177 British Journal of Health Psychology (2008), 13, 177 186 q 2008 The British Psychological Society The British Psychological Society www.bpsjournals.co.uk A longitudinal study of the relationship between psychological distress and recurrence of upper respiratory tract infections in chronic fatigue syndrome Susan Faulkner 1 * and Andrew Smith 2 1 University of Glamorgan, UK 2 Cardiff University, UK Objectives. Previous research has found that chronic fatigue syndrome (CFS) patients report increased susceptibility to upper respiratory tract illnesses (URTIs) when compared with healthy volunteers. This study aimed to replicate and extend this research by investigating the role of psychological distress (stress and negative mood) in the recurrence of URTIs in CFS patients as well as its role in the recurrence of CFS symptoms. Design. A 15-week diary study. Methods. Measures of psychological stress, negative mood, recurrence of URTIs and symptoms were recorded each week for a 15-week period. CFS patients (N ¼ 21), who had been assessed and diagnosed according to the Oxford criteria, were recruited from the Cardiff Chronic Fatigue Clinic and compared with a matched group of healthy controls (N ¼ 18). Frequency of occurrence of infectious illness and the relationship between psychological stress/negative mood and occurrence of illness were assessed. Results. CFS patients reported more URTIs than the controls. Stress scores (and negative mood) were significantly higher in the week prior to the occurrence of URTIs than in weeks when no subsequent illness occurred. High levels of psychological stress also preceded the severity of reported symptoms of fatigue in the CFS group. Conclusions. CFS patients reported more frequent URTIs than healthy controls and these recurrences were preceded by high levels of psychological stress. High levels of stress were also associated with greater subsequent fatigue. Possible explanations of these results are discussed. Chronic fatigue syndrome (CFS) is an illness in which the predominant clinical symptom is disabling fatigue together with various additional symptoms which might include myalgia, lymphadenopathy, cognitive impairment and depression. (Holmes et al., 1988) * Correspondence should be addressed to Susan Faulkner School of Humanities, Law and Social Sciences, University of Glamorgan, Pontypridd, CF37 1DL, UK (e-mail: sfaulkne@glam.ac.uk). DOI:10.1348/135910706X171469
178 Susan Faulkner and Andrew Smith CFS is defined by the fatigue and related symptoms being present for at least 6 months. Diagnosis is symptom-based and to date there is no agreed aetiology or distinguishing physical examination to confirm diagnosis (Afari & Buchwald, 2003). There have been a number of approaches to understanding the aetiology and progression of CFS which may be categorized into viral theories, theories of immunological vulnerability and psychosocial theories. These approaches are not mutually exclusive and a logical model for examining how these approaches may interact in either the onset or the progression of this illness would be the biopsychosocial model. Several viruses have been associated with CFS, including herpes virus, retroviruses and enteroviruses (Ablashi, 1994). The role of Epstein Barr virus in the development of CFS has been widely debated and there is substantial overlap between the symptoms of chronic glandular fever and those of CFS. Whilst there is no consistent evidence that CFS is caused by a specific viral infection (Hotopf & Wessely, 1994), there is evidence that a fatigue syndrome can develop following glandular fever (White et al., 1995b, 1995a) and viral meningitis (Hotopf, Noah, & Wessely, 1996). Any relationship between viral infection and CFS is likely to be mediated by immune function and there is evidence that common viral infections are higher in CFS sufferers than controls (Smith, Thomas, Borysiewicz, & Llewelyn, 1999). Given the fact that exposure to such viruses is unlikely to be higher than in the general population, this finding provides evidence for the argument that immunosuppressive factors in the host increases their vulnerability to frequent infectious illness. An alternative explanation, however, could be that this simply reflects greater symptom reporting amongst CFS patients. When investigating self-reported symptoms of stress and illness, it is important to control for the role of negative affect which is known to increase the likelihood of symptom reporting. (Watson & Pennebaker, 1989). Failure to consider the role of negative affect in symptom reporting is a common methodological shortcoming in much of the research in this area. In a study by Smith et al. (1999), it was found that CFS patients reported significantly more upper respiratory tract infections (URTIs) over a 10-week period than controls. This was a diary study where the weekly occurrence of URTIs were recorded, although this was based upon self reported illness and the study did not establish whether differences were due to increased symptom reporting or increased vulnerability to infectious illness (Smith et al., 1999). Further research, again using diary methodology, by Smith and Fox (2000), included objective measures of illness severity and infection and found that CFS patients were more susceptible to acute infections and the virological results showed that this was not due to biased reporting. There is now significant evidence demonstrating an association between psychological stress and increased susceptibility to infectious disease (e.g. Cohen & Williamson, 1991). This research is supported by the finding that the central nervous system (CNS), the endocrine system and the immune system interact with each other and that psychological stress can suppress immune function via the hypothalamicpituitary-adrenal axis (HPA) and the autonomic nervous system (Glaser & Kiecolt- Glaser, 1998). Furthermore, studies of experimentally induced URTIs have found that increased reported levels of psychological stress were associated with an increased risk of URTIs. The rate of both respiratory infection and clinical colds increased in a doseresponse manner with increases in the degree of psychological stress. In these studies, they were able to establish that URTIs were associated with host resistance and not with differential exposure to the virus. (Cohen, Tyrell, & Smith, 1991; 1993).
