Content uploaded by Lucinda Bateman
Author content
All content in this area was uploaded by Lucinda Bateman on Mar 28, 2025
Content may be subject to copyright.
Fatigue: Biomedicine, Health & Behavior
ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/rftg20
Chronic fatigue syndrome and co-morbid and
consequent conditions: evidence from a multi-site clinical
epidemiology study
Lucinda Bateman, Salima Darakjy, Nancy Klimas, Daniel Peterson, Susan M.
Levine, Ali Allen, Shane A. Carlson, Elizabeth Balbin, Gunnar Gottschalk &
Dana March
To cite this article: Lucinda Bateman, Salima Darakjy, Nancy Klimas, Daniel Peterson, Susan
M. Levine, Ali Allen, Shane A. Carlson, Elizabeth Balbin, Gunnar Gottschalk & Dana March
(2015) Chronic fatigue syndrome and co-morbid and consequent conditions: evidence from a
multi-site clinical epidemiology study, Fatigue: Biomedicine, Health & Behavior, 3:1, 1-15, DOI:
10.1080/21641846.2014.978109
To link to this article: https://doi.org/10.1080/21641846.2014.978109
Published online: 06 Dec 2014.
Submit your article to this journal
Article views: 340
View related articles
View Crossmark data
Citing articles: 7 View citing articles
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rftg20
Chronic fatigue syndrome and co-morbid and consequent
conditions: evidence from a multi-site clinical epidemiology study
Lucinda Bateman
a,b
, Salima Darakjy
c
, Nancy Klimas
b,d
, Daniel Peterson
b,e
, Susan
M. Levine
b,f
, Ali Allen
a,b
, Shane A. Carlson
a,b
, Elizabeth Balbin
g
, Gunnar Gottschalk
e
and Dana March
c
*
a
Fatigue Consultation Clinic, Salt Lake City, UT, USA;
b
Chronic Fatigue Initiative, New York,
USA;
c
Department of Epidemiology, Columbia University, New York, USA;
d
Departments of
Medicine and Clinical Immunology, Nova Southeastern University, Miami, FL, USA;
e
Sierra
Internal Medicine, Incline Village, NV, USA;
f
Susan Levine, MD, Private Practice, New York,
USA;
g
Department of Psychology and Behavioral Medicine, University of Miami, Miami, FL,
USA
(Received 31 August 2014; final version received 5 November 2014)
Background: Epidemiologic data that inform our understanding of the type,
frequency, and burden of co-morbidities with chronic fatigue syndrome is
limited. Purpose: To elucidate co-morbid and consequent conditions, using data
from a clinical epidemiology study of long-term CFS patients. Methods: Some
960 adults with CFS were identified at four sites specializing in the diagnosis
and treatment of CFS. Patients reported their demographics, CFS course, other
medical diagnoses, and current functioning. We determined associations
between: co-morbidities and a patient’s current health relative to their health
when diagnosed with CFS; CFS symptom severity at onset and subsequent
diagnosis with a co-morbid condition; and presence of a co-morbidity and
functional ability. We also modeled the change in CFS symptom severity over
time as it relates to the presence of a co-morbidity. Results: Of the sample, 84%
was diagnosed with one or more co-morbid conditions after CFS onset.
Fibromyalgia, depression, anxiety, and hypothyroidism were the most common
diagnoses. Nearly 60% of the sample reported a mental illness. Conclusions: In
general, co-morbid conditions reduced functional ability and were associated
with the worsening of CFS symptoms over time. This study provides important
new information on the prevalence of co-morbid conditions and their impact on
the course of CFS.
Keywords: co-morbidity; depression; anxiety; functional ability
Introduction
Chronic fatigue syndrome (CFS) is a severe, disabling condition, for which the etiology
remains unclear, although recent studies suggest neuroinflammation [1] and/or autoim-
munity [2] may play a role. A growing number of epidemiologic studies, whether com-
munity- or clinic-based, have contributed valuable information regarding the natural
© 2014 IACFS/ME
*Corresponding author. Email: dm2025@cumc.columbia.edu
Fatigue: Biomedicine, Health & Behavior, 2015
Vol. 3, No. 1, 1–15, http://dx.doi.org/10.1080/21641846.2014.978109
history and factors associated with the onset, persistence, and outcome of the illness.[3–
7] However, the wide range of study types, methods, and measures and case definitions
used provide an incomplete picture of this heterogeneous syndrome, which offers
opportunities for further characterization.
One such opportunity lay in the domain of conditions that are co-morbid with CFS of
long duration. While a number of studies have examined specific co-occurring conditions
or classes of conditionsin CFS patients, epidemiologic data that broadly inform our under-
standing of the type, frequency, and burden of co-morbidities is limited. For example, the
majority of investigations have focused on co-morbid Axis I psychiatric disorders [8–13]
and some have examined Axis II, specifically personality, disorders.[10,13–18]
Given that a suite of conditions characterized by fatigue complicate the diagnosis of
CFS [19,20] according to several widely used sets of diagnostic criteria,[21] there is a
dearth of studies that explore the emergence of both mental health and medical con-
ditions over time in well-defined cases of CFS. Indeed, few studies [22,23] have
explored the broad set of consequent conditions that may contribute to heterogeneity
in overall illness burden, functioning, and outcome in those with CFS.[24] Character-
ization of the patterns of co-morbid conditions would contribute to an understanding of
the overall burden of illness experienced by those with CFS, and may have impli-
cations, ultimately, for our understanding of the etiology of CFS.
