A Twin Study of Chronic Fatigue
DEDRA BUCHWALD, MD, RICHARD HERRELL, PHD, SUZANNE ASHTON, BS, MEGAN BELCOURT, BS,
KAREN SCHMALING, PHD, PATRICK SULLIVAN, MD, MICHAEL NEALE, PHD, AND JACK GOLDBERG, PHD
Objective: The etiology of chronic fatigue syndrome is unknown, but genetic influences may be important in its
expression. Our objective was to assess the role of genetic and environmental factors in unexplained chronic fatigue.
Methods: A classic twin study was conducted using 146 female-female twin pairs, of whom at least one member
reported ?6 months of fatigue. After completing questionnaires on symptoms, zygosity, physical health, and a
psychiatric interview, twins were classified using three increasingly stringent definitions: 1) chronic fatigue for ?6
months, 2) chronic fatigue not explained by exclusionary medical conditions, and 3) idiopathic chronic fatigue not
explained by medical or psychiatric exclusionary criteria of the chronic fatigue syndrome case definition. Concor-
dance rates in monozygotic and dizygotic twins were calculated for each fatigue definition along with estimates of
the relative magnitude of genetic and environmental influences on chronic fatigue. Results: The concordance rate
was higher in monozygotic than dizygotic twins for each definition of chronic fatigue. For idiopathic chronic
fatigue, the concordance rates were 55% in monozygotic and 19% in dizygotic twins (p ? .042). The estimated
heritability in liability was 19% (95% confidence interval ? 0–56) for chronic fatigue ?6 months, 30% (95%
confidence interval ? 0–81) for chronic fatigue not explained by medical conditions, and 51% (95% confidence
interval ? 7–96) for idiopathic chronic fatigue. Conclusions: These results provide evidence supporting the familial
aggregation of fatigue and suggest that genes may play a role in the etiology of chronic fatigue syndrome. Key words:
chronic fatigue, twins, concordance, genetics.
CDC ? Centers for Disease Control and Prevention;
CFS ? chronic fatigue syndrome; CI ? confidence
interval; DIS ? Diagnostic Interview Schedule; DZ ?
dizygotic; HLA ? human leukocyte antigen; MZ ?
Fatigue is a common complaint in primary care
settings (1–5), being reported by at least 20% of pa-
tients seeking care. Likewise, large community surveys
indicate that up to half of the general population ex-
periences fatigue, usually of limited duration (6–8). In
most cases the fatigue is transient, explained by pre-
vailing circumstances, relieved by rest, and of little
cause for concern. In both community and clinical
settings, fatigue is typically more frequent among
women than men. However, persistent and debilitat-
ing fatigue is the hallmark of chronic fatigue syndrome
(CFS), an illness characterized by disabling fatigue
associated with muscle pain, pharyngitis, and alter-
ations in mood, sleep, and neurocognition (9).
Many theories on the pathophysiology of CFS have
been proposed (10). Initially the prominence of infec-
tious, neurocognitive, and psychological symptoms
suggested a viral illness or psychiatric disorder (11,
12). Subsequently a variety of findings related to sleep
pathology and neuroendocrine, immunological, and
autonomic dysfunction were observed in subgroups of
patients with CFS (13–16). At the same time, our group
(anecdotal observations) and others (17) have noted
that CFS and other unexplained chronically fatiguing
conditions often affect several members of the same
Twin studies have been useful in elucidating the
relative contributions of genetic and environmental
factors to numerous medical and psychiatric disor-
ders, and they are especially helpful for the study of
diseases of unknown cause (18). Recently investigators
demonstrated a genetic influence on an index of gen-
eral fatigue in a sample of Australian twins (19). With
this in mind, we undertook a classical twin study to
investigate the possible genetic and nongenetic influ-
ences on CFS. In this study we examined the familial
clustering of fatigue by comparing concordance rates
in monozygotic (MZ) and dizygotic (DZ) female-female
twins for three definitions of fatigue and developed
initial estimates of the heritability of chronic fatigue.
Identification of Twins and Data Collection
Twins were recruited through a variety of sources, including
advertisements in patient support group newsletters (63%), clini-
cians and researchers familiar with CFS (12%), solicitations placed
on CFS electronic bulletin boards (8%), twin researchers or organi-
zations (5%), friends and relatives (3%), and various other means
(9%). All recruitment efforts emphasized that fatigued twins were
From the Departments of Medicine (D.B., S.A., M.B.) and Psychi-
atry and Behavioral Sciences (K.S.), University of Washington, Se-
attle, WA; the Division of Epidemiology–Biostatistics (J.G., R.H.),
University of Illinois, Chicago, IL; and the Virginia Institute for
Psychiatric and Behavioral Genetics (P.S., M.N.), Virginia Common-
wealth University, Richmond, VA.