Stress and infectious illness in CFS patients 179 Evidence for the role of psychological stress, depression and anxiety in CFS patients has been found in several studies (Smith et al., 1999; Wessely et al., 1995) although its precise role in the progression of CFS is not clearly understood. Much of the previous research has focused on psychological and psychiatric after effects or correlates of CFS or their role as precursors to the onset of CFS. It is possible that stress, depression and negative mood play a role in maintaining the illness via their effect on immune function, resulting in a negative cycle of stress, immunosuppresion, recurrence of infection or symptoms of the illness. Frequency of infectious illness, however, was found by Smith et al. (1999) not to influence the pathogenesis of CFS. If it is the case that CFS patients are more susceptible to URTIs and that immunosuppression in CFS patients mediates this relationship, then it follows that psychological stress may be an important factor in the recurrence of URTIs in CFS patients. The aim of this study was to investigate the longitudinal relationship between psychological stress and/or negative mood and recurrence of URTIs in CFS patients. The study also investigated the role of psychological stress in the severity of fatigue in CFS patients. This study specifically addressed the possible temporal association between stress and subsequent occurrence of URTIs as well as temporal association between psychological stress and experience of CFS symptoms, specifically fatigue. The study considered different aspects of the stress process, including perceived stress and negative mood. Method Design A 15-week diary study was conducted measuring the weekly incidence and severity of URTIs in CFS patients and healthy controls. Diaries were completed and returned weekly. Participants were asked to complete the diaries as far as possible at the same time each week. Participants Patients had previously attended the Cardiff Chronic Fatigue Syndrome Clinic and were volunteers from the Cardiff CFS research panel. Patients attending this clinic were given a full examination whereby a full clinical history was taken and the major features of the disease were assessed together with the standard clinical and biochemical investigations. Patients in the sample were formally diagnosed by clinicians as suffering from CFS using the Oxford criteria which excludes those suffering a co-morbid melancholic, psychotic or bipolar mood disorder. The healthy controls were partners or friends of the patients, the aim being both to control for exposure to infecting agents and to match for life-style factors. The sample comprised 21 CFS patients and 18 healthy controls. Measures The weekly diaries required participants to indicate whether or not they had suffered from any infectious illnesses that week and to specify which illness they had suffered. General health, levels of psychological stress and negative mood were recorded weekly (responses were measured on 5-point, Likert-type scales). The concept of psychological stress adopted for this study was based upon Lazarus s transactional model whereby the appraisal
180 Susan Faulkner and Andrew Smith of stress is a key factor in understanding its impact. (Lazarus & Folkman, 1984). They were also given a symptom checklist on which they recorded the presence and severity (not at all, moderate, severe) of a range of symptoms including symptoms associated with URTIs as well as CFS symptoms. Participants were required to report whether they were suffering from physical fatigue associated with their illness and the severity of fatigue. If they had been ill that week, further information about the illness and medication were requested. These measures have been used in previous research and have been found to correlate significantly with standardized measures of stress and illness (Faulkner, 2005). A range of standardized measures of stress and psychological health were taken at baseline. These included a Life Events Scale (based on Henderson, Byrne, & Duncan- Jones, 1981), the Perceived Stress Scale (Cohen, Kamarck, & Mermelstein, 1983), Trait Anxiety (Spielberger, Gorsuch, & Lushene, 1970), the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and the Positive and Negative Mood Scale (Zevon & Tellegen, 1982). Results Analysis of variance, analysis of covariance and related t tests (using BMDP) were used to analyse the data. Demographics and clinical profile The demographic characteristics of the patients and controls are shown in Table 1. These results show that the groups were well matched, except for gender. There were more females in the CFS group and more males in the control group. This was a consequence of using partners of the patients as controls which helped ensure similarity of life-style factors between the groups. Analyses of covariance were conducted to ensure that differences between the groups were not influenced by gender or age. Furthermore, the analysis investigating the temporal relationship between psychological stress and illness was within participants. Table 1. Demographic characteristics of CFS patients and controls CFS patients Controls Gender Male 4 11 Female 17 7 Mean (SD) age (years) 50 (12.11) range 29 77 49.6 (10.45) range 34 75 Marital status Single 3 1 Married 13 12 Divorced/separated 6 5 Education level Left school before 16 8 3 Completed secondary education 8 10 University graduate 5 5
Stress and infectious illness in CFS patients 181 The clinical profile of CFS patients is shown in Table 2 and the reported symptoms and symptom severity is reported in Table 3. This profile reflects the typical clinical characteristics of CFS patients. (Smith, Behan, Bell, Millar, & Bakheit, 1993). Most of the patients were recovering with occasional lapses (43%) or bad with some recovery (33%). Their views on what precipitated their illness shows that stress was perceived to be an important factor although infectious illnesses such as influenza, colds and glandular fever were nominated by significant numbers as being important in the onset of their illness. Table 2. Clinical profile of CFS patients Duration of illness Mean (months) 124.8 range 2 240 Current severity (no. of participants at each stage) Worse than at any stage 0 Bad 3 (15%) Bad with some recovery 7 (33%) Recovering with relapses 9 (43%) Almost recovered 2 (9%) What precipitated onset of CFS: Influenza 6 (32%) Sore throat 8 (42%) Glandular fever 3 (16%) Severe stomach upset 2 (11%) Stress 13 (68%) Other 6 (31%) Cross-sectional analysis of baseline psychosocial measures A series of one way analyses of covariance, with trait anxiety (negative affect) as a covariate were performed to explore the differences between CFS patients and controls on measures of psychological stress, negative mood and depression. The result of this analysis confirms the findings that CFS patients suffer significantly higher levels of psychological stress, negative mood and depression than controls and, furthermore, these differences do not appear to be a consequence of negative affect influencing their reported levels of stress and illness (see Table 4). Stress and negative mood Average weekly stress and negative mood over the diary period were calculated again controlling for negative affect. Analysis of covariance revealed significant differences between the groups on these measures. (Low scores ¼ more stress, more negative mood. The mean stress score for CFS patients, after covarying negative affect ¼ 41:52 [s:e: ¼ 1:95]; controls ¼ 45:11 [s:e: ¼ 2:14]. Fð1; 23Þ ¼4:76, p, :0005). The mean mood score after covarying negative affect was: CFS patients ¼ 45:45 [s:e: ¼ 1:51], controls ¼ 54:02 [s:e: ¼ 1:64], Fð1; 23Þ ¼16:28, p, :0001). Reporting of URTIs and other illnesses Figure 1 shows the percentage of CFS patients and healthy control participants reporting an URTI each week. It can be seen that the CFS patients reported significantly more URTIs over the diary period than controls. Analyses of covariance, controlling for
