ME/CFS Primer/Education Bulletins Severe ME/CFS in Adults - A report from the CHROME Database-Full Article
IACFS/ME Bulletin

 

Severe ME/CFS in Adults - A report from the CHROME Database

Derek Pheby1*, Peter Sneddon2, Inge Heinrich3

1Derek Pheby, BSc, MB,BS, LLM, MPhil, FFPH, Project Coordinator, National ME Observatory Visiting Professor, Faculty of Society and Health, Buckinghamshire New University, 106, Oxford Road, Uxbridge, Middlesex, UB8 1NA, UK
Tel.: +44 (0)1722 331283
e-mail : derekpheby@btinternet.com

2Dr. Peter Sneddon, M.A., PhD. School of Health and Social Sciences, Middlesex University and CHROME Research Group
7 Wilderton Road, London, N16 5QY, UK
e-mail : Peter.Sneddon@btinternet.com

3Dr. Inge Heinrich, MSc, PhD, Medical Statistician (retired),
“Thurne”, Gore Road, Eastry, Sandwich, Kent, CT13 0LP, UK. e-mail : inge.heinrich@virgin.net


ABSTRACT

BACKGROUND:    Case History Research on ME (CHROME) was established in 1994 to undertake research on severely affected patients. Since 1995 CHROME has collected data on volunteers with severe ME/CFS. All are medically diagnosed, conform to Fukuda criteria, are or were house-bound, and ill for at least two years. Participants complete initial and annual follow-up questionnaires, and report their condition at onset of illness, a year previously, on recruitment, and annually thereafter.

METHOD:    We report symptom severity and ability to undertake activities of daily living in 324 participants recruited between 1995 and 1998, their evolution through time, and, for later onset (age 17+) subjects, changes occurring between recruitment and the latest questionnaire. Significance was assessed using the McNemar test or its binomial equivalent. RESULTS:    For most features, subjects deteriorated between onset and recruitment, with subsequent improvement, often slow and uneven, among later onset cases. Complete recovery was unusual. There were marked individual variations, and many subjects remained severely incapacitated or deteriorated. All features examined were more prevalent and severe in patients with a younger age of onset. Age at onset was bimodal, with peaks at 11-20 and 31-40. Most subjects, at recruitment, had been ill for 2-10 years. The proportion of early onset cases rose significantly in the mid-1980s.

CONCLUSIONS:    Caution is necessary in interpreting the results, given problems of self- selection, recall bias, and subjective interpretations of symptom ratings. It is unclear whether the reported trends are inevitable features of ME/CFS, or perhaps due to poor management at the outset.


Word count:    Abstract:    250 words    Main text:    3,340 words


Key Words: Chronic fatigue syndrome, epidemiology, prognosis, severity, cognitive ability, sensorimotor disturbance.


INTRODUCTION

There have been few studies of severely ill people with ME/CFS, who constitute a substantial proportion of those with this illness. ME/CFS varies widely in severity and duration. In most subjects it is relatively mild and short lived, but in some it is both highly incapacitating and painful as well as capable of lasting for many years. The care needs of such people are poorly understood, due to lack of research (1). Most papers written on ME/CFS have focused on those less severely affected, having been carried out on ambulant patients able to attend clinics.

Case History Research on ME (CHROME) was established in 1994 to undertake research on more severely affected patients, who were defined as people with ME/CFS who are, or have been, housebound or bedbound, and unable to leave home without assistance. It was established with the intention of monitoring participants’ illness course for ten years, and of developing a database for prospective studies on people with severe ME. Participants are self-selected, and while they cannot be considered representative of the population with ME/CFS, they are a large group who manifest its clinical features in an extreme form. Inclusion criteria for incorporation in the CHROME database are a medically confirmed diagnosis of ME/CFS (diagnosed by a registered medical practitioner, either a consultant or a primary care physician), illness lasting for more than two years, conformity to the Fukuda case definition (2), and severe disease as defined above. In the initial questionnaire, participants were asked to indicate which category of medical practitioner had made the diagnosis, and if possible his or her name and clinical specialty.

CHROME is an initiative arising from within the ME/CFS community and thus reports those issues seen as significant by people with ME/CFS. Construction of the questionnaires was largely on the basis of the personal experiences of people with ME/CFS, rather than on that of the pre-existing research literature, though it should be appreciated that, at the time of inception of the project (1993-94), there was very little relevant epidemiological research literature on ME/CFS to draw on. In fact, we could only find two epidemiological reports prior to the start of the CHROME project that made reference to severity (3,4), and in neither of these was severity a major focus of the research.