Address reprint requests to: Dedra Buchwald, MD, Harborview
Medical Center, 325 9th Ave., Box 359780, Seattle, WA 98104.
Received for publication May 22, 2000; revision received April 9,
936Psychosomatic Medicine 63:936–943 (2001)
Copyright © 2001 by the American Psychosomatic Society
desired regardless of either the health of their co-twin or a definitive
diagnosis of CFS. Of the 204 pairs identified, complete data were
available for both members of 176 pairs (87%). Among these, 146
pairs were female-female, 10 were male-male, and 20 were male-
female twins. This study was restricted to female-female twins be-
cause the samples of male-male and male-female twins were too
limited to analyze. Written, informed consent was obtained from
each subject in accordance with the regulations of our institutional
Human Subjects Office.
All study subjects completed an extensive mailed survey ques-
tionnaire that included questions on fatigue and the CFS symptom
criteria according to the 1994 revised Centers for Disease Control
and Prevention (CDC) research definition (9). For nonfatigued twins,
a control version of questions was used that did not reference fa-
tigue. A trained research assistant administered the Diagnostic In-
terview Schedule (DIS), version III-A (20), by telephone to all twin
pairs in which at least one member reported fatigue of ?6 months in
duration. The DIS assigns diagnoses using a computer algorithm
based on criteria of the Diagnostic and Statistical Manual of Mental
Disorders, third edition revised (21). The interview included the
sections on major depression, dysthymia, generalized anxiety,
panic, agoraphobia, posttraumatic stress disorder, mania, bipolar
disorders, schizophrenia, eating disorders, somatization, and sub-
stance abuse/dependence. Depression with melancholic features
was scored without including symptoms attributable to CFS.
The questionnaire also included a checklist of self-reported med-
ical problems; twins indicated whether a condition was currently
active or had resolved and whether it had been evaluated by a
physician. The application of the CDC-defined exclusionary medical
conditions to the checklist was determined by consensus of two
general internists, a psychiatrist with expertise in CFS, an infectious
disease specialist, and an internist/emergency room physician with
knowledge of, but little exposure to, patients with CFS. Examples
from the comprehensive list of exclusionary disorders included (but
was not limited to) steroid-dependent asthma, infectious hepatitis,
diabetes, cancer (other than skin cancer), congestive heart failure,
stroke, cirrhosis, multiple sclerosis, and systemic lupus erythema-
tosus. To assess the performance of our list of exclusionary condi-
tions, subjects’ self-reported health conditions were compared with
physician confirmation of these diagnoses by means of chart review
and telephone contact with treating physicians for a subsample of
twins. Among 40 twins, we did not find any fatigued participant to
be ineligible due to an exclusionary condition that was missing or
inaccurately reported on the checklist. Conversely, no exclusionary
conditions were observed in any twin who self-reported good
Definition of Fatigue
We defined three progressively more stringent (but not mutually
exclusive) case definitions of fatigue. The first definition, chronic
fatigue, was based on the response to a single question: “Have you
been fatigued for at least 6 months?” No further inclusionary or
exclusionary conditions were applied.
The second definition, chronic fatigue not explained by medical
exclusions, classified all twins according to chronic fatigue using
data obtained solely from the mailed questionnaire. An algorithm
that defined chronic fatigue using the medical exclusionary compo-
nents of the CDC case definition for CFS was developed (9). The
exclusionary medical criteria were applied to both the chronically
fatigued and nonfatigued twins. To be classified as having medically
unexplained chronic fatigue, twins were required to report fatigue of
?6 months duration that was not lifelong and that resulted in a
substantial reduction of occupational, educational, social, or per-
sonal activities. Twins were excluded from this definition of chronic
fatigue if they had a body mass index ?45 (as stipulated in the CDC
criteria) or reported any of the exclusionary medical conditions.
The third definition (idiopathic chronic fatigue or chronic fatigue
not explained by medical and psychiatric exclusions) further re-
stricted our sample on the basis of DIS-generated psychiatric diag-
noses considered exclusionary by the CDC case definition (9). These
included lifetime mania, hypomania, bipolar disorder, schizophre-
nia, major depression with psychotic or melancholic features, an-
orexia or bulimia nervosa, and current alcohol or substance abuse/
dependence. Identical psychiatric exclusionary criteria were used
with both fatigued and nonfatigued twins.
Determination of Zygosity
Studies have shown that questions about childhood similarity in
twin pairs can be used to correctly classify zygosity with an accuracy
of 95% to 98% compared with biological indicators (22, 23). As part
of the mailed questionnaire all twins were asked questions about
childhood similarity. Initial classification of twin pairs was based on
concordant responses to the following question: “When you were
young, were you as alike as peas in a pod?” Twins were classified as
MZ if they both answered affirmatively; those pairs in which both
twins denied this degree of similarity were considered to be DZ.