182 Susan Faulkner and Andrew Smith Table 3. Reported symptoms and symptom severity in CFS patients Symptoms % Yes Mean severity (SD) range ¼ 1 5 Physical weakness 100 2.8 (1.3) Excessive fatigue 100 3.2 (1.4) Legs feel heavy 86 2.6 (1.2) Muscle pain 72 2.5 (1.2) Pain in chest 38 1.6 (0.9) Painful joints 76 2.5 (1.2) Nausea 67 2.1 (1.1) Indigestion 57 2.1 (1.2) Bloated stomach 81 2.6 (1.2) Wind 76 2.5 (1.2) Sore throat 52 2.1 (1.2) Headache 71 2.6 (1.3) Earache 29 1.3 (0.6) Sore eyes 71 2.2 (0.9) Sensitive to noise 81 2.5 (1.2) Sensitive to light 67 2.4 (1.3) Feeling hot/cold 86 2.9 (1.3) Sweating 67 2.3 (1.2) Shivering 71 2.3 (1.3) Swollen glands 35 1.7 (0.9) Racing heart 55 1.7 (0.9) Insomnia 71 2.5 (1.1) Depression 55 2.2 (1.2) Anxiety/panic 70 2.3 (1.3) Loss concentration 95 3.0 (1.4) Loss of memory 81 2.5 (1.1) Allergies 30 1.5 (0.7) negative affect, were used to calculate the overall difference in the number of URTIs during the diary period (controls mean ¼ 0:55 [SD ¼ 0:82]; CFS mean patients 6.93 [SD ¼ 3:95], Fð1; 24Þ ¼2:75, p, :001). Levels of URTIs in CFS patients in this study are similar to those reported by Smith et al. (1999). The differences between the groups in the number of other (non-infectious) illness were not significant (controls mean ¼ 1:45 [SD ¼ 4:50], CFS mean ¼ 4:31 [SD ¼ 4:35], Fð1; 24Þ ¼2:48, p, :13). Table 4. Adjusted means (s.e.) scores for CFS participants and controls on baseline psychosocial measures covarying negative affect Psychosocial variable Healthy controls N ¼ 16 Mean (s.e.) CFS patients N ¼ 21 Mean (s.e.) (F ) p Perceived stress 17.8 (9.3) 27.1 (8.1) (10.5) p, :002 Negative life-events 1.1 (1.6) 2.6 (2.1) (5.4) p, :02 B.D.I. 6.8 (9.03) 13.8 (5.9) (7.98) p, :007 Negative mood 10.3 (8.9) 20.9 (11.7) (9.06) p, :004
Stress and infectious illness in CFS patients 183 Figure 1. shows the percentage of CFS patients and control participants reporting infectious illness and other illness each week. Temporal relationships between psychosocial measures and occurrence of URTIs/fatigue This analysis investigated the temporal relationship between psychological well-being and occurrence of illness. The aim was to assess whether psychological stress and/or negative mood directly preceded the development of illness in CFS patients i.e. to assess the longitudinal relationship between stress and illness. This was a within-participant analysis which compared stress scores in the weeks prior to infectious illness and high levels of physical fatigue with the stress scores on weeks prior to no illness/low fatigue. Within-group analysis using related t tests were calculated to investigate whether the differences in stress scores were significant. A participant s stress score in the week prior to an illness was compared with their stress score in the weeks in which there was no subsequent illness. In order to control for the effect of illness on levels of stress, stress scores for those who had been ill in the previous week were excluded. The same analysis was also carried out for negative mood (i.e. mood scores the week prior to an illness were examined and compared with mood scores the week prior to no illness ). The same set of analyses was carried out firstly for the CFS participants only and secondly for the healthy control group. Table 5 shows the mean stress scores for CFS patients: (a) in the weeks prior to an URTI and (b) in the weeks where no subsequent URTI occurred. There were significant differences between the scores in these weeks indicating that high levels of psychological stress preceded the occurrence of URTI in CFS patients. The same analysis was conducted for the relationship between negative mood and infectious illness and it can be seen that scores for negative mood were significantly higher in weeks prior to the development of URTI, compared with the weeks where there was no subsequent URTI. This demonstrates that both these measures of psychological stress were significantly higher prior to the development of URTIs. The analysis was repeated for the severity of reported fatigue in CFS patients. A similar relationship was found between psychological stress and reporting of fatigue. Scores on psychological stress were significantly higher in the weeks which preceded moderately severe or extremely severe fatigue compared with the weeks that CFS sufferers reported no fatigue. Scores for negative mood, however, were not significantly higher in the week prior to physical fatigue.