An Australian epidemiological study (5) reported that 43% of people with CFS were unable to attend school or work, and therefore had a more severe form of ME than those who could. Another report from the same source (6) suggested that two distinct subclasses, of more and less severely affected patients, could be identified, comprising 27% and 73% of subjects respectively. A recent member survey (7) showed that 89% of respondents claimed to have been bed-ridden or house-bound at some stage, and that 30% were still severely affected. There is likely to have been both selection and response bias, as the response rate of 28% was very low. The fact that 81% of respondents were female also suggested bias, as although most epidemiological studies have confirmed a preponderance of females among adults with ME/CFS, this proportion appears higher than among people with ME/CFS of all degrees of severity, and a 1991 GP study (4) revealed that illness severity was the same in males and females.

There have been more studies of risk factors for poor prognosis than for severity, though they are linked as prognosis for severely ill patients is worse than for the less severely ill, in both children (8) and adults (9,10). Social, psychological and physical factors have all been shown to be associated with severe illness and poor prognosis, e.g. not having a solicitous ‘significant other’ (11), not being married (12), having cognitive difficulties (13), attributing complaints to a somatic cause (14), behavioural disengagement (15), accommodating to the illness (16), fatigue, somatic symptoms and cognitive difficulty (17), clarity of thinking and fewer symptoms deemed to be somatoform in nature (12). Factors associated with severe ME in children include specific viral triggers, start date and socio-economic status (8). In adults, comorbidity can predispose to severe illness (18), as also can exercise prior to falling ill (19,20). A role for familial factors is suggested by a twin study (21) and a study of an epidemic outbreak in New York state (22).

Most of these studies are uncorroborated, and mostly what has been demonstrated is association, not causation. For example, the various psychological factors claimed to be causal may be a consequence of severe ME, and for the most part study designs do not enable causality to be determined. Additionally, definitions of severity have varied widely, and may be vague (23). There have been few attempts to quantify severity in ways which are reproducible, using validated instruments, and there have been problems when they have been used. Buchwald et al, for example, explored the use of SF-36 in chronic fatigue syndrome, but found that it was a poor discriminator between chronic fatigue and chronic fatigue syndrome (24). This may reflect its multiaxial nature (25), consisting of eight dimensions, not all of which are related to levels of disability. Other studies have concentrated on developing validated instruments to measure fatigue (26), and one of these, the Chalder Fatigue Scale (27), is now widely used in the UK. However, fatigue is only one of the features of ME/CFS, and it is not clear to what extent the Chalder Fatigue Scale is effective in identifying people with ME/CFS whose disability may for example cause them to experience difficulty in activities of daily living.

The aim of this study was to report on clinical features in 324 participants recruited to the CHROME database from 1995, and, for later onset (age 17+) subjects, to report on changes following recruitment.

METHODS

CHROME has recruited volunteer subjects from 1995. The project is a longitudinal cohort study, in which it was originally intended to continue recruitment and follow-up for ten years. It has a large database of information on people who have had severe ME/CFS for two years or more.    One of the central objectives of the project was to have gathered enough data after ten years to be able to identify and report any prognostic factors relating to the likelihood and speed of recovery.

Each participant completed an initial questionnaire which described in detail how they were at the initial onset of the illness, how they were a year previous to that, and how they were at the point of completion of the questionnaire. These three points in time are referred to subsequently in this report as onset, baseline and recruitment respectively. Questionnaires were completed by carers in some cases, and they were asked to reflect faithfully the subject’s responses. However, we do not know to how many subjects this pertained, nor the extent of the assistance carers gave, as this information was not collected.

Participants are sent an annual follow up questionnaire. Both questionnaires were designed in-house to reflect specific areas of concern regarding people with severe ME/CFS, rather than making use of pre-existing validated instruments, which, as indicated above, have not proved particularly useful. With over three hundred participants, the database is large and evolving. The number of follow-up questionnaires completed and returned by participants ranged from 1 (11.1% of participants) to 10 (15.1%) or 11 (0.6), with an average of 5.53. More than half the respondents (56.5%) had returned at least 5 annual follow-up questionnaires. The detailed response, including the cumulative response, is summarised in Table 1 and Figure 1.