Further refinement and classification of twin pairs was based on
additional childhood similarity indicators (eg, eye and hair color
and general physical characteristics). Full details of our zygosity
classification algorithm can be found elsewhere (24).
Two measures of twin concordance were calculated as descrip-
tive indices of the familial clustering of fatigue. The pairwise con-
cordance rate gives the percentage of affected pairs that are concor-
dant (ie, both members of the pair meet the definition for chronic
fatigue); it is simply defined as the ratio of fatigue-concordant twins
divided by the sum of the number of fatigue-discordant and fatigue-
concordant twins. The proband concordance rate gives an estimate
of the prevalence of fatigue among twin siblings of probands. Con-
cordance rates and associated 95% exact confidence intervals are
presented separately for each of the three definitions of fatigue (?6
months of fatigue, chronic fatigue not attributable to medical exclu-
sion criteria, and idiopathic chronic fatigue not explained by med-
ical or psychiatric exclusion criteria). The two-tailed Fisher’s exact
test was used for hypothesis testing to examine whether the differ-
ence between concordance rates in MZ and DZ pairs was significant
for each of the three definitions of chronic fatigue. When MZ rates
were higher than DZ rates, this was interpreted as evidence for a
genetic influence on chronic fatigue.
We also conducted a separate statistical analysis comparing the
concordance rates for the most stringent definition of idiopathic
chronic fatigue after further excluding all twins with a lifetime
history of major depression on the structured psychiatric interview.
Although this is not part of the formal CDC case definition for CFS,
we were concerned that any differences in concordance rates for
idiopathic chronic fatigue might be the result of comorbid depres-
sion, which is known to have a genetic component (25).
Estimates of the relative contribution of genetic and nongenetic
sources of phenotypic variability for each definition of chronic fa-
tigue are derived using a multifactorial model of inheritance that
assumes both polygenetic and environmental effects (26). The num-
ber of concordant and discordant pairs, in combination with exter-
nal estimates of the population prevalence for the three different
definitions of chronic fatigue, were used to derive estimates of the
GENETICS AND CHRONIC FATIGUE
937 Psychosomatic Medicine 63:936–943 (2001)
relative proportion of variance in liability for chronic fatigue attrib-
utable to additive genetic (commonly referred to as heritability),
common environment, and unique environment. Prevalence esti-
mates of chronic fatigue in females were obtained from a recently
completed population-based study of CFS (L. Jason, personal com-
munication, 1998); these prevalences were 4.2%, 2.1%, and 0.4%
for each of our three increasingly stringent definitions of chronic
fatigue. The mean age of the female chronic fatigue cases in the
population-based study was very similar to the average age in the
twin sample (44 vs. 46 years). The structural equation modeling
program Mx (27) was used to derive the estimates of additive genetic
(a2), common environment (c2), and unique environment (e2) effect
as well as the associated 95% confidence intervals (28).
Because this was a self-selected sample of twins, we attempted to
account for possible ascertainment biases that could influence our
estimates of genetic and nongenetic effects (29). We conducted a
series of alternative structural equation models based on varying the
ratio of the probability of ascertaining fatigue-concordant versus
fatigue-discordant twin pairs. When this ratio was 1, there was no
differential ascertainment. An ascertainment correction ratio of 2
implies that a fatigued twin with a fatigued co-twin is twice as likely
to be ascertained as a fatigued twin with a healthy co-twin. Different
models were estimated with ascertainment correction factors that
ranged from 1 to a maximum value of 4.
Table 1 presents the concordance rates in MZ and
DZ twins according to the three increasingly stringent
definitions of chronic fatigue. The concordance rates
for chronic fatigue of ?6 months duration were greater
in MZ twins than DZ twins, although this difference
was not statistically significant (p ? .253). A similar
pattern of higher MZ than DZ concordance rates was
observed for chronic fatigue not attributable to the
medical exclusion criteria; this difference was also not
significant (p ? .109). For idiopathic chronic fatigue
not explained by the medical or psychiatric exclusion
criteria, the difference in concordance rates became
larger, with the MZ pairwise (38%) and proband (55%)
concordance rates significantly elevated (p ? .042)
compared with the respective DZ rates (11% pairwise
and 19% proband). To explore whether depression
could be confounding our analyses, we reanalyzed
chronic fatigue not explained by the medical or psy-
chiatric exclusion criteria after excluding twins if ei-
ther had a history of major depression. In this highly
refined sample of 25 MZ and 11 DZ twins, the differ-
ence in concordance rates remained statistically sig-
nificant (p ? .016) with a 40% pairwise and 57%
proband rate in the MZ twins and a pairwise and
proband rate of 0% in the DZ twins.