184 Susan Faulkner and Andrew Smith Table 5. Mean (SD) scores for stress/mood in the week prior to illness and the week prior to no illness for CFS participants CFS patients df Prior to URTI Mean (SD) Prior to URTI Mean (SD) Matched t Sig. Stress and infectious illness: 18 2.09 (0.46) 3.04 (0.75 26.04 p, :0000 Mood and infectious illness: 19 2.48 (0.69) 3.07 (0.66) 23.76 p, :001 Prior to high fatigue Prior to low fatigue Stress and physical fatigue. 19 2.45 (0.53) 2.67 (0.58) 7.95 p, :000 Mood and physical fatigue. 19 2.88 (0.48) 2.92 (0.46) 1.89 p, :07 ns Note. Low score ¼ greater stress and negative mood. Healthy controls There were no significant differences in the weekly stress and negative mood scores in the week prior to the development of URTIs or the reporting of fatigue compared to weeks with no subsequent illness or fatigue for the healthy control group. Discussion These results show that psychological stress and negative mood are cumulatively higher for CFS patients than for healthy controls. These differences were maintained when trait anxiety was co-varied, suggesting that the differences were not simply a consequence of higher levels of symptom reporting in highly anxious individuals. This is an important issue in self-reported measures of this kind which are frequently taken at face value. The fact that the differences between the groups remained after taking negative affectivity into consideration avoids the possible biased reporting associated with this personality profile. The results in this study also confirm earlier findings of Smith et al. (1999) that CFS patients report significantly more URTIs than healthy controls. The study demonstrated, furthermore, that levels of psychological stress in CFS patients fluctuate from week-toweek and that there were significantly higher levels of stress and negative mood in the weeks prior to a recurrence of URTI than in weeks which did not precede an infectious illness. This indicates that there is a longitudinal relationship between psychological stress and reported development of URTI in the CFS group with higher levels of psychological stress and negative mood preceding the occurrence of reported URTIs. The healthy control group, however, was less affected by weekly fluctuations in stress and did not report higher levels of stress in the week prior to developing an infectious illness. This could either be simply a result of the fact there were very low levels of infectious illness in the control group or, alternatively, it could be explained in terms of the differential effects of acute and chronic stress. Cohen et al. (1998) found that acute stress (lasting less than one month) did not alter susceptibility to colds whereas enduring stress was associated with greater susceptibility to rhinovirus-induced colds. In this study, the CFS patients reported significantly higher scores on standardized, baseline measures of stress (chronic stress) and whilst reporting fluctuations in weekly stress, this maybe seen as exacerbating these existing high levels of stress in the CFS group. The control group, who were generally low on standard measures of stress and reported weekly fluctuations, might be viewed as experiencing acute stress. It can be
Stress and infectious illness in CFS patients 185 concluded that those suffering from CFS were subject to higher levels of psychological stress when compared with healthy controls and the experience of stress was directly associated with recurrence of URTIs and the severity of the symptoms of chronic fatigue. As mentioned earlier, these findings are unlikely to reflect general biases in symptom reporting because (a) personality characteristics known to increase symptom reporting were controlled and (b) the symptoms reported were specific to URTIs and random symptoms not related to infectious illness were not reported. A plausible explanation for these findings is that a significant relationship between stress and illness, mediated by the immune system, is important in understanding the progression and pattern of CFS following diagnosis. Sufferers of this illness may be characterized as suffering from chronically high levels of psychological stress, immunosuppression and subsequent recurrences of infectious illness and fatiguerelated symptoms. These results fit logically with the findings of studies which suggest that CFS patients have abnormal immune system function (Straus, 1966). They also support the findings of research which has demonstrated that psychological stress increases susceptibility to infectious illness discussed in the introduction (Cohen et al., 1991; Segerstrom & Miller, 2004). It would be better, of course, where possible to provide virological evidence of URTI to control for any tendency on the part of CFS patients to make somatic attributions for their symptoms, however, previous research by Smith and Fox (2000) which did confirm virology found support for the reliability of reported URTIs in CFS patients. Furthermore, Cohen et al. study of stress and induced URTIs (nasal challenge studies) suggests that this would be a logical explanation. Whilst this study did not measure immunological changes, other possible reasons for the relationship between stress and illness were eliminated making an explanation based on immune system changes more plausible. Conclusions CFS patients reported a statistically significant greater number of URTIs compared when with the control group. The occurrence of reported infectious illness as well as the severity of the symptoms of physical fatigue was preceded by significantly higher levels of weekly psychological stress and negative mood than in the weeks prior to no recurrence. References Ablashi, D. V. (1994). Viral studies of chronic fatigue syndrome. Clinical Infectious Diseases, 18(Suppl. 1), S130 S133. Afari, N., & Buchwald, D. (2003). Chronic fatigue syndrome: A review. American Journal of Psychiatry, 160, 221 226. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J. E., & Erbaugh, J. K. (1961). Psychometric properties of the Beck Depression Inventory: Twenty five years of evaluation. Clinical Psychology Review, 8, 77 100. Cohen, S., Frank, E., Doyle, W. J., Skoner, D. P., Rabin, B. S., & Gwaltney, J. M. (1998). Types of stressors that increase susceptibility to the common cold in adults. Health Psychology, 17(3), 214 223. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behaviour, 24(4), 385 396.
