adjusted odds ratio. Chronic fatigue and minor psychiatric morbidity after viral meningitis: a controlled study 507

Table 5 Characteristics ofpatients with chronicfatigue syndrome Chronic Non-chronic J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 May 1996. Downloaded from fatigue fatigue syndrome syndrome (n = 20) (n = 139) Statistic P value OR (95% CI)

Age (y (SD)) 34-1 (11 1) 31-8 (10-4) *z = - 0-76 0 44 Females (%) 13 (65) 74 (54) X2 = 0-85 0-36 1-58 1 df (0-6-4 2) Duration of stayin hospital 5-4 (3 5) 5-4 (5-1) *z = -0 5 0-64 (days (SD)) Time taken off work (days 62-6 (99-1) 37-3 (63 4) *z = - 1 42 0-15 (SD)) Duration offollowup 172 (6-3) 15-5 (7-7) *z= - 114 0-25 (months (SD)) Psychiatrichistory(%) 11 (55) 29 (21) X2 = 10 5 0-001 3 05 1 df (1 2-8-1) Depression on BDI (%) 9 (35) 20 (13) X2= 7-24 0 007 3-77 1 df (14-10 4) Distress on GHQ-12 (%) 16 (80) 60 (44) x2 = 8-81 0 003 5 00 1 df (16-157) *Mann-Whitney U test.

Table 6 Adjusted odds ratios after logistic regression analysis to accountforpotential confounding variables, age, sex, and duration offollow up Chronicfatigue Chronicfatigue Chronicfatigue syndrome syndrome + Psychiatric history P = 0-55 P = 0-02 P = 0 006 OR = 1-33 OR = 3-58 OR = 7-82 (0-5-3 4) (1-2-10-6) (1-8-34 3) Prolonged convalescence P = 0-002 P = 0-02 P = 0 12 OR = 4-85 OR = 4-93 OR = 3-8 (1-8-11 7) (1-3-18 8) (0-7-20 9) Values in parentheses are 95% CI.

problems and this relation became stronger as support infection as a specific risk the severity of fatigue increased. Those with factor for chronic fatigue syndrome. Secondly, chronic fatigue and chronic fatigue syndrome the prevalence of chronic fatigue syndrome for had had a longer convalescence than those the sample as a whole was considerably higher without fatigue states, but this difference did than that seen in primary care samples. not maintain statistical significance in those Thirdly, the variables which best predict with chronic fatigue syndrome +. Finally, chronic fatigue syndrome at follow up are pre- those with fatigue states were more likely to be vious psychiatric morbidity and a prolonged in psychological distress as measured on the convalescence. GHQ-12 and to have depression as measured These findings must be interpreted with on the BDI. care. This was a postal survey and a response rate of 77% leaves the results open to response http://jnnp.bmj.com/ LOGISTIC REGRESSION bias, although no differences between respon- To disentangle the potentially confounding ders and non-responders could be identified. variables of sex, duration of follow up, age, The sample sizes used limit the statistical previous service use for emotional problems, power of this study, and it is possible that and prolonged time offwork after viral illness, a small but significant differences between the forced entry logistic regression analysis was two groups could have been missed. The rarity performed. Prolonged time off work was of adult viral meningitis makes it expensive coded as above and below median (21 days) and impractical to perform large scale on October 2, 2021 by guest. Protected copyright. taken off work after the viral illness. This prospective studies except after an epidemic. allowed odds ratios (ORs) for the risk of There are also demographic differences fatigue states if an individual had had a psychi- between the two groups. There was an excess atric history or prolonged convalescence. of women among the cases. We would have Psychiatric history was no longer a statistically expected a higher rate of psychiatric disorder significant association with chronic fatigue, and fatigue among women so this potential but was associated with chronic fatigue syn- confounder is likely to push the results toward drome and chronic fatigue syndrome +. a positive finding (an excess of fatigue among Prolonged convalescence was associated with those exposed to viral meningitis). The other chronic fatigue and chronic fatigue syndrome. main difference was that the duration of follow Table 6 shows the adjusted ORs for these risk up was longer in the viral meningitis group. factors. This potential confounder should work in the opposite direction-that is, to increase the level of morbidity among the controls, who would Discussion be at an earlier stage of their convalescence. There were three main findings in this study. Our analyses suggest that these potential con- Firstly, there were no statistically significant founders are not powerful-there was no differences in outcome between the group who excess of fatigue among the female responders had viral meningitis compared with other viral and the cases of fatigue identified were at a infections. This suggests that there is little to similar stage in their convalescence to those 508 Hotopf, Noah, Wessely