Using data from a multi-site clinical epidemiology study of long-term CFS patients
diagnosed in CFS specialty practices, we sought to elucidate, both descriptively and ana-
lytically, co-morbid and consequent (i.e., after CFS onset) conditions. The objectives of
this paper, therefore, are: (1) to determine the individual prevalence and total number of
co-morbid and consequent conditions; (2) to examine relations among co-morbid and con-
sequent conditions and patient characteristics; (3) to determine associations between the
reported severity of CFS symptoms at illness onset and co-morbid and consequent con-
ditions; (4) to examine associations between co-morbid and consequent conditions and
current health status and functioning; and (5) to determine associations between the
change in symptom severity over time and co-morbid and consequent conditions.
Methods
Study participants
English-speaking adults (18 years and older) who were clinically diagnosed with CFS per
the Fukuda [25] criteria upon initial assessment were identified from four clinical sites spe-
cializing in CFS diagnosis and treatment: Miami, Florida (n= 401), Sierra, Nevada (n=
353), New York, New York (n= 316), and Salt Lake City, Utah (n= 360). A telephone
interview was conducted with patients who had been diagnosed at one of the sites at
least five years earlier, with an emphasis on patients who initially presented to the clinical
study site 10 years or more prior. Of the 1430 patients originally identified by chart review,
355 individuals were lost to follow-up, 59 patients were deceased, and 57 individuals
declined to participate in the study. Thus, a total of 960 patients completed the survey
across the four clinical study sites: Miami, Florida (n= 237), Sierra, Nevada (n=221),
New York, New York (n= 256), and Salt Lake City, Utah (n= 246).
Measures
From June 2012 through February 2013, patients were contacted via telephone by
trained interviewers and asked a series of demographic questions along with questions
2L. Bateman et al.
about their illness course, co-morbidities, current functioning, and treatment. A struc-
tured questionnaire was developed by the clinical investigators and staff (LB, NK, DP,
SML, SAC) with five goals: (1) to establish a longitudinal patient database to document
the natural history of CFS; (2) to determine conditions that are co-morbid with and con-
sequent to CFS; (3) to identify factors that contribute to CFS recovery and relapse; (4)
to evaluate the presentation and trajectory of CFS symptom change over time; and (5) to
identify effective treatments in the course of illness reported by the patient. The ques-
tionnaire presented patients with specific response categories for certain questions and
open-ended response options for others.
Demographics
The demographic variables included age, sex, height, weight, marital status, race, eth-
nicity, country of birth, and education level. Race was defined by the study participants;
we subsequently categorized responses into two groups (white or non-white) for all
statistical analyses to adjust for potential differences between the groups with respect
to the outcome measures.
Illness course
Patients reported their CFS onset type (gradual or sudden). Gradual onset was defined
as the emergence of CFS symptoms over several weeks to months without a clear trig-
gering event. CFS symptoms that definitely started within a 1–30 day window were
considered sudden onset. Duration of illness and time elapsed since the initial diagnosis
by the specialist were determined from the medical record. Patients rated the initial and
current severity of nine CFS symptoms. Symptoms included post-exertional malaise,
impaired memory or concentration, unrefreshing sleep/sleep difficulties, headache,
muscle pain, joint pain, sore throat, lymph node pain/tenderness (the eight defining
symptoms of the Fukuda criteria) plus the addition of orthostatic intolerance, and
whether they had ever been diagnosed by a medical professional with any other con-
dition(s) since/after their CFS onset. The severity of each CFS symptom (reported as
absent, mild, moderate, or severe) was converted to a scalar variable (absent = 0,
mild = 1, moderate = 2, and severe = 3).
For each patient, we calculated initial and current global CFS symptom severity
scores as the sum of the initial and current severity ratings, respectively, for all nine
CFS symptoms. In addition, global CFS symptom severity change scores were calcu-
lated as the current global score minus the initial global score. Possible values for the
global CFS symptom severity change score ranged from −27 (indicating complete
symptom remission, i.e., reporting the most severe ratings for all nine symptoms at
CFS onset to reporting “absent”for all nine CFS symptoms currently) to a hypothetical
27 (indicating maximum worsening over time of all nine CFS symptoms).
Patients also reported on the number and duration of remissions from CFS, as well as
current functioning (engagement in work, school, or an equivalent activity) and their
current health status relative to their health status at the time they were diagnosed with CFS.
Co-morbid conditions
Specific co-morbid conditions of interest included both medical diagnoses and psycho-
logical or mental health conditions. Medical diagnoses listed included fibromyalgia;
Fatigue: Biomedicine, Health & Behavior 3
severe spine problems (lumbar or cervical disc disease, stenosis, radiculopathy, or
degenerative joint disease [DJD] requiring surgery, injections, physical therapy,
steroids, or strong medications); common endocrine conditions (hypothyroidism, meta-
bolic syndrome/type 2 diabetes), other autoimmune disease, primary sleep disorders
(narcolepsy, sleep apnea, periodic limb movement disorder [PLMD] or restless leg syn-
drome [RLS]), and neurological disease. Sex-specific conditions included endometrio-
sis or menopause in women, and low testosterone in men. Patients also reported cancer
malignancies, and the type of cancer (i.e., site or system affected). Mental health con-
ditions listed included depression, anxiety, post-traumatic stress disorder (PTSD), and
bipolar disorder (considered exclusionary for initial CFS diagnosis). Patients were also
asked to report any other major diagnoses.
Statistical analyses
First, chi-square analysis and one-way ANOVA were used to determine differences across
clinical sites with respect to patient characteristics and the prevalence of co-morbidities.
Second, a principal component factor analysis was performed on the 14 non-sex-
specific co-morbid conditions to determine whether they could be condensed into
fewer variables. Correlations among the extracted factors were allowed.