186 Susan Faulkner and Andrew Smith Cohen, S., Tyrrell, D. A. J., & Smith, A. P. (1991). Psychological stress and susceptibility to the common cold. New England Journal of Medicine, 25, 606 612. Cohen, S., Tyrrell, D. A. J., & Smith, A. P. (1993). Negative life events, perceived stress negative affect and susceptibility to the common cold. Journal of Personality and Social Psychology, 64(1), 131 140. Cohen, S., & Williamson, G. (1991). Stress and infectious disease in humans. Psychological Bulletin, 109, 5 24. Faulkner, S. (2005). The role of psychosocial factors in the progression and recurrence of illness: A study of Herpes Virus infectious illness and Chronic Fatigue Syndrome. Unpublished doctoral dissertation, Cardiff University, Wales. Glaser, R., & Kiecolt-Glaser, J. K. (1998). Stress-associated immune modulation: Relevance to viral infections and chronic fatigue syndrome. American Journal of Medicine, 105(3A), 35S 42S. Henderson, S., Byrne, D. G., & Duncan Jones, P. (1981). Neurosis and the social environment. New York: Academic Press. Holmes, G. P., Kaplan, J. E., Gantz, N. M., Komaroff, A. L., Schonberger, L. B., Straus, S. E., et al. (1988). Chronic fatigue syndrome: A working case definition. Annals of Internal Medicine, 108, 387 389. Hotopf, M., Noah, N., & Wessely, S. (1996). Chronic fatigue and minor psychiatric morbidity after viral meningitis: A controlled study. Journal of Neurology, Neurosurgery and Psychiatry, 60, 504 509. Hotopf, M. H., & Wessely, S. (1994). Viruses, neurosis and fatigue. Journal of Psychosomatic Research, 38, 499 514. Lazarus, R. S., & Folkman, S. (1984). Stress and the coping process. New York: Springer. Segerstrom, S. C., & Miller, G. E. (2004). Psychological stress and the human immune system: A meta-analytic study of 30 years of inquiry. Psychological Bulletin, 130(4), 601 630. Smith, A., Behan, P., Bell, W., Millar, K., & Bakheit, M. (1993). Behavioural problems associated with chronic fatigue syndrome. British Journal of Psychology, 84, 411 423. Smith, A., & Fox, J. D. (2000). Chronic Fatigue Syndrome, Report to the Linbury Trust. Smith, A., Thomas, M., Borysiewicz, L., & Llewelyn, M. (1999). Chronic fatigue syndrome and susceptibility to upper respiratory tract illness. British Journal of Health Psychology, 4, 327 335. Spielberger, C. D., Gorsuch, R., & Lushene, R. (1970). The State-Trait Anxiety Inventory (STAI) test manual Form X. Palo Alto: Consulting Psychologists Press. Straus, S. (1966). Chronic fatigue syndrome. In S. Rosen (Ed.), Medicine for the public (pp. 1 50). Bethesda, MD: National Institute of Health. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress and distress: Exploring the central role of negative affectivity. Psychological Review, 96(2), 234 254. Wessely, S., Chalder, T., Hirsch, S., Pawlikowska, T., Wallace, P., & Wright, D. J. M. (1995). Post infectious fatigue: A prospective study in primary care. Lancet, 345, 1333 1338. White, P., Grover, S., Kangro, H., Thomas, J. M., Amess, J., & Clare, A. W. (1995b). The validity and reliability of the fatigue syndrome that follows glandular fever. Psychological Medicine, 25, 917 924. White, P., Thomas, J., Amess, J., Grover, S. A., Kangro, H. O., & Clare, A. W. (1995a). The existence of a fatigue syndrome after glandular fever. Psychological Medicine, 25, 907 916. Zevon, M. A., & Tellegen, A. (1982). The structure of mood change: An idiographic/nomethetic analysis. Journal of Personality and social Psychology, 43, 111 122. Received 23 June 2006; revised version received 9 November 2006