Figure 1 presents estimates of the relative contribu-
tion of genetic and nongenetic effects for the three
definitions of chronic fatigue. For chronic fatigue of
?6 months, genetic effects were relatively modest, ac-
counting for 19% of the variance (95% CI ? 0–56),
whereas common environmental effects accounted for
69% of the variance (95% CI ? 32–89). For chronic
fatigue not explained by medical conditions, common
environmental effects still predominated (a2? 30%,
95% CI ? 0–81; c2? 58%, 95% CI ? 9–89). Idio-
pathic chronic fatigue not explained by medical or
psychiatric exclusionary criteria displayed the largest
genetic influence with a heritability of 51% (95% CI ?
7–96). As shown in Table 2, adjusting for various lev-
els of differential ascertainment of concordant to dis-
cordant pairs yielded results that were generally con-
TABLE 1.Concordance for Chronic Fatigue in Monozygotic and Dizygotic Female Twin Pairs
Zygosity Fatigue ?6 monthsa
Chronic Fatigue Not
Due to Medical
Chronic Fatigue Not
Due to Medical or
Monozygotic, N of pairs
Concordant, N of pairs
Discordant, N of pairs
Pairwise concordance %
Proband concordance %
Dizygotic, N of pairs
Concordant, N of pairs
Discordant, N of pairs
Pairwise concordance %
Proband concordance %
ap ? .253 by Fisher’s exact test for comparing the MZ and DZ concordances rates.
bp ? .109 by Fisher’s exact test for comparing the MZ and DZ concordances rates.
cp ? .042 by Fisher’s exact test for comparing the MZ and DZ concordances rates.
D. BUCHWALD et al.
938Psychosomatic Medicine 63:936–943 (2001)
sistent. For chronic fatigue of ?6 months, the
estimates of genetic effects ranged from 19% with no
ascertainment adjustment to a maximum of 30% with
an ascertainment correction ratio of 4; likewise, for
chronic fatigue not explained by medical conditions,
genetic effects ranged from 30% to 43%. Common
environmental effects, which were substantial for both
fatigue of ?6 months and chronic fatigue not ex-
plained by medical conditions, were reduced as the
ascertainment correction ratio was increased. For
chronic fatigue not explained by medical or psychiat-
ric exclusionary criteria, the genetic effects increased
from 51% to 65% as the ascertainment ratio increased
from 1 to 4, whereas the common environmental in-
fluence was diminished from 42% to 7%.
At the beginning of the century a popular diagnosis
that bears some resemblance to modern-day CFS was
neurasthenia. At the time most authorities on neuras-
thenia favored a distinction between an inherited and
Fig. 1.Estimates of genetic and environmental influence on the liability to chronic fatigue with 95% confidence intervals.
TABLE 2.Estimates of the Percentage of Variance in Liability to Chronic Fatigue Due to Additive Genetics, Common Environment,
and Unique Environment Under Different Twin Ascertainment Assumptions
Fatigue ?6 months
Chronic Fatigue Not Due to
Chronic Fatigue Not Due to
Medical or Psychiatric
aThe ratio is defined as of the probability of ascertaining fatigue-concordant versus fatigue-discordant twin pairs. When the ratio ? 1, it is
assumed that there is no differential ascertainment of concordant compared with discordant pairs. Ratios ?1 suggest that fatigue-concordant
twins are more likely than fatigue-discordant twins to be identified in the Chronic Fatigue Twin Registry.
bIncludes five individuals excluded for a body mass index ?45.
GENETICS AND CHRONIC FATIGUE
939Psychosomatic Medicine 63:936–943 (2001)
an acquired form (30). Moreover, it was commonly
observed that the illness ran in families and could
begin in early life (30). Although a greater appreciation
of the methodological difficulties in separating con-
genital and acquired factors developed, a role of heri-
tability was never eliminated; instead it became un-
fashionable to ascribe neurasthenia to inherited
factors. Only in the context of “effort syndrome,” a
condition of easy fatigability with exertion, were care-
ful studies performed; these studies showed a modest
genetic contribution (31). No further studies were con-
ducted for several decades on the heritability of fatigu-
Modern genetic technologies unavailable to the pre-
vious generation of neurasthenia researchers are now
accessible, but few investigators have applied these to
CFS. In one of the first reports, a Scottish group at-
tempted to determine HLA status in patients with CFS
but were inexplicably hampered by difficulties in ob-
taining lymphocyte expression of HLA markers (32).