Third, to determine associations among co-morbid conditions and other character-
istics of CFS, three types of regression analyses were performed, yielding crude and
adjusted odds ratios (OR) and 95% confidence intervals (CI). Ordinal logistic
regression was used to determine the associations between the presence of a co-morbid-
ity factor or cancer malignancy and a patient’s current health relative to his or her health
when diagnosed with CFS. Next, logistic regression was used to examine the associ-
ations between global CFS symptom severity at onset and the subsequent diagnosis
with a co-morbid condition, and the associations between the presence of a co-morbid-
ity factor or cancer malignancy and functional ability (engagement in work, school, or
regular equivalent activity). Finally, linear regression was used to model the change in
CFS symptom severity over time as it relates to the presence of a co-morbidity factor or
cancer malignancy.
In order to determine potential covariates for adjusted models, Pearson correlation
coefficients were calculated for co-morbidity factors and patient characteristics as well
as cancer malignancy and patient characteristics. Adjusted regression models accounted
for: clinical site; patient characteristics including age, sex, race, country of birth, and
education; and CFS characteristics (e.g., illness duration or initial symptom severity).
All statistical tests were two-sided. Missing data were missing at random, and we
reported the number of valid responses. All analyses were performed using IBM
SPSS Statistics version 21 (IBM Corporation, USA).
The clinical data used for this study were collected employing HIPAA compliant
methods [26] within the individual clinics. Prior to central compilation and analysis,
the data were completely de-identified. This study was reviewed and deemed exempt
by both the Western Institutional Review Board and the Institutional Review Board
of Columbia University Medical Center.
Results
As shown in Table 1, the majority of participants in this survey sample are middle-aged,
white women who were born in the United States (US) and are highly educated. These
4L. Bateman et al.
Table 1. Sample characteristics.
Characteristic Total survey Florida Nevada New York Utah
Sample Subsample Subsample Subsample Subsample P-value
a
(N= 960) (n= 237) (n= 221) (n= 256) (n= 246)
Current age [years] (M, SD) 55.1 ± 12.3 56.4 ± 11.5 59.0 ± 11.1 55.6 ± 9.2 50.0 ± 15.0 <0.001
Age at CFS onset [years] (M, SD) 35.3 ± 11.8 37.5 ± 13.0 36.8 ± 10.7 35.6 ± 10.0 31.7 ± 12.4 <0.001
Sex (%)
Female 79.8 80.6 65.2 87.5 84.1 <0.001
BMI [kg/m
2
] (M, SD) 25.71 ± 5.52 25.34 ± 5.14 25.99 ± 5.21 25.57 ± 5.86 25.98 ± 5.75 0.50
%
Underweight [<18.50] 4.5 4.7 2.8 4.3 6.1 0.40
Normal [18.50–24.99] 48.8 48.7 47.2 52.7 46.3 0.49
Overweight [25.00–29.99] 26.7 28.9 31.3 21.5 26.0 0.09
Obese [≥30.00] 19.9 17.7 18.7 21.5 21.5 0.63
Marital status (%)
Single 23.4 22.3 21.6 31.6 17.5 0.002
Married/partnered 57.5 57.1 59.6 49.2 64.6 0.005
Separated/divorced/widowed 19.1 20.6 18.8 19.1 17.9 0.21
Race (%)
White 94.9 92.9 93.6 94.4 98.4 0.03
Ethnicity (%)
Hispanic 4.6 13.2 0.9 3.2 0.9 <0.001
Foreign born (%) 6.5 16.4 2.7 6.3 1.2 0.003
Education (%)
College degree 63.9 65.6 61.8 73.7 54.1 <0.001
CFS duration
b
[years] (M, SD) 15.4 ± 6.2 12.7 ± 5.4 19.0 ± 6.4 17.5 ± 5.3 12.7 ± 5.1 <0.001
Notes:
a
P-value for comparison across site (chi-square statistic for categorical variables; one-way ANOVA for continuous variables).
b
Time from clinical diagnosis.
M= mean; SD = standard deviation; BMI = body mass index; kg = kilograms; m = meters; CFS = chronic fatigue syndrome.
Fatigue: Biomedicine, Health & Behavior 5
demographics are typical of clinic-based [27,28] as opposed to population-based CFS
samples. There were statistical differences across site with respect to all demographic
variables except body mass index (BMI). The Nevada subsample was older and had
a larger proportion of men. Overall, racial and ethnic diversity were low and the
majority of Hispanic and foreign-born patients belonged to the Florida and
New York subsamples. The mean CFS duration was lower for the Florida and Utah
subsamples, which were very similar in this respect. Likewise, mean CFS duration
was similar for the Nevada and New York subsamples.
Specific co-morbid conditions of interest are presented in Table 2. A high pro-
portion of the total survey sample (84.0%) reported having been diagnosed with at
least one co-morbid condition after the onset of CFS. Fibromyalgia, depression,
anxiety, and hypothyroidism were the most common diagnoses. Nearly three-fifths
of the sample (57.7%) reported a mental health condition (depression, anxiety,
PTSD, and/or bipolar disorder). A substantial proportion reported cancer malignancies,
the majority of which were skin cancer (7.5%) among patients in Florida and Nevada.
Nine individuals had multiple cancer diagnoses. Women reported slightly more co-
morbidities than men, although the differences were statistically significant for
Florida and Utah only.
Pearson correlation coefficients for patient characteristics and co-morbid conditions
appear in Table 3. Our principal component factor analysis condensed 14 co-morbid
conditions into four co-morbidity factors: (1) anxiety, depression, and PTSD; (2)
hypothyroidism and other autoimmune disease; (3) narcolepsy; and (4) neurological
disease. The strongest correlations were observed between cancer and age; the
hypothyroidism and other autoimmune disease factor and age, sex, and BMI; and the
anxiety, depression, and PTSD factor and CFS duration. Foreign-born status was not
statistically correlated with any co-morbidity factors or cancer malignancy; therefore
we did not adjust for it in multivariable regression.
