Journal of Psychosomatic Research, Vol 53, Issue 3, 749-753 Date: September 2002 URL: http://www.sciencedirect.com/science/journal/00223999 Natural course and predicting self-reported improvement in patients with chronic fatigue syndrome with a relatively short illness duration ------------------------------------------------------------------------ Sieberen P. van der Werf(a), Berna de Vree(a), Maurice Alberts(a), Jos W. M. van der Meer(b), Gijs Bleijenberg(a) Netherlands Fatigue Research Group Nijmegen a Department of Medical Psychology, University Medical Centre, Post Box 9101, 6500 HB, Nijmegen, The Netherlands b Department of General Internal Medicine, University Medical Centre, Nijmegen, The Netherlands Corresponding Author Contact Information Corresponding author. Tel.: +1-24-361-3234; fax: +31-24-361-3425; email: s.vanderwerf@cksmps.azn.nl Received 7 September 2001; accepted 11 January 2002. Available online 4 September 2002. Abstract Objective: To describe the course of fatigue in chronic fatigue syndrome (CFS) patients with a relatively short duration of complaints and to test which psychosocial factors predict spontaneous improvement 1 year later. Methods: Seventy-nine patients with a complaint duration of less than 2 years were tested at baseline and 78 of the same group at 1-year follow-up. During this time period, no systematic intervention took place. Self-reported improvement and fatigue severity were the main outcome measures. Results: Forty-six percent (95% confidence intervals, 95CI=35-58%) of the patients with a short illness duration reported to be improved. This was a significantly (chi^2=20.3, P<.001) higher percentage compared to the 20% (95CI=15-26%) self-reported improvement in a previously published natural-course study among 246 CFS patients with a longer illness duration. Persistence of complaints after 1-year follow-up was associated with high baseline levels of experienced concentration problems, less strong psychosocial causal explanations for the complaints, and higher levels of the experienced lack of social support. Baseline fatigue severity predicted fatigue severity at follow-up. Conclusion: The results showed that CFS patients with a relatively short duration of complaints had a more favourable outcome compared to patients with a long illness duration. The data also indicated that complete recovery only occurred in patients with a complaint duration of less than 15 months. This finding has important implications, since it suggests that after such a time period spontaneous recovery hardly occurs. Author Keywords: Chronic fatigue syndrome; Natural course Introduction Chronic fatigue syndrome (CFS) is characterised by severe disabling fatigue of definite onset, lasting longer than 6 months, and for which no somatic cause can be found [1]. Follow-up studies in patients with CFS showed that only small proportions of these patients, varying between 0% and 8%, recovered spontaneously [2, 3, 4, 5 and 6]. The means of the illness duration in these studies ranged between 3 to 9 years, while the follow-up periods varied between 1 and 3 years. These differences make direct comparisons between these studies difficult. Nevertheless, the reported number of patients becoming symptom-free indicated that prognosis in CFS is rather poor. Follow-up data of patients who suffered prolonged fatigue (>1 month) but did not fulfill the criteria of 6-month illness duration showed that approximately half of the patients reported full recovery of their symptoms at follow-up [7 and 8]. These discrepancies between recovery rates in CFS patients and recovery rates in studies concerning patients suffering prolonged fatigue do suggest a reversed relation between complaint duration and spontaneous recovery of symptoms. Some prospective studies have indeed found that illness duration in CFS was a predictor of persistence of fatigue [2 and 9]. Therefore, there is a need for investigating prognosis in CFS patients with a relatively short duration of complaints. In a previous study of natural course in CFS patients with a median duration of complaints of 5 years, 20% of the CFS patients reported improvement (including 3% recovery) in complaints after a 1 View Image -year follow-up period [2]. In this study, the persistence of fatigue was not only predicted by long illness duration, high level of baseline fatigue, and functional disability, but also by the lack of perceived control over symptoms (low self-efficacy expectations) and strong physical attributions. The importance of physical attributions was confirmed by the cohort study of Wilson et al. [4], but illness duration, premorbid psychiatric