Our earlier study (28) focused on subjects with early onset disease (i.e. before age seventeen). The only comparison over time possible then was between reported symptoms and disabilities at onset and recruitment. In this paper, we report symptoms reported by 324 participants recruited between 1995 and 1998, and their evolution through time. For later onset (age 17+) subjects (245 in number), we report changes occurring between recruitment and the latest questionnaire. In addition, the details of the reported symptoms for a year before recruitment and for the first follow up questionnaire have been included.

As this was a subject-led project, rather than one initiated by researchers, the project organisers did not seek ethical approval, nor indeed would it have been appropriate for them to have done so. All the participants were volunteers. Consent to their participation was sought in every case, for the most part via telephone discussions, and in no case was it refused. Since participation was entirely optional, the fact of completing and submitting the initial and follow-up questionnaires could be taken to indicate implicit consent.

Three or four point scales were used to indicate severity. Examples are listed in Table 3. For analysis, categories were grouped into less severe and more severe respectively. This follows the procedure adopted in our earlier study (28), and focuses attention on those who are most incapacitated by the illness.

Symptoms Studied
The questionnaires cover numerous symptoms. A comprehensive internal CHROME document prepared by Koliadin (29) with Heinrich details the overall results. This paper reports the variables studied by Gibbons et al, viz. ability to memorise and pass on a message, ability to walk to the lavatory, in prolonged standing, arm raising, obtaining refreshing sleep, muscle pains in lower limbs, and nausea.

Statistical Analysis
New participants are constantly being recruited, while others drop out or die. Therefore the samples at the three different time points are not identical. However the statistical tests of change, the McNemar test or its binomial equivalent, are repeated measures tests of direction of change for dichotomous data, and so the same participants are being compared over time with themselves. Any differences found can therefore not be the consequence of different samples being used.

It should be noted that, because seven variables have been studied, significant results may have been obtained by chance. Accordingly, following the Bonferroni procedure, we have determined a more appropriate level for statistical significance of P<0.007.

The CHROME database is kept in Microsoft Access. SPSS for Windows, Version 10, was used for the statistical analysis. To transfer the data from the ACCESS format to an SPSS one, programs were written in Microsoft Visual Basic, Version 3.

Changes in clinical features between onset and recruitment were assessed using the McNemar test, for participants with early (i.e. age below 17) and later onset disease respectively. To compare the frequency of severe problems at recruitment in participants in these two groups, the Chi Square test (with continuity correction) was used. The change between recruitment and the latest questionnaire completed was also examined, using the binomial equivalent of the McNemar test.

RESULTS

Age at Onset
The distribution of age at onset is bimodal, with one peak for the 11-20 age range, and a second one at age 31 to 40 (figure 2). This applies irrespective of year of recruitment. All clinical features examined were more prevalent in severe form at recruitment in subjects with early onset disease than in those who reported a later onset. This difference was statistically significant in respect of ability to walk to the lavatory, and prolonged standing.

Year of Onset
Year of onset is summarised in Table 2, which also shows the proportions by year of early and late onset cases, and 95% confidence intervals for the proportions of early and late onset cases It should be noted that, though there is considerable variation from one period of onset to another, the number of cases arising within any one time period is relatively small, and 95% confidence intervals are wide. However, there was clearly a change in the proportion of early onset cases from the mid 1980s onwards. Among cases originating before 1987, early onset cases comprise 17.8% of all cases (19 out of 107). Among cases arising from 1987 onwards, early onset cases constitute 42.4% of the total (59 out of 139). This was highly significant (Chi2, with Yates’ correction = 8.339; p = 0.004).

Figure 3 shows a peak year of onset for early onset cases of 1994, while for late onset cases the peak years were 1987 and 1992. It is not clear whether this is a recruitment artefact, or represents the impact of some environmental exposure. For most of the new participants, onset was within the last decade. The majority of subjects, at recruitment, had been ill for between two and ten years.

Severity of Symptoms
Table 3 displays the proportions of participants reporting various degrees of severity for selected symptoms, and of difficulty performing specific tasks, at onset, at recruitment (baseline), and in the latest questionnaire. For every symptom except nausea, the proportion of participants reporting that it was severe increased from onset to recruitment, but decreased again by the most recent follow-up, though not to the level at onset. Muscle pain, for example, followed this characteristic pattern of increase in symptom severity until recruitment, but some improvement thereafter. Changes in both time periods were highly significant (P<0.001). At recruitment some 89% of the participants suffered from some muscle pain in their lower limbs, and even in the most recent follow-up questionnaires, some 49% of participants still report this (Table 4). Again, there was marked individual variation, with 29 out of 227 participants improving from "onset" to "recruitment".