Results from the regression analysis for the association between initial CFS
symptom severity and co-morbid conditions are shown in Table 4. In the crude
model, each one-point increase in global symptom severity increased the odds of an
anxiety, depression, and/or PTSD diagnosis by 3%. The adjusted model –which fac-
tored potential confounding by site, age, sex, race, education, and CFS duration –
revealed statistically significant increased odds of an anxiety, depression, and/or
PTSD diagnosis (5%); a hypothyroidism and/or other autoimmune disease diagnosis
(5%); and a narcolepsy diagnosis (12%) with each one-point increase in global
symptom severity.
Results from the regression analysis for the association between co-morbid con-
ditions and current health and functioning are shown in Table 5. In general, patients
subsequently diagnosed with a co-morbid condition were less likely to report better
current health as compared to their health at the time they were diagnosed with CFS.
This relationship was true for all non-sex-specific conditions analyzed except bipolar
disorder, which contributed to greater odds of health status improvement over time,
although it was not a statistically significant association. With respect to functional
ability, patients with any co-morbid condition other than depression or sleep apnea
were less likely to engage in work, school, or equivalent regular activity. However,
fibromyalgia is the only condition for which the association with decreased functional
ability was statistically significant. It is important to note that less than half of the
sample (41.1%) reported currently working, attending school, and/or engaging in
regular activity equivalent to work or school.
6L. Bateman et al.
Table 2. Prevalence (%) and average number of co-morbid conditions.
Condition (valid responses
b
) Total survey Florida Nevada New York Utah
Sample Subsample Subsample Subsample Subsample P-value
a
(N= 960) (n= 237) (n= 221) (n= 256) (n= 246)
Fibromyalgia (n= 945) 61.0 65.2 38.5 73.1 65.0 <0.001
Severe spine problem (n= 913) 26.5 39.3 22.3 16.1 27.6 <0.001
Hypothyroidism (n= 921) 35.0 38.5 27.5 32.3 40.7 0.01
Other autoimmune disease (n= 909) 15.2 22.8 13.8 11.8 12.3 0.003
Type 2 diabetes or metabolic syndrome (n= 917) 10.3 12.0 6.4 15.1 7.7 0.008
Narcolepsy (n= 914) 3.1 4.8 3.6 1.4 2.4 0.18
Sleep apnea (n= 916) 21.9 27.6 17.7 11.9 29.3 <0.001
PLMD or RLS (n= 916) 17.4 16.8 15.6 10.9 25.2 0.001
Endometriosis
c
(n= 726) 20.1 22.2 24.6 10.9 23.7 0.003
Menopause
c
(n= 746) 60.3 72.3 62.5 60.8 47.6 <0.001
Low testosterone
d
(n= 184) 36.4 45.5 37.0 25.0 33.3 0.35
Cancer malignancy (n= 916) 16.4 21.0 21.5 13.8 9.8 0.001
Depression (n= 928) 47.4 52.4 33.5 36.8 65.0 <0.001
Anxiety (n= 921) 39.7 54.7 23.7 30.8 48.0 <0.001
PTSD (n= 918) 12.7 20.3 8.8 10.7 11.0 0.001
Bipolar disorder (n= 912) 3.0 3.9 3.7 0.5 3.7 0.11
Neurological disease (n= 896) 9.4 11.0 10.9 4.7 10.6 0.07
Count
e
(M
f
, SD)
Men (n= 67) 2.7 ± 2.1 3.6 ± 2.5 2.0 ± 1.7 2.4 ± 2.1 2.9 ± 1.9 0.08
Women (n= 509) 3.6 ± 2.1 4.4 ± 2.3 2.9 ± 2.0 2.9 ± 1.8 4.0 ± 1.9 <0.001
Notes:
a
P-value for comparison across site (chi-square statistic).
b
Total Nranges from 194 to 960 for this domain of questions, indicating both missing data and sex-specific responses. Data presented reflect percentages of valid responses.
c
Women only.
d
Men only.
e
Out of non-sex-specific conditions queried in survey: fibromyalgia, severe spine problem, hypothyroidism, other autoimmune disease, type 2 diabetes or metabolic syndrome,
narcolepsy, sleep apnea, PLMD or RLS, cancer malignancy, depression, anxiety, PTSD, bipolar disorder, and neurological disease.
f
Value is the aggregate mean, calculated as the sample mean of the average number of co-morbid conditions per individual.
PLMD = periodic limb movement disorder; RLS = restless leg syndrome; PTSD = post-traumatic stress disorder.
Fatigue: Biomedicine, Health & Behavior 7
Table 3. Correlations (Pearson) between co-morbid conditions and patient characteristics.
Co-morbidity factor
c
Cancer
Patient
characteristics
Anxiety/
Depression/
PTSD
Hypothyroid/
Autoimmune
disease Narcolepsy
Neurological
disease Malignancy
Age −0.036 0.105
b
0.020 0.048 0.258
b
Gender
d
0.117
b
0.245
b
−0.052 0.007 0.039
BMI 0.088
b
0.100
b
0.029 0.012 0.025
Race
e
0.072
a
−0.021 0.016 −0.020 −0.022
Foreign born
f
−0.001 −0.058 −0.022 −0.054 −0.025
Education
g
0.069
a
−0.017 0.012 0.032 0.004
CFS duration −0.109
b
0.004 0.015 −0.067
a
0.101
b
CFS onset
type
h
−0.089
b
−0.077
a
0.041 0.081
a
0.010
Notes:
a
P< 0.05.
b
P< 0.01.
c
Derived from principal component factor analysis (oblique rotation) of the 14 non-sex-specific conditions, which
yielded four co-morbidity factors: (1) anxiety, depression, and PTSD; (2) hypothyroidism and other autoimmune
disease; (3) narcolepsy; and (4) neurological disease.
d
Dichotomous variable (men versus women); reference group is men.
e
Dichotomous variable (white versus non-white); reference group is white.
f
Dichotomous variable (US born versus foreign born); reference group is US born.
g
Dichotomous variable (bachelor’s/graduate/professional degree versus other); reference group is bachelor’s/
graduate/professional degree.
h
Dichotomous variable (gradual versus sudden); reference group is gradual.