diagnosis, and cell-mediated immune function, did not predict outcome. Another prospective study reported long illness duration and low internal locus of control to be significant predictors for persisting fatigue [9]. Based on their data, the authors suggested that interventions should enhance perceived control. Bombardier and Buchwald [3] surveyed large groups (>400) of persons with chronic fatigue and found that prognostic indicators differed for both groups. However, the presence of dysthymia at baseline predicted poorer outcome in both groups. An uncontrolled intervention study found that CFS patients with a shorter illness duration had a lower incidence of psychiatric disorder [10]. The authors of this study also reported that the effectiveness of counselling was greater in patients with a relatively short illness duration. The average complaint duration in their sample was 21 months, and after counselling 80% of their patients reported improvement at follow-up. In summary, most of the described studies suggested a possible relation between illness duration and spontaneous recovery, while psychosocial factors were consistently found to predict outcome. In spite of this, most of the data were limited to patients with a relatively long illness duration. Consequently, this follow-up study was aimed at assessing improvement rates in patients with a recent diagnosis of CFS (maximum illness duration of 2 years) and to compare them with the findings in our previous natural-course study. Furthermore, it was tested whether similar or other psychosocial factors as found in our previous natural-course study predicted outcome. Based on the literature, we expected higher improvement and recovery rates in patients with a relatively recent diagnosis of CFS. Method PARTICIPANTS This study was carried out at the University Medical Centre Nijmegen in the Netherlands and was approved by the Medical Ethical Committee. Informed consent was obtained prior to the start of the study. All patients with a major complaint of fatigue referred to the outpatient clinic of the Department of Internal Medicine were assessed by means of detailed history, physical examination, and computer assessment of questionnaires. Patients were eligible for the study if they met the U.S. Centers for Disease Control and Prevention criteria for CFS or idiopathic fatigue [1]. The patients had also to fulfill operational criteria for severe fatigue and severe functional impairment. These criteria were defined as a score of 35 or more on the subscale fatigue severity of the Checklist Individual Strength (CIS-Fatigue) and a sum score of 700 or more on eight subscales of the Sickness Impact Profile (SIP8). Additional inclusion criteria were aged between 18 and 65 years and a symptom duration of less than 2 years. Consecutive patients fulfilling the inclusion criteria of this study were approached. In this study, no record was kept about the number of participants who refused to participate. In total, 79 CFS patients with a relatively short duration of complaints (mean 1.4 years, minimum 6 months, and maximum 24 months) completed baseline assessment. Sixty-six (85%) of the 79 patients were female. The mean age of this sample was 34.8 years (range: 16-57 years). Educational level was rated with a seven-point educational scoring system applicable to the situation in the Netherlands (1=only primary school, 7=university degree). The mean educational score of this sample was 4.4, approximating 11 years of education. Seventy-five percent of the total sample worked (paid job) before the illness started while 29% of the sample indicated that they had worked (paid job) in the month preceding assessment. Seventy-eight of the 79 patients participated at follow-up, 1 year later. Comparison group Improvement ratings were compared with a previous natural-course study in which the same measurement method was used. The improvement ratings of this comparison group (n=199, 77% female) pertained to patients fulfilling similar operational inclusion criteria for fatigue severity and disability as used in the current study. The mean illness duration of this sample was 8.8 years and 51 patients had an illness duration of 2 years or less. MEASURES Improvement at follow-up Self-reported change At follow-up, patients were asked to evaluate their present situation in comparison with the situation at initial assessment. Patients could choose out of the following four categories: (1) recovered, (2) complaints improved, (3) similar complaints, (4) worse. Fatigue severity at follow-up Fatigue severity at follow-up was measured with the subscale CIS-Fatigue severity of the CIS. The CIS-Fatigue severity subscale has eight items rated on a seven-point scale. Scores can range between 8 (nonfatigued) and 56 (extremely fatigued) and scores of 35 or higher are indicative of severe fatigue [11 and 12]. Predictor variables Psychosocial Dimensions 1. Functional impairment: The total score of the eight subscales of the SIP8 was used. [13] 2. Psychological distress: Psychological distress was rated by the sum score of the depression (SCL-DEP) and anxiety (SCL-ANX) subscales of the Symptom Checklist-90 (SCL-90) [14]. 