who were well. When these potential con- Assuming that the result of high levels of founders were adjusted for by logistic regres- chronic fatigue syndrome are not due to bias J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 May 1996. Downloaded from sion analysis the lack of association between or confounding, it is necessary to construct a viral meningitis and subsequent fatigue model to bring our other findings together. remained. Previous service use for emotional disorder The morbidity in the group as a whole was and a prolonged period of time off work after high. The prevalence of chronic fatigue syn- the viral illness were associated with chronic drome was higher than that of previous sur- fatigue syndrome. These variables are not veys using the same instruments in primary independent. Logistic regression analysis care. For example, Wessely et aP1 found that showed that prolonged convalescence was a 9 9% and 1*3% of patients presenting to pri- predictor for chronic fatigue and chronic mary care with any viral infection had chronic fatigue syndrome. A history of previous con- fatigue or chronic fatigue syndrome respec- sultation for psychiatric disorder predicted tively when followed up for six months after chronic fatigue syndrome and chronic fatigue infection. These rates did not differ from those syndrome +. in the comparison group of patients presenting The finding that prolonged convalescence is in primary care for other reasons. Our esti- a risk factor for chronic fatigue syndrome mates for chronic fatigue syndrome (12-6%) could be interpreted in two ways. Firstly, those were considerably higher and we will examine who had a more serious viral illness would be possible reasons for this finding. more likely both to have a longer time off work It may be that all the viral illnesses studied and be more prone to chronic fatigue syn- act directly to cause chronic fatigue syndrome. drome. This explanation must be balanced For example, there is now some evidence that against the finding that the duration of stay in hepatitis A may cause chronic fatigue.26 hospital, which is predominantly a doctor led Unfortunately, without a control group admit- assessment of severity, is no greater in those ted to hospital for non-viral illness it is impos- who later developed chronic fatigue syndrome. sible to test this hypothesis. However, there Secondly, duration of time off work is a mea- are two findings which are against this view: sure of the patient's assessment of his or her (1) there was no relation between the duration symptoms and is more sensitive to the of initial admission to hospital, which may be patient's illness behaviour, which may in turn used as a proxy measure of severity and be associated with chronic fatigue syndrome. fatigue; and (2) there was no relation between Could a better understanding of the patho- duration of follow up and fatigue. If exposure genesis of fatigue be gleaned from these find- to severe viral infection on its own was respon- ings? Prolonged time off work after a viral sible for these high rates of fatigue, those who infection may be the principle mechanism of were early in their convalescence would be fatigue. Previous studies have found that bed expected to be most severely affected. rest or prolonged convalescence is a predictor The increased rates of chronic fatigue and of fatigue and subjective symptoms in infec- chronic fatigue syndrome could have been due tious mononucleosis'8 28 and postvaccinial (yel- to bias. With a response rate of 77%, response low ) illness.29 Further, there is evidence bias is not a sufficient explanation, but may to suggest that poor prognosis in established play a part, because those with fatigue may be chronic fatigue syndrome may relate to limit- more inclined to respond. Selection bias is an ing exercise and changing or leaving employ- http://jnnp.bmj.com/ alternative explanation: the processes which ment.30 The deconditioning caused by such led to admission of these patients to hospital changes may be responsible for fatigue. may have contributed to their increased preva- The relation between psychiatric morbidity lence of chronic fatigue syndrome on follow and chronic fatigue syndrome is well estab- up. Most patients with a viral infection are not lished8 and in a prospective study one of the admitted to hospital. This would apply to the most powerful predictors of chronic fatigue