PTSD = post-traumatic stress disorder; BMI = body mass index; CFS = chronic fatigue syndrome.
Table 4. Odds ratios (OR) and 95% confidence intervals (CI) for the association between
global initial CFS symptom severity
a
and co-morbid conditions.
Model 1 Model 2
(Crude) (Adjusted)
d
OR 95% CI OR 95% CI
Co-morbidity factor
e
Anxiety/Depression/PTSD 1.03
b
1.00–1.06 1.05
c
1.03–1.08
Hypothyroidism/Other autoimmune disease 1.02 0.99–1.04 1.05
c
1.03–1.08
Narcolepsy 1.04 0.96–1.12 1.12
c
1.04–1.21
Neurological disease 1.02 0.98–1.06 1.04 1.00–1.09
Cancer malignancy 1.01 0.97–1.04 1.01 0.98–1.04
Notes:
a
Global initial symptom severity (range: 0–27) equals the sum of initial severity ratings for all nine CFS
symptoms.
b
P< 0.05.
c
P< 0.01.
d
Adjusted for site, age, sex, race, education, CFS duration.
e
Derived from principal component factor analysis (oblique rotation) of the 14 non-sex-specific conditions, which
yielded four co-morbidity clusters: (1) anxiety, depression, and PTSD; (2) hypothyroidism and other autoimmune
disease; (3) narcolepsy; and (4) neurological disease.
OR = log odds ratio; CI = confidence interval; PTSD = post-traumatic stress disorder.
8L. Bateman et al.
Results from the regression analysis for the association between change in global
CFS symptom severity over time and co-morbid conditions are presented in Table 6.
Individuals diagnosed with a co-morbid condition were more likely to have higher
global symptom severity change scores, which would loosely indicate the worsening
of CFS symptoms over time. The adjusted model –which accounted for demographic
variables and initial CFS symptom severity –showed that narcolepsy, PLMD or RLS,
sleep apnea, an autoimmune disease, fibromyalgia, and depression had the strongest
influence on the worsening of CFS symptoms, with statistically significant ORs
greater than 5.00. The associations between changes in CFS symptom severity over
time and both cancer malignancy and bipolar disorder were not statistically significant.
Table 5. Odds ratios (OR) and 95% confidence intervals (CI) for the association between co-
morbid conditions and current health and functioning.
Model 1 Model 2
(Crude) (Adjusted)
c
OR 95% CI OR 95% CI
Current health status
d
Fibromyalgia 0.32
b
0.26–0.40 0.34
b
0.25–0.48
Severe spine problem 0.77 0.58–1.03 0.68
a
0.50–0.94
Hypothyroidism 0.81 0.01–1.06 0.70
a
0.52–0.93
Other autoimmune disease 0.67
a
0.47–0.94 0.61
a
0.42–0.89
Type 2 diabetes or metabolic syndrome 0.46
b
0.30–0.70 0.40
b
0.25–0.62
Narcolepsy 0.42
a
0.21–0.85 0.43
a
0.20–0.91
Sleep apnea 0.65
b
0.48–0.88 0.66
a
0.47–0.93
PLMD or RLS 0.71
a
0.51–0.98 0.69
a
0.48–0.98
Cancer malignancy 0.94 0.67–1.32 0.91 0.62–1.34
Depression 0.77
a
0.60–0.99 0.72
a
0.54–0.95
Anxiety 0.73
a
0.56–0.94 0.71
a
0.53–0.95
PTSD 0.75 0.52–1.10 0.75 0.49–1.14
Bipolar disorder 1.84 0.82–4.14 1.53 0.67–3.48
Neurological disease 0.60
a
0.39–0.93 0.57 0.35–0.92
Any functioning (work, school, regular activity)
Fibromyalgia 0.62
b
0.47–0.81 0.51
b
0.35–0.72
Severe spine problem 0.54 0.67–1.23 0.88 0.62–1.27
Hypothyroidism 0.93 0.70–1.23 0.88 0.63–1.22
Other autoimmune disease 0.80 0.55–1.16 0.65 0.42–1.00
Type 2 diabetes or metabolic syndrome 0.54 0.33–0.86 0.64 0.37–1.11
Narcolepsy 0.66 0.30–1.47 0.53 0.21–1.34
Sleep apnea 1.03 0.75–1.42 0.97 0.65–1.42
PLMD or RLS 0.84
a
0.59–1.20 0.73 0.48–1.10
Cancer malignancy 0.75 0.52–1.08 1.00 0.65–1.54
Depression 1.27 0.97–1.65 0.90 0.65–1.23
Anxiety 1.05 0.80–1.37 0.88 0.63–1.22
PTSD 0.75 0.50–1.12 0.69 0.42–1.11
Bipolar disorder 0.73 0.32–1.66 0.50 0.21–1.23
Neurological disease 0.62 0.39–1.01 0.59 0.34–1.03
Notes:
a
P< 0.05.
b
P< 0.01.
c
Adjusted for site, age, sex, race, education, CFS duration, global initial CFS symptom severity.
d
Ordinal variable coded: 0 = worse; 1 = the same; 2 = better, compared to health status at CFS onset.
OR = log odds ratio; CI = confidence interval; PLMD = periodic limb movement disorder; RLS = restless leg
syndrome; PTSD = post-traumatic stress disorder.