3. Dissatisfaction with the extent of social support: A subscale of the Social Support List (SSL) was used to rate the perceived discrepancy in actual support and wanted support (SSL-D) [15]. 4. Physical activity: The Physical Activity Rating Scale (PARS) served as measure of physical activity levels. Patients were asked to rate the frequency of 10 different activities over the last 2-week period on a five-point scale. These 10 ratings were then averaged into a total activity score that could vary between 0 (low physical activity level) and 5 (high physical activity level) [16]. 5. Self-efficacy: Sense of control over symptoms was measured with five questions (SE-5) that had been used in our previous studies. The score could range between 5 and 25. Higher score indicates higher levels of self-efficacy concerning fatigue [2 and 11]. 6. Causal attributions: Beliefs regarding the physical and psychosocial causes of fatigue complaints were measured with the Causal Attributions List (CAL). The CAL consists of two subscales and patients could indicate on each of the two 4-item subscales how strongly they agreed (scale 1-5) with the possible causes of their complaints. High scores on these two scales indicate stronger physical or stronger psychosocial attributions [2 and 11]. 7. Concentration problems: The CIS subscale concentration (CIS-CON) served as a measure of experienced concentration problems [2 and 17]. 8. Sleeping problems: The SCL-90 subscale sleeping problems (SCL-SLP) served as a measure of sleeping problems, e.g., items like cannot sleep, frequently wake up at night. 9. Focussing on bodily symptoms: The SCL somatisation (SCL-SOM) subscale was used to represent this dimension. 10. Fatigue severity: Baseline CIS-Fatigue score. STATISTICAL ANALYSES Improvement ratings between the two patient samples were compared with chi^2 statistics and 95% confidence intervals (95CI) were provided. The sample was divided into a group of patient who reported to be recovered or improved and a group of patients who reported no change or worsening of symptoms at follow-up. To compare these two samples with respect to their baseline psychosocial dimensional scores and mean illness duration, t tests were used. Additionally, the combination of baseline psychosocial dimensions that could best predict self-reported improvement and fatigue severity at follow-up was tested. Testing was carried out two-sided with an alpha set at .05 and, in case variables showed a nonnormal (-11) distribution, appropriate (log) transformations were applied. Results SELF-REPORTED IMPROVEMENT At follow-up, 6 (8%) patients indicated that they had no complaints anymore, 30 (38%) patients reported that they had less complaints, 29 patients reported that their complaints remained similar, and 13 (17%) patients indicated that their complaints had become worse. Thus, 46% (95CI=35-58%) of the patients reported to be improved or recovered, which was a significantly (chi^2=21.2, P<.001) higher percentage compared to the 19% (95CI=13-25%) improvement in the comparison group of 199 patients with a higher mean illness duration. Post hoc analysis within the comparison group also revealed a significant (chi^2=15.8, P<.001) difference in improvement rates between patients with a short and longer illness duration. In total, 36% (95CI=24-52%) of the 51 patients with a short illness duration (<=2 years) reported either full recovery (n=4) or improvement (n=15), while 12% (95CI=7-19%) of the patients with an illness duration of more than 2 years reported either full recovery (n=1) or improvement (n=17). Fisher's exact testing showed that full recovery was significantly (P<.05) more likely in the group with short illness duration (8%, 95CI=2-19%) compared to the subgroup with a long illness duration (<1%, 95CI=0-3%). RELATION BETWEEN SELF-REPORTED IMPROVEMENT AND BASELINE FATIGUE SEVERITY MEASURES In Table 1, baseline psychosocial dimensional scores of the group recovered and improved patients (improved group) were compared to the group of patients who reported no change or worsening of symptoms (nonimproved group) at follow-up. The improved