Another investigator, who did not experience techni-
cal problems, reported no specific association of HLA
class II antigens (33). In contrast, American researchers
found that HLA-DQ3 and HLA-DR5 antigens were sig-
nificantly associated with CFS (34). Taken together,
these and other studies of HLA type have produced
conflicting results and have failed to show any consis-
tent association with CFS (11, 35).
Several other investigations have suggested that a
combination of host and environmental factors may be
involved in the etiology of CFS. One report identified
a family in which 5 of 6 siblings and 3 other immediate
family members developed CFS as adults (17). Blood
samples were collected and tested for natural killer
cell activity over a 2-year period from the 8 affected
and 12 unaffected family members and 8 normal con-
trol subjects. The cytotoxic activity of the affected
immediate family members was significantly lower
than that of the control subjects, with unaffected fam-
ily members intermediate between these two groups.
Although a familial influence (ie, common environ-
ment) could also explain their findings, the authors
concluded that the natural killer cell dysfunction in
this family may have resulted from a genetically deter-
mined immunological abnormality predisposing to
CFS. Others have recently described a new pattern of
autoantibodies to gangliosides, phospholipids, and es-
pecially serotonin in CFS patients (36). The observa-
tion that these antibodies were present among family
members of CFS patients was interpreted in favor of an
inherited predisposition to the illness. Finally, in a
community-wide investigation of pediatric cases of a
CFS-like illness, the best predictive model of illness
produced significant estimates of relative risk for both
environmental and genetic/familial effects, including
a risk of 23.3 for symptoms among other family mem-
Twin studies have been widely used to estimate the
heritability of numerous complex disorders and con-
ditions. Recent research has estimated that heritable
influence accounts for 39% of the variation in major
depression (38), 54% in osteoarthritis (39), and 44% in
blood pressure in the elderly (40). However, twin and
family studies have not been extensively used to ex-
amine the heritability of CFS and similar chronically
fatiguing conditions. An exhaustive, computerized
search of the medical literature of the last decade
yielded only four citations on twins and fatigue, all
using a large volunteer registry of Australian twins (19,
41–43). These publications examined shorter periods
of fatigue (eg, 1 week to 1 month) and not CFS-like
illnesses (19, 41–43) among children and adolescents
as reported by their parents (41) and older individuals
(43). Interestingly, several of these studies found that
the genetic and environmental influences on fatigue
were for the most part independent of those for anxi-
ety, depression, and psychological distress (19, 42, 43).
Despite different populations and highly variable case
definitions of fatigue, these investigations, like our
own, found evidence to support a familial effect.
Publications on objective measures of energy metab-
olism have also demonstrated that many characteris-
tics that contribute to aerobic capacity are heritable
(44–49). For example, level of habitual physical activ-
ity, resting metabolic rate, energy cost of exercise, en-
durance performance response to training, and maxi-
mum oxygen consumption (VO2max) are all, to some
extent, genotype-dependent (45, 47). In this regard, an
intriguing recent report claims to have identified a
gene for physical performance (50). Finally, the aware-
ness of symptoms depends, in large part, on how the
central nervous system processes information (ie, an
individual’s perception). Because MZ twins have re-
markably similar cortical structures, electroencephalo-
graphic patterns, neurotransmitter activity, and auto-
nomic nervous system activation symptom perception
clearly could have a heritable basis (51).
This study has several potential limitations. First, as
with all classical twin studies, there is the assumption
that the rearing experience of MZ and DZ twins is
comparable. If this so-called equal environment as-
sumption is violated, the estimates of genetic influ-
ence could be falsely inflated. However, this assump-
psychiatric disorders and found to be quite tenable
(52, 53). Second, it is important to acknowledge that
the method we used to identify the twin sample was
not ideal. We used a selected sample of volunteer
D. BUCHWALD et al.
940Psychosomatic Medicine 63:936–943 (2001)
twins who responded to solicitations from a variety of
sources. This approach was a practical solution to
identifying a relatively large sample of twins with
chronic fatigue. The more desirable strategy of system-
atically identifying twins with chronic fatigue and CFS
from a well-defined, truly population-based twin reg-
istry is not readily accomplished. The use of selected
samples of twins has a long history in twin studies of
medical and psychiatric disorders (18) and has re-
cently been applied successfully in a study of the
genetics of homosexuality (28). However, this ap-
proach could result in potentially biased ascertain-
ment. Although we repeatedly emphasized in all ad-
vertisements and contacts with support groups and
physicians that probands could participate regardless
of either a definitive diagnosis or the health of their
co-twin, it is likely that subjects screened themselves
as eligible or ineligible. In this study, an ascertainment
bias could have occurred if twin pairs in which both
members experienced fatigue were more likely to vol-
unteer and complete the lengthy booklet. We ad-
dressed this in our analysis by constructing a series of
ascertainment scenarios to examine the impact on her-
itability estimates. Another type of ascertainment bias
might have occurred if there was differential partici-
pation of chronic fatigue–concordant pairs according
to zygosity. If an excess of MZ concordant twins had
volunteered, this could have resulted in an overesti-
mate of the influence of genetics. Without a popula-
tion-based sample of twins, it is very difficult to esti-
mate the extent of this type of bias.