As regards nausea, most participants, both early and later onset, suffered to a moderate or severe extent. Unlike other symptoms, there was no obvious deterioration from "onset" to "recruitment", and in fact the number of severe cases decreased from 35.4% to 24.5% (P=0.006), but there was no clear improvement thereafter.

Activities of Daily Living
A similar pattern was found in respect of ability to perform tasks associated with daily living. All the clinical features examined were more prevalent in severe form at recruitment in subjects with early onset disease than in those who reported a later onset. This difference was statistically significant in respect of ability to walk to the lavatory. There did not appear to be any significant difference in the course of symptom severity by year of onset.

The proportion of subjects reporting that performance of tasks was extremely difficult or
impossible tended to increase from onset to recruitment, and then to decrease by the latest
follow-up questionnaire, though not to the level reported at onset. There was significant worsening between onset and recruitment among both early and late onset cases in ability to walk to the lavatory and prolonged standing (see table 4). Among the later onset cases, there was also a marked tendency to improvement between the baseline and latest questionnaire results. This was significant in respect of all tasks examined except ability to pass on a message and ability to walk to the lavatory.

As shown in Tables 2 and 3 the prevalence of difficulty walking to the lavatory increased from onset to a year before recruitment. It then stabilised and thereafter slowly improved. The deterioration between onset and recruitment was not statistically significant (P=0.048) given our revised α- threshold of P<0.007, and neither was the improvement from recruitment to the most recent follow-up (P=0.009). Again, there were marked individual variations from this pattern. 62 of 164 participants who reported no severe difficulties at “onset” considered that they were present at “recruitment”,  but 41 of 66 participants who were severely disabled at “onset” had a non-severe rating at “recruitment” (P=0.048). Between "recruitment" and "latest", though, twenty patients moved from the Severe to the Not Severe categories while only 6 moved in the opposite direction. Despite this, the majority of the participants still reported difficulty in walking to the lavatory in the latest follow-up questionnaires.

Both prolonged standing (to iron, cook etc.), and raising arms above head (to comb hair, etc.) show the same pattern of dramatic deterioration from the onset of the illness and then some gradual improvement. Arm raising, while causing problems for most participants, was considerably less debilitating than the inability to stand for prolonged periods.

Problems with standing were even more widespread than problems with walking, or lower limb muscle pain. At recruitment, 94% had difficulty with standing, 80% finding it extremely difficult or impossible. Even though there was significant improvement since then (P<0.002), the latest questionnaire data indicate that 62% of participants still have extreme difficulties.

Cognitive Impairment
An indicator of cognitive impairment was whether or not participants were able to memorise and pass on a message correctly. Of the early onset group, 28% found this extremely difficult or impossible at onset, compared with 26.5% of the participants with later onset. Both groups show a deterioration, which was not significant, from onset to recruitment, the above percentages rising to 38% and 33.1% in the early and later onset groups respectively, with a slight but not significant improvement on follow-up. The overall trend conceals marked individual variations. Thus, of 61 people who found it extremely difficult or impossible to pass on a message, 28 had a less severe problem by recruitment, but 43 of the 169 participants who did not have a severe problem at onset did have one by recruitment.

Sleep
Another frequent problem was difficulty in obtaining refreshing sleep. Again, this was worse at recruitment than at onset, though not significantly so. The gradual improvement thereafter, to the latest follow-up, was significant (P<0.001), but left most participants finding it difficult or extremely difficult to obtain restorative sleep.

DISCUSSION

Like our previous report (28), this paper explores a topic where little research has been carried out to date, viz. the chronological progression of severe ME/CFS. This study demonstrates that, while most symptoms tend to improve after an initial period of deterioration, the improvement is slow and uneven, with many sufferers staying at the same level of severe incapacity, and a small minority continuing to deteriorate. However, caution is necessary in interpreting the results, in particular in extrapolating to the overall population with severe ME/CFS, given that the participants were self-selected, and that recollection of symptoms at onset may involve recall bias. In addition, the scales used were capable of subjective interpretation, and may not therefore have had the same meaning to all participants. We are unable at present, on the basis of available data, to determine either the reliability or the validity of the data. It should be noted that the time period over which participants were asked to recall their symptoms at onset varied very much from participant to participant, affecting the likelihood of recall bias.