group reported significantly less concentration problems at baseline and reported significantly more psychosocial causal attributions. The differences on both baseline fatigue severity (P=.05) and dissatisfaction with the extent of social support (P=.05) approached significance. Improved patients tended to report less baseline fatigue and less baseline dissatisfaction with social support. Table 1. Self-reported change and the relation with CIS-Fatigue severity and daily observed fatigue at baseline and follow-up ------------------------------------------------------------------------------------- - Baseline measurements Mean (S.D.) Univariate analysis, t value, df, significance ------------------------------------------------------------------------------------- - Recovered and No change or improved (n=36) worsening (n=42) ------------------------------------------------------------------------------------- Fatigue severity 48.6 (6.8) 51.0 (5.7) t= 1.96, df(75), P=.05 (CIS-Fatigue) Functional impairment 1593 (576) 1801(599) t=-1.56, df(76), ns (SIP8 total) Psychological distress 39.3 (9.3) 45.3 (16.3) t=-1.87, df(76), ns (SCL-ANX and SCL-DEP) Physical causal 14.1 (3.01) 14.6 (2.8) t=-0.74, df(76), ns attributions (CAL) Psychosocial causal 9.0 (1.8) 8.1 (2.1) t= 2.05, df(76), P<.05 attributions (CAL) Self-efficacy (SE-5) 15.9 (3.5) 15.0 (3.1) t= 1.20, df(76), ns Physical activity 44.2 (11.2) 44.2 (9.4) t=-0.00, df(75), ns ratings (PARS) Concentration problems 26.3 (5.9) 28.4 (7.0) t= 2.32, df(76), P<.05 (CIS-CON) Sleep problems (SCL-SLP) 7.4 (3.2) 8.3 (3.8) t= 1.07, df(76), ns Focussing on bodily 27.6 (6.2) 30.1 (9.2) t=-1.60, df(76), ns symptoms (SCL-SOM) Dissatisfaction with the 40.4 (6.2) 44.5 (11.2) t=-1.98, df(76), P=.05 extent of social support (SSL-D) Baseline illness 16.3 (4.7) 16.6(5.3) t=-0.30, df(76), ns duration (in months) ------------------------------------------------------------------------------------- - The table shows raw test scores and their standard deviations. Skewed data were log-transformed before significance testing. PREDICTION OF SELF-REPORTED IMPROVEMENT OF COMPLAINTS AND FATIGUE SEVERITY AT FOLLOW-UP Self-reported improvement A similar stepwise multiple regression analysis entering all the dimensional scores and disease duration at baseline resulted in a significant model [R2 adj.=.18, F=6.2, df(3,73), P<.01] with three significant predictors. Self-reported improvement was predicted by lower levels of experienced concentration problems (beta=-.33, P<.01), stronger psychosocial attributions concerning the cause of the complaints (beta=-.29, P<.01), and less dissatisfaction with amount of social support (beta=-.24, P=.03). Fatigue severity at follow-up A similar stepwise multiple regression analysis with fatigue severity at follow-up as dependent variable resulted in a significant model [R2 adj.=.11, F=10.2, df(1,73), P<.01]. Fatigue severity at baseline was the single significant predictor for fatigue severity at follow-up (beta=.35, P<.01), while experienced concentration problems did approach significance as a predictor (beta=.21, P=.05). Discussion This study specifically targeted CFS patients with a relatively short illness duration. The findings suggested that a shorter illness duration was associated with higher rates of spontaneous improvement and recovery. Since no details were available concerning nonparticipants, there is some uncertainty whether the investigated sample might have been biased towards a more favourable illness course. Nevertheless, a post hoc analysis among equally fatigued and disabled patients of a previous natural-course study provided further support for our finding that improvement and full recovery were more likely in patients with a short illness duration. Spontaneous recovery, however, was still rare, and all of the recovered patients in the present sample had an illness duration of less than 1 View Image years. The finding that many (36%) patients with a short illness duration reported improvement could imply higher success rates for early interventions in CFS, since it might be easier to treat patients who experienced a positive change in their complaints. The regression analyses showed that more baseline concentration problems predicted persistence or worsening of symptoms at follow-up and a similar trend was observed for the prediction of fatigue severity. A previous neuropsychological study of our group indicated that the experience of concentration problems in CFS was not related to actual performance on neuropsychological tests. It was suggested that experienced concentration problems might reflect higher levels of distress or a