Fatigue: Biomedicine, Health & Behavior 9
Discussion
The present study is one of the largest and most comprehensive assessments of con-
ditions and disorders co-morbid with chronic fatigue syndrome to date; only a
handful of previous studies [12,13,23] have assessed both medical and mental health
conditions co-morbid with CFS, and only one has assessed co-morbid symptoms.
[29] This large multi-site clinical epidemiology study in the US ascertained 960 indi-
viduals with a confirmed diagnosis of CFS and informed treatment history in one of
four geographically dispersed CFS specialty clinics. Data regarding the co-morbid con-
ditions, which occurred after the onset of CFS, were collected systematically using a
structured survey administered largely over the telephone. Stratifying by site revealed
demographic differences among the CFS patients, which may relate to the underlying
population in the different geographic locations. In addition, the clinical focus and
approach varied by site. However, it is difficult to isolate whether this is due to the inter-
est and/or expertise of each clinician or a direct response to the needs of the local CFS
patient community. More than likely, it is a combination of both factors.
A decade after illness onset, less than half of the participants are able to maintain the
activity required to work. An overwhelming majority (84%) of the study participants
reported at least one co-morbid condition after the onset of CFS, consistent with the
figures reported by Dansie and colleagues in a population-based sample of demographi-
cally similar individuals with CFS-like illness.[12] Consistent with a report by Aaron
Table 6. Odds ratios (OR) and 95% confidence intervals (CI) for the association between
change in global CFS symptom severity
a
over time and co-morbid conditions.
Model 1 Model 2
(Crude) (Adjusted)
d
OR 95% CI OR 95% CI
Change in global CFS symptom severity over
time
Fibromyalgia 5.75
c
2.31–13.92 6.61
c
2.44–17.87
Severe spine problem 8.06
c
2.90–22.42 10.90
c
4.36–27.25
Hypothyroidism 3.95
c
1.55–10.07 4.21
c
1.83–9.69
Other autoimmune disease 4.50
b
1.27–15.91 10.01
c
3.33–30.08
Type 2 diabetes or metabolic syndrome 6.99
b
1.58–31.03 5.24
c
1.44–19.13
Narcolepsy 18.56
b
1.32–261.91 31.72
c
3.38–297.97
Sleep apnea 23.13
c
7.88–67.90 13.94
c
5.20–37.37
PLMD or RLS 7.43
c
2.26–24.39 10.42
c
3.67–29.61
Cancer malignancy 0.94 0.27–3.24 0.84 0.28–2.55
Depression 16.83
c
1.48–8.96 5.41
c
2.42–12.11
Anxiety 3.84
c
1.53–9.62 4.50
c
1.95–10.34
PTSD 1.62 0.42–6.27 3.49
b
1.04–11.74
Bipolar disorder 0.57 0.04–7.71 2.68 0.29–24.61
Neurological disease 2.60 0.54–12.52 4.06
b
1.02–16.10
Notes:
a
Change (range: –27 to 18) equals the current global CFS symptom severity score minus the initial global
CFS symptom severity score.
b
P< 0.05.
c
P< 0.01.
d
Adjusted for site, age, sex, race, education, CFS duration, global initial CFS symptom severity.
OR = odds ratio; CI = confidence interval; PLMD = periodic limb movement disorder; RLS = restless leg
syndrome; PTSD = post-traumatic stress disorder.
10 L. Bateman et al.
and colleagues,[23] which assessed a number of co-morbid clinical conditions in a con-
trolled co-twin study, fibromyalgia was the most common co-morbid condition in our
sample and was statistically associated with reduced function in routine life activities.
Nearly 60% of our participants reported the emergence of mental health symptoms
(depression, anxiety, PTSD, and/or bipolar disorder), which is strikingly similar to
other published reports addressing clinic- [10–12] and community-based [30]
samples. The prevalence of mental health diagnoses in our sample is somewhat
lower than a prospective study of CFS patients by Wessely and colleagues [8] and
higher than a recent report from a Belgian sample of CFS patients.[13] The mental
health diagnoses reported by participants were not necessarily determined by diagnostic
criteria or structured interviews, but the high prevalence is sobering. Depression as a co-
morbid condition is notably high in patients with chronic illness (15–25%) compared to
healthy primary care patients (5–10%), and the highest rates (40–50%) are in patients
with neurological illness.[31] In our study, the higher prevalence of depression reported
by the Utah subsample could have resulted from differences in clinical approaches, for
example a physician who actively checks in with patients about mental health issues.
Furthermore, numerous studies have demonstrated that depression is more preva-
lent in women,[32,33] and the Utah subsample included a greater proportion of
women than the Florida and Nevada subsamples. However, the lower prevalence of
depression in the New York subsample, which relative to other sites comprised the
largest proportion of women, contradicts this explanation. The high rates of affective
disorders in our CFS subjects may reflect the severely disabling nature of the disorder,
high symptom burden, lower quality of life, and biological changes in the brain second-
ary to chronic illness.[34] Our subjects had been diagnosed and treated by CFS special-
ists, but a 2003 population-based study [35] showed that less than 20% of cases had
been diagnosed or treated for CFS by any physician, a situation which could addition-
ally increase the rates of secondary depression or anxiety.
The prevalence of any cancer in this sample of CFS patients was about 16%, which
is approximately four times the prevalence of cancer in the adult population in the US in
2009. Excluding skin cancers, which were most prevalent in the Florida and Nevada
subsamples, the prevalence of any cancer was about 8%, or twice the prevalence in
the 2009 US adult population.
1
According to the National Cancer Institute, certain
types of cancer may be increased in the CFS population, most notably non-Hodgkin’s
lymphomas (OR 1.34–1.51).[36] Clinical interest in mantle and B-cell lymphoma is
being pursued separately in a subsample of the present study.