These make cross-sectional comparisons between groups problematic. Longitudinal comparisons, where each participant is acting as their own control, reduce this problem of interpretation. The binomial equivalent of the McNemar test for correlated proportions is appropriate for such within-group comparisons. Additionally, the size of the cohort creates the possibility of important trends not appearing significant (type II error).
Among the participants in our study, age at onset was bimodal, with peaks at ages 11-20 and 31 to 40, with a higher prevalence of severe clinical features at recruitment among those with earlier onset. Cases arising from 1987 onwards included a much higher proportion of early onset cases than was apparent among cases originating earlier. Among the group of participants as a whole, there was a consistent trend of deterioration from onset to recruitment, with a tendency, among later onset cases, to improvement thereafter. This was true of most symptoms, performance of activities of daily living, cognitive function, physical disability, and sleep, but there were also substantial individual variations from this overall pattern.

A recent systematic review has identified some of the problems of investigating prognosis in people with chronic fatigue. It concluded, from its review of seventeen cohort studies, that most were methodologically flawed due to bias caused by low response rates or loss to follow-up, small sample size, and long symptom duration at recruitment, which made it difficult to study the prognosis of early fatigue (28). A major strength of our current study was that it avoided the problems of low response rate, loss to follow-up and small sample size. A weakness, though, was long symptom duration at recruitment, which could render the study liable to be affected by recall bias. The deterioration noted between onset and recruitment could be at least to some extent artefactual, since being housebound or bedbound was an inclusion criterion for participation in the study. Another weakness was that perceptions of symptom severity were subjective. Nevertheless, the study does confirm some important observations about the course of the illness, including the persistent, long-drawn-out nature of severe ME, the tendency to recovery, the fact that recovery is rarely complete, and that there are marked individual variations. It is unclear whether the reported trends are inevitable features of the illness, or due to factors such as poor management in the initial stages.

AUTHORS' CONTRIBUTIONS

The statistical analysis was largely carried out by Peter Seddon, assisted by Derek Pheby. All authors contributed to planning the study. This report was largely prepared by Derek Pheby, with considerable input also from Peter Seddon and Inge Heinrich.

ACKNOWLEDGEMENTS

Thanks are due to CHROME, in particular to Ray Gibbons and Chris Richards, for maintenance of and access to the database, to the CHROME Research Group, especially Rachel Barker, Catherine Manley, Ken Manley, Colin Parratt and Emma Toy, to Victor Koliadin for his initial detailed overview of the data, and to all the participants in the CHROME project for contributing their data and supporting the initiative.

REFERENCES

1)    ME Research UK Research Database. Available at: http://meresearch.org.uk/information/researchdbase/index.html. Accessed 4th. December, 2009.
2)    Fukuda K, Straus SE, Hickie I, et al. The chronic fatigue syndrome: a comprehensive approach to its definition and study, International Chronic Fatigue Syndrome Study Group. Annals of Internal Medicine 1994; 121:953-959.
3)    McDonald E, David AS, Pelosi AJ, Mann AH. Chronic fatigue in primary care attenders (1993). Psychol Med 1993; 23(4):987-98
4)    Ho-Yen DO, McNamara I. General practitioners' experience of the chronic fatigue syndrome. British Journal of General Practice 1991; 41:324-6.
5)    Lloyd AR, Hickie I, Boughton CR, et al. Prevalence of chronic fatigue syndrome in an Australian population. Medical Journal of Australia 1990; 153(9):522-8.
6)    Hickie I, Lloyd A, Hadzi-Pavlovic D, et al. Can the chronic fatigue syndrome be defined by distinct clinical features?. Psychological Medicine 1995; 25(5):925-35.
7)    Action for ME. Severely neglected: ME in the UK. London. March 2001.