Other possible limitations are the methods we used
to define and measure fatigue and exclusionary medi-
cal conditions. Although a clinical examination of all
twins was not feasible, we closely followed the CFS
symptom and psychiatric criteria as articulated by the
CDC. In addition, our three increasingly stringent mea-
sures of fatigue were consistent with those used in
several recent epidemiological studies (L. Jason, L
Steele, S. Reyes, personal communication, 1998) and
permitted us to assess the magnitude of genetic influ-
ence across these different definitions. We were like-
wise concerned with the use of self-reported health
conditions. This methodology could result in either an
overestimate or underestimate of the actual rate of
idiopathic chronic fatigue depending on whether in-
dividuals did not report conditions that might be ex-
clusionary or incorrectly claimed to have been diag-
nosed with conditions they did not actually have.
However, we presented evidence that the measure-
ment of exclusionary medical conditions using self-
report was adequate. Lastly, we considered the possi-
bility that the heritable effects on fatigue might be a
surrogate for a genetic influence on major depression,
because this diagnosis is common in chronic fatigue
(54), has known genetic influences (25), and is not an
exclusionary psychiatric criterion for CFS. When we
excluded all twins with a lifetime history of major
depression from the definition of idiopathic chronic
fatigue, the MZ concordance rates remained higher
than the DZ rates. Similarly, researchers using the
Australian Twin Registry have demonstrated that fa-
tigue seems to be under genetic influence that is dis-
tinct from the genetic influence on depression (19, 42,
In conclusion, the present study supports the famil-
ial aggregation of fatiguing illnesses. Furthermore,
these data are consistent with the presence of a genetic
influence on idiopathic chronic fatigue. An important
observation was that heritability increased as our def-
inition of chronic fatigue was made more stringent.
This suggests that future research would benefit by
distinguishing CFS from milder forms of chronic fa-
tigue. Our study provides evidence that there is a
genetic vulnerability to idiopathic chronic fatigue. In
this respect, CFS and related conditions may be simi-
lar to other disorders in which situational and envi-
ronmental factors enhance the likelihood of illness
expression. A practical implication for the non-ill pro-
band will be to examine which of these factors might
be important triggers and perpetuators of CFS and to
identify those that might be amenable to change.
The authors thank the participants in the University
of Washington Twin Registry for their cooperation,
patience, and goodwill, and Dr. Leigh Sawyer, Program
Officer, National Institute of Allergy and Infectious
Diseases, for her encouragement and support. We also
acknowledge Anthony Komaroff and Seth Eisen for
their insightful comments on this manuscript and our
advisory panel, who with sage advice and ongoing
encouragement improved our scientific efforts. This
work was funded by Grant U19 AI38429 from the Na-
tional Institutes of Health (D.B.).
1. Bates DW, Schmitt W, Buchwald D, Ware NC, Lee J, Thoyer E,
Kornish RJ, Komaroff AL. Prevalence of fatigue and chronic
fatigue syndrome in a primary care practice. Arch Intern Med
2. Buchwald D, Sullivan J, Komaroff A. Frequency of “chronic
active Epstein-Barr virus infection” in a general medical prac-
tice. JAMA 1987;257:2303–7.
3. Cathebras PJ, Robbins JM, Kirmayer LJ, Hayton BC. Fatigue in
primary care: prevalence, psychiatric comorbidity, illness be-
havior, and outcome. J Gen Intern Med 1992;7:276–86.
4. Kroenke K, Wood DR, Mangelsdorff AD, Meier NJ, Powell JB.
Chronic fatigue in primary care: prevalence, patient character-
istics, and outcome. JAMA 1988;260:929–34.
GENETICS AND CHRONIC FATIGUE
941Psychosomatic Medicine 63:936–943 (2001)
5. McDonald E, David AS, Pelosi AJ, Mann AH. Chronic fatigue in
primary care attenders. Psychol Med 1993;23:987–98.
6. Pawlikowska T, Chalder T, Hirsch SR, Wallace P, Wright DJ,
Wessely SC. Population based study of fatigue and psychologi-
cal distress. BMJ 1994;308:763–6.
7. Chen M. The epidemiology of self-perceived fatigue among
adults. Prev Med 1986;15:74–81.