Consistent with the notion that co-morbid conditions can result in greater levels of
fatigue or burden of illness,[24] participants in our sample with co-morbid conditions
had worse current health, lower levels of current functioning, and a slight worsening in
CFS symptoms over time. The latter was most strongly influenced by primary sleep dis-
orders (narcolepsy, PLMD or RLS, sleep apnea), autoimmune disease, fibromyalgia,
and depression. Prior work addressing the impact of psychiatric disorders on CFS out-
comes has been mixed, with poorer outcomes in some studies [24] but not others.[11]
Strengths and limitations
As noted above, the present study is among the largest and most comprehensive inves-
tigations to assess the development of co-morbid conditions in a population diagnosed
with CFS. The study survey elicited information on 17 specific co-morbid medical and
mental health conditions; our 960 participants also had the opportunity to provide
Fatigue: Biomedicine, Health & Behavior 11
information on other disorders. This study therefore, to our knowledge, collected data
on more co-morbid conditions in a larger sample of individuals with CFS than any
study published previously. In addition, we examined associations between the
number and type of co-morbid conditions and change in CFS symptom severity over
time, as well as indicators of current functioning.
This study, while both large and involving a well-defined population, is cross-sec-
tional and relies on retrospective reports. With this cross-sectional investigation, we
have created a large cohort with a unique opportunity for longitudinal follow-up. Indi-
viduals included in this multi-site clinical epidemiology study first presented to one of
four specialty CFS clinical sites, on average, approximately 15 years prior to partici-
pation. However, any impact of recall bias would only serve to diminish the observed
associations presented here. In addition, this study is not community-based and thus
comprises a largely female, relatively advantaged midlife population. However,
many published clinical CFS studies have similar sample demographics, rendering
possible meaningful comparisons.
Lastly, while pharmacological treatments may influence the occurrence and severity
of co-morbid conditions, we were unable to obtain complete treatment data for partici-
pants. Therefore the regression analyses in this study on co-morbid conditions did not
account for the effect of pharmacological interventions. However, a separate study on
the most effective treatments for CFS is planned.
Clinical significance
Regardless of the case definition employed to make a diagnosis, CFS is a serious unmet
medical need and public health issue. The prevalence is estimated to be 0.76%–3.28% in
the general population, based on clinical assessment and self-report respectively.[37]
Although direct treatments for CFS have not been identified, the common and significant
co-morbid conditions revealed in this study provide distinct targets for preventive and
supportive care. Primary sleep disorders, chronic widespread pain, metabolic syndrome,
and mental health symptoms are well-known conditions that warrant careful assessment
and treatment compatible with the primary underlying illness of CFS. More careful
screening and management of co-morbid conditions may also lead to better understand-
ing of the underlying risk factors, target organs, and pathophysiology of CFS as well.
Conclusion
In conclusion, this study adds valuable knowledge to the epidemiologic literature addres-
sing chronic fatigue syndrome, at once providing evidence of consistency with previously
published work and important new information on the prevalence and burden of co-morbid
conditions and their impact on thecourse of CFS and functional outcomes. Future research
may address a variety of questions affiliated with the overall burden of physical and mental
conditions co-morbid with CFS, including specific illness subtypes and potential endophe-
notypes, clinical management strategies, and treatment implications.
Acknowledgements
DM had full access to all the data in the study and takes responsibility for the integrity of the data
and the accuracy of the data analysis. Author contributions: LB, NK, DP, SML, and SAC devel-
oped the CFS questionnaire. LB, NK, DP, SML, AA, SAC, EB, and GG collected the data. SD
12 L. Bateman et al.
and DM performed the statistical analysis and, with the assistance of LB, interpreted the results.
All authors were involved in manuscript preparation, and approved the final manuscript. The
authors acknowledge the generous financial support of the Hutchins Family Foundation and
the Chronic Fatigue Initiative (CFI). We thank Scott Carlson, MBA, Peter Wolczynski, and
Stella Lee, MPH of CFI as well as Anthony Komaroff, MD of Brigham and Women’s Hospital
for their support, attendees of the 2014 IACFS/ME meeting in San Francisco, CA, and of course,
the CFS community, and the individuals who dedicated their time and effort to participating in
this study. Competing interests: the authors declare they have no competing interests.
Notes on Contributors
Lucinda Bateman, MD, is an internist and the Founder of the Fatigue Consultation Clinic in Salt
Lake City, Utah.
Salima Darakjy, MPH, is a data analyst in the Department of Epidemiology at Columbia
University.
Nancy Klimas, MD, is the Chair of the Department of Clinical Immunology at Nova Southeast-
ern University where she also serves as a Professor of Medicine and the Scientific Director of the
Institute for Neuro Immune Medicine.
Daniel Peterson, MD, is an internist in private practice in Incline Village, Nevada. He is the
Founder and a Scientific Advisory Board Member of Simmaron Research.
Susan M. Levine, MD, is an infectious disease specialist in private practice in New York City.
Ali Allen, RN, is a Senior Research Coordinator at the Fatigue Consultation Clinic in Salt Lake
City, Utah.
Shane A. Carlson, MA, is an Assistant Research Coordinator at the Fatigue Consultation Clinic
in Salt Lake City, Utah.
Elizabeth Balbin, M., is a Research Coordinator in the Department of Psychology and Behav-
ioral Medicine at the University of Miami.
Gunnar Gottschalk, B.S., is a Project Coordinator at Simmaron Research in Incline Village,
Nevada.
Dana March, Ph.D., is an Assistant Professor of Epidemiology at Columbia University.
Note
1. The most recent available data from the American Cancer Society (http://www.cancer/org/
cancer/cancerbasics/cancer-prevalence) and the US Census Bureau (via CDC Wonder;
http://wonder.cdc.gov/) estimates the 2009 prevalence of any cancer (all sites) in the US
is 4.1% (=12,549,000 persons with cancer/306,771,529 people in the US population).