stronger focus on bodily symptoms [17]. Although physical causal attributions have been found to predict the course of fatigue in CFS [2 and 4], the extent of physical causal attributions did not predict complaint persistence in this group. However, when the patients favoured more strongly a psychosocial explanation for their complaints, the prognosis was better. Perhaps that most CFS patients suppose a physical cause for their complaints, but that some of these patients not rule out that, besides a physical cause, additional factors could play a role in their symptoms. It is also possible that patients with a relatively short illness are more open to different causal explanations, but that the need for a single more conclusive physical causal explanation strengthens when the complaints persist and no medical answers can be given over a prolonged period of time. Baseline dissatisfaction with the extent of social support turned out to be another significant predictor. This finding could indicate that early in the course of the illness, reactions of friends, relatives, and coworkers play a crucial role. Negative reactions or little understanding could provoke psychological distress, which in turn could promote the persistence of symptoms. Since this study showed that spontaneous improvement or recovery occurred more frequently in a relatively early phase of the illness, it would be interesting to see whether the self-rated improvement in these patients continued to complete recovery or whether relapse occurred in the relatively few recovered patients. References 1. K Fukuda, SE Straus, I Hickie, MC Sharpe, JG Dobbins and A Komaroff, The chronic fatigue syndrome: a comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Ann Intern Med 121 12 (1994), pp. 953-959. 2. JHWM Vercoulen, CMA Swanink, JFM Fennis, JMD Galama, JWM van-der-Meer and G Bleijenberg, Prognosis in chronic fatigue syndrome: a prospective study on the natural course. J Neurol, Neurosurg Psychiatry 60 5 (1996), pp. 489-494. 3. CH Bombardier and D Buchwald, Outcome and prognosis of patients with chronic fatigue vs. chronic fatigue syndrome. Ann Intern Med 155 19 (1995), pp. 2105-2110. 4. A Wilson, I Hickkie, AR Lloyd, PD Hadzi, C Boughton, J Dwyer and D Wakefield, Longitudinal study of outcome of chronic fatigue syndrome. BMJ, Br Med J 308 (1994), pp. 756-759. 5. U Tirelli, G Marotta, S Improta and A Pinto, Immunological abnormalities in patients with chronic fatigue syndrome. Scand J Immunol 40 (1994), pp. 600-601. 6. PK Peterson, CH Schenk and R Sherman, Chronic fatigue syndrome in Minnesota. Minn Med 74 5 (1991), pp. 21-26. 7. AF Valdini, S Steinhardt, J Valicenti and A Jaffe, A one-year follow-up of fatigues patients. J Fam Pract 26 1 (1988), pp. 33-38. 8. E Nelson, J Kirk, G McHugo, R Douglass, J Ohler, J Wasson and M Zubkoff, Chief complaint fatigue: a longitudinal study from the patient's perspective. Fam Pract Res J 6 4 (1987), pp. 175-188. 9. C Ray, S Jefferies and WR Weir, Coping and other predictors of outcome in chronic fatigue syndrome: a 1-year follow-up. J Psychosom Res 43 4 (1997), pp. 405-415. 10. MF Shanks and YD Ho, A clinical study of chronic fatigue syndrome. Br J Psychiatry 166 6 (1995), pp. 798-801. 11. JHMM Vercoulen, CMA Swanink, JFM Fennis, JMD Galama, JWM van-der-Meer and G Bleijenberg, Dimensional assessment of chronic fatigue syndrome. J Psychosom Res 38 5 (1994), pp. 383-392. 12. JHMM Vercoulen, M Alberts and G BleijenbergDe Checklist Individual Strength (CIS). Gedragstherapie 32 2 (1999), pp. 131-136. 13. M Bergner, RA Bobbit, WB Carter and BS Gilson, The Sickness Impact Scale: development and final revision of a health status measure. Med Care 19 (1981), pp. 787-805. 14. WA Arrindell and H Ettema, Dimensional structure, reliability and validity of the Dutch version of the Symptom Checklist (SCL-90): data based on a phobic and a "normal" population. Ned Tijdschr Psychol Grensgebieden 36 2 (1981), pp. 77-108. 15. Van-Sonderen E. Sociale Steun Lijst-Interacties (SSL-I) en Sociale Steun Lijst-Discrepanties (SSL-D). 1993. Report. Groningen: Noordelijk Centrum voor Gezondheidsvraagstukken. 16. JHMM Vercoulen, E Bazelmans, CMA Swanink, JFM Fennis, JMD Galama, PJH Jongen, OR Hommes, JWM van-der-Meer and G Bleijenberg, Physical activity in chronic fatigue syndrome: assessment and its role in fatigue. J Psychiatr Res 31 6 (1997), pp. 661-673. 17. JHMM Vercoulen, E Bazelmans, CMA Swanink, JMD Galama, JFM Fennis, JWM van-der-Meer and G Bleijenberg, Evaluating neuropsychological impairment in chronic fatigue syndrome. J Clin Exp Neuropsychol 20 2 (1998), pp. 144-156. -------- (c) 2002 ScienceDirect / Elsevier Science B.V.