8)    Rangel L, Garralda ME, Levin M, Roberts H. The course of severe chronic fatigue
syndrome in childhood. J. Roy. Soc.Med. 2000; 93(3):129-34.
9)    Pheley AM, Melby D, Schenck C, et al. Can we predict recovery in chronic fatigue
syndrome?. Minnesota Medicine 1999; 82(11):52-6.
10)    Hill NF, Tiersky LA, Scavalla VR, et al. Natural history of severe chronic fatigue
syndrome. Archives of Physical Medicine & Rehabilitation 1999; 80(9):1090-4.
11)    Schmaling KB, Smith WR, Buchwald DS. Significant other responses are associated with fatigue and functional status among patients with chronic fatigue syndrome.
Psychosomatic Medicine 2000; 62(3):444-50.
12)    Hartz AJ, Kuhn EM, Bentler SE, et al. Prognostic factors for persons with idiopathic
chronic fatigue. Archives of Family Medicine 1999; 8(6):495-501.
13)    Friedberg F, Dechene L, McKenzie MJ, Fontanetta R. Symptom patterns in long- duration chronic fatigue syndrome. Journal of Psychosomatic Research 2000; 48(1):59-68.
14)    Vercoulen JH, Swanink CM, Galama JM, et al. The persistence of fatigue in chronic fatigue syndrome and multiple sclerosis: development of a model. Journal of Psychosomatic Research 1998; 45(6):507-17.
15)    Ray C, Jefferies S, Weir WR. Coping and other predictors of outcome in chronic fatigue syndrome: a 1-year follow-up. Journal of Psychosomatic Research 1997; 43(4):405-15.
16)    Ray C, Jefferies S, Weir WR. Coping with chronic fatigue syndrome: illness responses and their relationship with fatigue, functional impairment and emotional status. Psychological Medicine 1995; 25(5):937-45.
17)    Ray C, Weir WR, Cullen S, Phillips S. Illness perception and symptom components in chronic fatigue syndrome. Journal of Psychosomatic Research 1992; 36(3):243-56.
18)    White KP, Speechley M, Harth M, Ostbye T. Co-existence of chronic fatigue syndrome with fibromyalgia syndrome in the general population. A controlled study. Scandinavian Journal of Rheumatology 2000; 29(1):44-51.
19)    MacDonald KL, Osterholm MT, LeDell KH, et al. A case-control study to assess possible triggers and cofactors in chronic fatigue syndrome. American Journal of Medicine 1996; 100(5):548-54.
20)    Van Houdenhove B, Onghena P, Neerinckx E, Hellin J. Does high 'action-proneness' make people more vulnerable to chronic fatigue syndrome? A controlled psychometric study. Journal of Psychosomatic Research 1995; 39(5):33-40.
21)    Farmer A, Scourfield J, Martin N, et al. Is disabling fatigue in childhood influenced by genes? Psychological Medicine 1999; 29(2):279-82.
22)    Bell KM, Cookfair D, Bell DS, et al. Risk factors associated with chronic fatigue syndrome in a cluster of pediatric cases. Reviews of Infectious Diseases 1991; 13 Suppl 1:S32-38.
23)    Cox DL, Findley LJ. Severe and very severe patients with chronic fatigue syndrome perceived outcome following an inpatient programme. Journal of Chronic Fatigue Syndrome 2000; 7(3):33-47.
24)    Buchwald D, Pearlman T, Umali J, et al. Functional status in patients with chronic fatigue syndrome, other fatiguing illnesses, and healthy individuals. American Journal of Medicine 1996; 101(4):364-70.
25)    Anon. Comparing the Generic SF Surveys. Quality Metric. 2005. Available at: www.qualitymetric.com/products/comparesfsurveys.aspx. Accessed February 19, 2009.
26)    Schwartz JE, Jandorf L, Krupp LB. (1993). The measurement of fatigue: a new instrument. Journal of Psychosomatic Research 1993; 37(7):753-62.
27)    Chalder T, Berelowitz G, Pawlikowska T, et al. Development of a fatigue scale. Journal of Psychosomatic Research 1993; 37(2):147-53.
28)    Gibbons R, Pheby D, Richards C, Bray F. Severe ME/CFS of Juvenile Onset. Journal of the Chronic Fatigue Syndrome 1998; 4(4):67-80.
29)    Koliadin V. (A report on the CHROME Database.) Internal report - CHROME Project. 2000.
30)    Nijrolder I, van der Horst H, van der Windt D. Prognosis of fatigue. A systematic review. Journal of Psychosomatic Research 2008; 64(4):335-49.


TABLE 1:

TABLE 2:  

.

TABLE 3:



TABLE 4:



FIGURE 1:



FIGURE 2:

FIGURE 3:
 


Bulletin of the IACFS/ME. 2009-10;17(4):143-167. © 2010 IACFS/ME

To download this paper as a PDF file, click here.

(You will need Adobe Reader to open this file. If you do not have Adobe Reader, click here to download.)

Return to Bulletin of the IACFS/ME, Volume 17, Issue 4