infections seen among controls-for example, and chronic fatigue syndrome after a viral on October 2, 2021 by guest. Protected copyright. influenza and chickenpox. No studies are infection was a high score on the GHQ-12.25 available on the admission rates for viral Psychiatric morbidity is also an indicator of meningitis; however, there are degrees of poor prognosis of chronic fatigue syn- severity for the illness, ranging from asympto- drome.3' 33 There is good evidence that psychi- matic cases to the full syndrome of meningi- atric morbidity is an important determinant of tis'5 and it is reasonable to suppose that a functional recovery from other illnesses for proportion of patients with viral meningitis do example, recovery after myocardial infarction34 not consult or are not diagnosed as such. and back pain.35 Previous psychiatric morbid- There is some support for this explanation ity might be a risk factor by leading to pro- from a prospective study of consulting behav- longed convalescence, or alternatively may act iour after influenza infection.27 The attack rate directly due to the considerable overlap for clinically evident infection with Asian between depression, , and fatigue. influenza in the 1957 pandemic was increased Subsequent research should aim to com- almost threefold in high scorers on the pare recovery from a range of infectious and Minnesota multiphasic personality inventory. non-infectious physical illnesses with prospec- This increase applied to clinically evident ill- tive information on psychiatric morbidity and ness, not actual disease, as shown by a rise in changes in behaviour after acute illnesses. In titres. This explanation is unlikely to this way it may be possible to predict those at fully account for the rates of chronic fatigue risk of a poor outcome after physical illness, syndrome that we detected. and to plan early interventions. Chronic fatigue and minor psychiatric morbidity after viral meningitis: a controlled study 509