8. Hammond E. Some preliminary findings on physical complaints
from a prospective survey of 1,064,004 men and women. Am J
Public Health 1964;54:11–23.
9. Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff
A. The chronic fatigue syndrome: a comprehensive approach to
its definition and study. Ann Intern Med 1994;121:953–9.
10. Komaroff AL, Buchwald DS. Chronic fatigue syndrome: an up-
date. In: Coggins CH, Hancock EW, Levitt LJ, editors. Annual
review of medicine. Vol 49. Palo Alto (CA): Annual Reviews;
1998. p. 1–13.
11. Straus SE, Tosato G, Armstrong G, Lawley T, Preble OT, Henle
W, Davey R, Pearson G, Epstein J, Brus I. Persisting illness and
fatigue in adults with evidence of Epstein-Barr virus infections.
Ann Intern Med 1985;107:7–16.
12. Manu P, Matthews DA, Lane TJ, Tennen H, Hesselbrock V,
Mendola R, Affleck G. The mental health of patients with a chief
complaint of chronic fatigue: a prospective evaluation and fol-
low-up. Arch Intern Med 1988;148:2213–7.
13. Buchwald D, Pascualy R, Bombardier C, Kith P. Sleep disorders
in patients with chronic fatigue. Clin Infect Dis 1994;18(Suppl
14. Demitrack MA, Dale JK, Straus SE, Laue L, Listwak SJ, Kruesi
MJ, Chrousos GP, Gold PW. Evidence for impaired activation of
the hypothalamic-pituitary-adrenal axis in patients with
chronic fatigue syndrome. J Clin Endocrinol Metab 1991;73:
15. Landay AL, Jessop C, Lennette ET, Levy JA. Chronic fatigue
syndrome: clinical condition associated with immune activa-
tion. Lancet 1991;338:707–12.
16. Bou-Holaigah I, Rowe PC, Kan J, Calkins H. The relationship
between neurally mediated hypotension and the chronic fatigue
syndrome. JAMA 1995;274:961–7.
17. Levine PH, Whiteside TL, Friberg D, Bryant J, Colclough G,
Herberman RB. Dysfunction of natural killer activity in a family
with chronic fatigue syndrome. Clin Immunol Immunopathol
18. Hubric Z, Robinette CD. The study of human twins in medical
research. N Engl J Med 1984;310:435–41.
19. Hickie I, Kirk K, Martin N. Unique genetic and environmental
determinants of prolonged fatigue: a twin study. Psychol Med
20. Robins LN, Helzer JE. Diagnostic Interview Schedule (DIS): ver-
sion III-A. St Louis: Department of Psychiatry, Washington Uni-
versity School of Medicine; 1985.
21. DSM-III-R. Diagnostic and statistical manual of mental disor-
ders. 3rd ed revised. Washington DC: American Psychiatric
22. Torgersen S. The determination of twin zygosity by means of a
mailed questionnaire. Acta Genet Med Gemellol (Roma) 1979;
23. Eisen SA, Neuman R, Goldberg J, Rice J, True W. Determining
zygosity in the Vietnam Era Twin Registry: an approach using
questionnaires. Clin Genet 1989;35:423–32.
24. Buchwald D, Herrell R, Ashton S, Belcourt M, Schmaling K,
Goldberg J. Chronic fatigue twin registry: method of construc-
tion, composition, and zygosity assignment. Twin Res 1999;2:
25. Kendler KS, Heath A, Martin NG, Eaves LJ. Symptoms of anxiety
and depression in a volunteer twin population: the etiologic role
of genetic and environmental factors. Arch Gen Psychiatry 1986;
26. Khoury MJ, Beaty TH, Cohen BH. Fundamentals of genetic ep-
idemiology. New York: Oxford University Press; 1993.
27. Neale MC. Mx: statistical modeling. 2nd ed. Richmond (VA):
Department of Psychiatry, Medical College of Virginia; 1997.
28. Neale MC, Miller MB. The use of likelihood-based confidence
intervals in genetic models. Behav Genet 1997;27:113–20.
29. Bailey JM, Pillard RC, Neale MC, Agyei Y. Heritable factors
influence sexual orientation in women. Arch Gen Psychiatry
30. Wessely S, Hotopf M, Sharpe M. Chronic fatigue syndromes.
New York: Oxford University Press; 1997.
31. Wheeler E, White P, Reed E, Cohon M. Familial incidence of
neurocirculatory asthenia (‘anxiety neurosis,’ ‘effort syndrome’).
J Clin Invest 1948;27:562–82.
32. Behan PO, Bakheit AM. Clinical spectrum of post-viral fatigue
syndrome. Br Med Bull 1991;47:793–808.
