Prevalence of CFS
Primary care studies
The study by David et al. [16] discussed earlier found
only one case of a CFS-like illness in a sample of 611
primary care attendees. A later study by the same group
[19] used a 2-stage procedure. They screened 686
patients attending primary care physicians and identified
77 individuals with chronic fatigue. Of these, 65 underwent
a comprehensive medical and psychological evaluation.
Twenty-six percent met the British criteria for
CFS. Many of these studies had small sample sizes and
inappropriate sampling procedures.
In one of a series of London-based studies, Wessely
et al. [20] conducted a prospective study to determine the
prevalence of CFS in 1199 primary care attendees with
symptoms of common infection and 2267 seen for other
reasons. They managed a follow-up rate of 84% at 6
months. At 6 months, 9.9% of the infectious group and
11.7% of the control group reported chronic fatigue and
there was no difference in the proportions of the case
groups fulfilling criteria for CFS. The study also looked at
the prevalence of various definitions of CFS. The overall
prevalence of CDC-1994, Oxford, Australian and CDC-
1988 criteria were 2.6, 2.2, 1.4 and 1.2%, respectively;
these values fell further when co-morbid psychiatric
disorders were excluded. This compares with an American
primary care study which found rates of 0.3, 0.4 and
1% for CDC-1988, British and Australian definitions,
respectively [21].
Community-based studies
Price et al. [22] used the Diagnostic Interview Schedule
(DIS) to assess the presence of CFS according to the
CDC-1988 definition. While 23% of the participants
reported experiencing persistent fatigue sometime during
their lives, only 1 of 13 538 peoplemet a diagnosis of CFS
fulfilling all exclusion and inclusion criteria. This gave a
prevalence figure of 7/100 000. The study of Price et al.
[22], though, has numerous methodological problems.
The definition of CFS did not correspond to other
definitions. The other problem was that symptoms were
assessed on a lifetime basis, leading to the possibility that
cases where medical or psychiatric conditions existed at a
different time may have been excluded. The lack of
medical work-up or laboratory analyses was also a failing.
The CDC conducted a community-based survey in
San Fransisco [23]. Using telephone interviews, 14 627
adults were studied of whom 4.3% reported chronic
fatigue. Having excluded cases with medical or psychiatric
diagnoses that could potentially explain fatigue, the
investigators identified 1.8% of the population as having
idiopathic chronic fatigue and 0.2% as having a CFS-like
illness.
Jason et al. [24] studied the point prevalence of CFS in
an ethically diverse random community sample. The
study was a methodologically sophisticated one with a
2-stage procedure. In the first stage, the patients were
screened using a questionnaire and those who screened
positive for a CFS-like illness were interviewed on the
Structured Clinical Interview for the Diagnostic and
Statistical Manual (SCID) in the second stage. This was
followed by a detailed medical examination and investigations.
The final diagnosis of CFS was made by two
independent raters. The study found a prevalence of
0.42% in the sample.
The wide disparity in the figures for prevalence of CFS
in various studies (see Table 2) is likely to be due to
differences in methodology rather than a true difference
in prevalance. It seems clear that random communitybased
samples, an extensive search for a medical
abnormality and restrictive case definitions result in
lower prevalence figures. It is noteworthy that a recent
Australian study found a prevalence rate of 1.5% in the
population for neurasthenia, a related disorder [25].