Subtypes 375studies involves issues related to sampling and participant se-
lection. A majority of investigations have employed nonran-
dom, medically referred samples. Different types of illnesses
are probably now contained within the CFS construct, which
makes it even more dif“cult to identify commonalities in all
people with this diagnosis. As an example, See and Tilles
(1996) found that, with alpha interferon treatment, quality-
of-life scores improved signi“cantly only for those with
NK cell dysfunction, suggesting that this subgroup of CFS
patients might have bene“ted from the enhancement of
cytokine production that was produced by the interferon-
stimulated NK cells and that this might have led to restora-
tion of normal immune function.
Findings from many empirical investigations of CFS sug-
gest that subtypes of patients can be distinguished with re-
spect to the mode of illness onset (whether gradual or sudden;
DeLuca, Johnson, Ellis, & Natelson, 1997; Komaroff, 1988,
1994; Levine, 1997; Reyes et al., 1999). However, contro-
versy exists as to whether prognosis of individuals with CFS
is affected by the experience of sudden versus gradual onset.
Reyes and associates (1999) examined symptoms experi-
enced at illness onset for individuals with either sudden or
gradual onset of CFS and found that those with sudden onset
reported signi“cantly more symptoms at onset than those
with gradual onset, and symptoms were more likely to be of
infectious nature, including fever, sore throat, chills, and ten-
der lymph nodes. This is consistent with other research
(Komaroff, 1988, 1994), which has suggested that sudden
onset of CFS may be indicative of viral/infectious illness.
Over time, however, symptom patterns among individuals in
the Reyes et al. (1999) sample became more similar for those
with sudden and gradual onset, and probability of recovery
was not affected by mode of onset. In support of this “nding,
an examination of the ”uctuation of symptoms and outcome
of CFS over time (Hill, Tiersky, Scavalla, & Natelson, 1999)
found that mode of illness onset was not predictive of positive
or negative illness outcomes. Contrary to these “ndings, how-
ever, Levine (1997) has found that individuals with sudden
onset have a better prognosis than those with gradual onset.
In a random community sample, Jason, Taylor, Kennedy,
Song, et al. (2000) found that individuals with sudden CFS
onset were signi“cantly more likely to experience more se-
vere sore throat pain and more severe fatigue following exer-
cise. One possible explanation for increased severity of these
two symptoms in the sudden onset group may involve an in-
creased likelihood of viral etiology in this subgroup of partic-
ipants. Individuals with sudden onset were also more likely
to experience lifetime psychiatric diagnosis.
The presence of a stressful life event preceding or
precipitating onset of CFS is another factor that has been
investigated and may differentiate subgroups of patients with
CFS (Ray, Jeffries, & Weir, 1995; Salit, 1997; Theorell,
Blomkvist, Lindh, & Evangard, 1999). Some evidence indi-
cates that individuals with CFS have experienced a higher
frequency of negative life events in the time directly preced-
ing the onset than matched controls (Salit, 1997; Theorell
et al., 1999). In a community-based sample, Jason, Taylor,
Kennedy, Song, and associates (2000) found that individuals
who were experiencing unusually severe stress at the time
of CFS onset reported lower levels of vitality and lower
emotional role functioning. Taylor and Jason (2001) found
prevalence rates of sexual and physical abuse history among
individuals with CFS were comparable to those found in indi-
viduals with other conditions involving chronic fatigue, in-
cluding medically based conditions. In addition, relative to
those with CFS who report such history, most individuals
with CFS did not report histories of interpersonal abuse.
Other researchers (Ray et al., 1995) have found that the pres-
ence of positive life events causing moderate or major life
change are associated with lower fatigue and impairment
scores in individuals with CFS, while negative life events do
not impact these outcomes.
Findings from a number of empirical investigations of
CFS conducted in England, the United States, and Australia
suggest that two subtypes of patients can be distinguished in
terms of symptom severity, functional level, and psychiatric
status (Friedberg & Jason, 1998; Hickie et al., 1995; Manu,
Lane, & Matthews, 1988). In a sample of patients with CFS
analyzed by Hickie and associates (1995), distinctive sub-
types emerged: (a) a •somatization-likeŽ group, including
those who have a higher prevalence of CFS symptoms
and atypical symptoms, greater disability attributed to CFS
and psychiatric symptoms, and a greater percentage unem-
ployed and; (b) a •CFSŽ group, including those individuals
with lower prevalence of CFS and atypical symptoms, less
disability attributed to CFS and psychiatric symptoms, and a
greater percentage employed.
In a U.S. sample, Manu and associates (1988) found a bi-
modal distribution of symptoms among 100 chronic fatigue
patients, including 21 patients with 10 to 15 symptoms and
79 patients with 0 to 9 symptoms. In two follow-up studies of
patients with CFS, persistent symptoms and disability at the
follow-up were associated with eight or more medically
unexplained symptoms at time one (Bombardier & Buch-
wald, 1995; Clark et al., 1995). A comparison group of non-
CFS chronic fatigue patients exhibited fewer symptoms and
higher functioning (Bombardier & Buchwald, 1995).
Lange et al. (1999) found no MRI differences between
those with CFS and healthy controls; however, when the CFS
group was divided into those with and without a psychiatric