33. Middleton D, Savage D, Smith D. No association of HLA class II
antigens in chronic fatigue syndrome. Dis Markers 1991;9:47–9.
34. Keller RH, Lane JL, Klimas N, Reiter WM, Fletcher MA, van Riel
F, Morgan R. Association of HLA class II antigens and chronic
fatigue immune dysfunction syndrome. Clin Infect Dis 1994;18:
35. Marrie TJ, Ross L, Montague TJ, Doan B. Post-viral fatigue syn-
drome. Clin Ecol 1987;5:5–10.
36. Klein R, Berg PA. High incidence of antibodies to 5-hydroxy-
tryptamine, gangliosides and phospholipids in patients with
chronic fatigue and fibromyalgia syndrome and their relatives:
evidence for a clinical entity of both disorders. Eur J Med Res
37. Bell KM, Cookfair D, Bell DS, Reese P, Cooper L. Risk factors
associated with chronic fatigue syndrome in a cluster of pedi-
atric cases. Rev Infect Dis 1991;13(Suppl 1):S32–8.
38. Kendler KS, Prescott CA. A population-based study of lifetime
major depression in men and women. Arch Gen Psychiatry
39. Spector TD, Cicuttini F, Baker J, Loughlin J, Hart D. Genetic
influence on osteoarthritis. BMJ 1996;312:940–3.
40. Hong Y, de Faire U, Heller DA, McClearn GE, Pedersen N.
Genetic and environmental influences on blood pressure in
elderly twins. Hypertension 1994;24:663–70.
41. Farmer A, Scourfield J, Martin N, Cardno A, McGuffin P. Is
disabling fatigue in childhood influenced by genes? Psychol
42. Hickie I, Bennett B, Lloyd A, Heath A, Martin N. Complex
genetic and environmental relationships between psychological
distress, fatigue and immune functioning: a twin study. Psychol
43. Kirk KM, Hickie IB, Martin NG. Fatigue as related to anxiety and
depression in a community-based sample of twins aged over 50.
Soc Psychiatry Epidemiol 1999;34:85–90.
44. Bouchard C, Dionne FT, Simoneau JA, Boulay MR. Genetics of
aerobic and anaerobic performances. Exerc Sport Sci Rev 1992;
45. Bouchard C, Lesage R, Lortie G, Simoneau JA, Hamel P, Boulay
MR, Perusse L, Theriault G, Leblanc C. Aerobic performance in
brothers, dizygotic and monozygotic twins. Med Sci Sports Ex-
46. Fagard R, Bielen E, Amery A. Heritability of aerobic power and
anaerobic energy generation during exercise. J Appl Physiol
D. BUCHWALD et al.
942Psychosomatic Medicine 63:936–943 (2001)
47. Hamel P, Simoneau JA, Lortie G, Boulay MR, Bouchard C. He- Download full-text
redity and muscle adaptation to endurance training. Med Sci
Sports Exerc 1986;18:690–9.
48. Simoneau JA, Bouchard C. Genetic determinism of fiber type
proportion in human skeletal muscle. FASEB J 1995;9:
49. Tremblay A, Poehlman ET, Despres JP, Theriault G, Danforth E,
Bouchard C. Endurance training with constant energy intake in
identical twins: changes over time in energy expenditure and
related hormones. Metabolism 1997;46:499–503.
50. Montgomery HE, Marshall R, Hemingway H, Myerson S, Clark-
son P, Dollery C, Hayward M, Holliman DE, Jubb M, World M,
Thomas EL, Brynes AE, Saeed N, Barnard M, Bell JD, Prasad K,
Rayson M, Talmud PJ, Humphries SE. Human gene for physical
performance. Nature 1998;393:221–2.
51. Watson D, Pennebaker JW. Situational, dispositional, and ge-
netic bases of symptom reporting. In: Skelton JA, Croyle RT,
editors. Mental representation in health and illness. New York:
Springer Verlag; 1991. p. 61–84.
52. Hettema JM, Neale MC, Kendler KS. Physical similarity and the
equal-environment assumption in twin studies of psychiatric
disorders. Behav Genet 1995;25:327–35.
53. Kendler KS, Gardner CO. Twin studies of adult psychiatric and
substance dependence disorders: are they biased by differences
in the environmental experiences of monozygotic and dizygotic
twins in childhood and adolescence? Psychol Med 1998;28:
54. Manu P, Matthews DA, Lane TJ. The mental health of patients
with a chief complaint of chronic fatigue: a prospective evalu-
ation and follow-up. Arch Intern Med 1988;148:2213–7.
GENETICS AND CHRONIC FATIGUE
943 Psychosomatic Medicine 63:936–943 (2001)