MH was supported by a grant from the Wellcome Trust. We clinical, sociomedical, electroencephalographic and psy- thank Professor Banatvala, Dr F Nicholson, Professor chological study. Acta Psychiatr Scand 1958;(suppl 126): J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.60.5.504 on 1 May 1996. Downloaded from Carmichael, Dr S Sutherland, and Dr B Bannister for their 1-115. advice and assistance in identifying subjects. 18 White PD, Grover SA, Kango HO, Thomas JM, Ames SJ, Clare AW. The validity and reliability of the fatigue syn- drome that follows glandular fever. Psychol Med 1995; 1 Noah N, Urquhart A. Virus meningitis and in 25:917-24. 1979. 7 Infect 1980;2:379-83. 19 Goldberg D. The detection of psychiatric illness by question- 2 Yousef G, Bell E, Mann G, et al. Chronic enterovirus infec- naire. London: Oxford University Press, 1972. tion in patients with postviral fatigue syndrome. Lancet 20 Beck AT, Ward CH, Mendelson M. An inventory for mea- 1988;i: 146-50. suring depression. Arch Gen Psychiatry 1961;4:561-71. 3 Calder BD, Wamock PJ. Coxsackie B infection in a 21 Chalder T, Berelowitz C, Pawlikowska T. Development of a Scottish general practice. Jf R Coil Gen Pract 1984;34: fatigue scale.3Psychosom Res 1993;37:147-54. 15-9. 22 Stewart AD, Hays RD, Ware JE. The MOS short-form 4 Calder BD, Wamock PJ, McCartney RA, Bell EJ. general health survey. Med Care 1988;26:724-32. Coxsackie B and the post viral fatigue syndrome: a 23 Ray WA, Fought RL, Decker MD. Psychoactive and prospective study in general practice. JI R Coll Gen Pract the risk of injurious motor vehicle crashes in elderly dri- 1987;37:11-4. vers. AmJ7 Epidemiol 1992;136:873-83. 5 Keighley BD, Bell EJ. Sporadic myalgic encephalomyelitis 24 Fukuda K, Straus S, Hickie I, Sharpe M, Dobbins J, in a rural practice. _J R Coil Gen Pract 1983;33:339-41. Komaroff A. The chronic fatigue syndrome: a compre- 6 Bell EJ, McCartney RA, Riding MH. Coxsackie B virus hensive approach to its definition and study. Ann Intern and myalgic encephalomyelitis. J R Soc Med 1988;81: Med 1994;121:953-9. 329-32. 25 Wessely S, Chalder T, Hirsch S, Palikowska T, Wallace P, 7 Gow JW, Behan WM, Clements GB, Woodall C, Riding Wright DJM. Post-infectious fatigue: prospective cohort M, Behan PO. Enteroviral RNA sequences detected by study in primary care. Lancet 1995;345:1333-8. polymerase chain reaction in muscles of patients with 26 Berelowitz GJ, Burgess AP, Thanabalasingham T, Murray- post viral fatigue syndrome. BMg 199 1;302:692-6. Lyon IM, Wright DJM. Post-hepatitis syndrome revis- 8 Hotopf MH, Wessely SC. Viruses, neurosis and fatigue. Jf ited. Hepatitis 1996 (in press). Psychosom Res 1994;38:499-514. 27 Cluff LE, Canter A, Imboden JB. Asian influenza: infec- 9 Miller NA, Carmichael HA, Calder BD, et al. Antibody to tion, disease and psychological factors. Arch Intern Med Coxsackie B virus in diagnosing postviral fatigue syn- 1966;117: 159-63. drome. BMJ 1991;302:140-3. 28 Dalrymple W. Infectious mononucleosis: 2. relation of bed 10 Lynch S, Seth R. Postviral fatigue syndrome and VP-1 anti- rest and activity to prognosis. Postgrad Med 1961;35: gen. Lancet 1989;2:1160-1. 345-9. 11 Halpin D, Wessely S. VP-1 antigen in chronic fatigue syn- 29 Benjamin JE, Hoyt RC. following postvaccinial drome. Lancet 1989;1:1028-9. () hepatitis. JAMA 1945;128:319-23. 12 Gow JW, Behan WM, Simpson K, McGarry F, Keir S, 30 Wilson A, Hickie I, Uloyd A, et al. Longitudinal study of Behan PO. Studies on enteroviruses in patients with outcome of chronic fatigue syndrome. BMJ 1994;308: chronic fatigue syndrome. Clin Infect Dis 1994;(suppl 1): 756-9. S126-9. 31 Sharpe M, Hawton K, Seagroatt V, Pasvol G. Follow up of 13 Swanink C, Melchers W, van der Meer J, et al. patients presenting with fatigue to an infectious diseases Enteroviruses and the chronic fatigue syndrome. Clin clinic. BMJ 1992;305:147-52. Infect Dis 1994;19:860-4. 32 Clark RR, Katon W, Russo J, Kith P, Sintay M, Buchwald 14 Clements GB, McGarry F, Nairn C, Galbraith DN. D. Chronic fatigue: risk factors for symptom persistence in Detection of enterovirus specific RNA in serum: the rela- a 2/2 year follow up study. Am Y Med 1995;98: tionship to chronic fatigue. JMed Virol 1995;45:156-61. 187-95. 15 Jubelt B, Lipton HL. Enterovirus infection. In: McKendall 33 Wilson A, Hickie I, Lloyd A, et al. Longitudinal study of RR, ed. Handbook of clinical . Amsterdam: outcome of chronic fatigue syndrome. BMg 1994;308: Elsevier 1989. 756-9. 16 Lepow ML, Coyne N, Thompson LB. A clinical epidemio- 34 Ladwig KH, Roll G, Breithardt G, Budde T, Borggrefe M. logic and laboratory investigation of aseptic meningitis Post-infarction depression and incomplete recovery 6 during the four year period 1955-1958: II the clinical months after acute myocardial infarction. Lancet disease and its sequelae. N Engl _J Med 1962;266: 1994;343:20-3. 1188-93. 35 Burton K, Tillotson M, Main C, Hollis S. Psychosocial 17 Muller R, Nylander I, Larsson L, Widen L, Frankenhauser predictors of outcome in acute and sub-chronic low back M. Sequelae of primary aseptic : a trouble. Spine 1995;20:722-8. http://jnnp.bmj.com/ on October 2, 2021 by guest. Protected copyright.