The prevalence of any cancer in this CFS sample is 15.6% (=150/960). Excluding skin
cancer, the prevalence of any cancer in this sample is 8.1% (=78/960).
References
[1] Nakatomi Y, Mizuno K, Ishii A, Wada Y, Tanaka M, Tazawa S, Onoe K, Fukuda S, Kawabe J,
Takahashi K, Kataoka Y, Shiomi S, Yamaguti K, Inaba M, Kuratsune H, Watanabe Y.
Neuroinflammation in Patients with Chronic Fatigue Syndrome/Myalgic Encephalomyelitis:
An 11C-(R)-PK11195 PET Study. J Nucl Med. 2014;55(6):945–950.
[2] Fluge O, Bruland O, Risa K, Storstein A, Kristoffersen EK, Sapkota D, Naess H, Dahl O,
Nyland H, Mella O. Benefit from B-lymphocyte depletion using the anti-CD20 antibody ritux-
imab in chronic fatigue syndrome. A double-blind and placebo-controlled study. PLoS One.
2011;6(10):e26358.
[3] Cairns R, Hotopf M. A systematic review describing the prognosis of chronic fatigue syndrome.
Occup Med (Lond). 2005;55(1):20–31.
Fatigue: Biomedicine, Health & Behavior 13
[4] Ciccone DS, Chandler HK, Natelson BH. Illness trajectories in the chronic fatigue syndrome: a
longitudinal study of improvers versus non-improvers. J Nerv Ment Dis. 2010;198(7):486–493.
[5] Jason LA, Porter N, Brown M, Anderson V, Brown A, Hunnell J, Lerch A. CFS: A Review of
Epidemiology and Natural History Studies. Bull IACFS ME. 2009;17(3):88–106.
[6] Jason LA, Porter N, Hunnell J, Brown A, Rademaker A, Richman JA. A natural history study of
chronic fatigue syndrome. Rehabil Psychol. 2011;56(1):32–42.
[7] Crawley E, Collin SM, White PD, Rimes K, Sterne JA, May MT. Treatment outcome in adults
with chronic fatigue syndrome: a prospective study in England based on the CFS/ME National
Outcomes Database. QJM. 2013;106(6):555–565.
[8] Wessely S, Chalder T, Hirsch S, Wallace P, Wright D. Psychological symptoms, somatic symp-
toms, and psychiatric disorder in chronic fatigue and chronic fatigue syndrome: a prospective
study in the primary care setting. Am J Psychiatry. 1996;153(8):1050–1059.
[9] Torres-Harding SR, Jason LA, Cane V, Carrico A, Taylor RR. Physicians’diagnoses of psychia-
tric disorders for people with chronic fatigue syndrome. Int J Psychiat Med. 2002;32
(2):109–124.
[10] Ciccone DS, Busichio K, Vickroy M, Natelson BH. Psychiatric morbidity in the chronic fatigue
syndrome: are patients with personality disorder more physically impaired?. J Psychosom Res.
2003;54(5):445–452.
[11] Matsuda Y, Matsui T, Kataoka K, Fukada R, Fukuda S, Kuratsune H, Tajima S, Yamaguti K,
Kato YH, Kiriike N. A two-year follow-up study of chronic fatigue syndrome comorbid with
psychiatric disorders. Psychiatry Clin Neurosci. 2009;63(3):365–373.
[12] Dansie EJ, Furberg H, Afari N, Buchwald D, Edwards K, Goldberg J, Schur E, Sullivan PF.
Conditions comorbid with chronic fatigue in a population-based sample. Psychosomatics.
2012;53(1):44–50.
[13] Mariman A, Delesie L, Tobback E, Hanoulle I, Sermijn E, Vermeir P, Pevernagie D, Vogelaers
D. Undiagnosed and comorbid disorders in patients with presumed chronic fatigue syndrome. J
Psychosom Res. 2013;75(5):491–496.
[14] Johnson SK, DeLuca J, Natelson BH. Personality dimensions in the chronic fatigue syndrome: a
comparison with multiple sclerosis and depression. J Psychiatr Res. 1996;30(1):9–20.
[15] Henderson M, Tannock C. Objective assessment of personality disorder in chronic fatigue syn-
drome. J Psychosom Res. 2004;56(2):251–254.
[16] Courjaret J, Schotte CK, Wijnants H, Moorkens G, Cosyns P. Chronic fatigue syndrome and
DSM-IV personality disorders. J Psychosom Res. 2009;66(1):13–20.
[17] Nater UM, Jones JF, Lin JM, Maloney E, Reeves WC, Heim C. Personality features and person-
ality disorders in chronic fatigue syndrome: a population-based study. Psychother Psychosom.
2010;79(5):312–318.
[18] Kempke S, Luyten P, Claes S, Goossens L, Bekaert P, Van Wambeke P, Van Houdenhove B.
Self-critical perfectionism and its relationship to fatigue and pain in the daily flow of life in
patients with chronic fatigue syndrome. Psychol Med. 2013;43(5):995–1002.
[19] Reeves WC, Wagner D, Nisenbaum R, Jones JF, Gurbaxani B, Solomon L, Papanicolaou DA,
Unger ER, Vernon SD, Heim C. Chronic fatigue syndrome–a clinically empirical approach to its
definition and study. BMC Med. 2005;3:19.
[20] Jones JF, Lin JM, Maloney EM, Boneva RS, Nater UM, Unger ER, Reeves WC. An evaluation
of exclusionary medical/psychiatric conditions in the definition of chronic fatigue syndrome.
BMC Med. 2009;7:57.
[21] Christley Y, Duffy T, Martin CR. A review of the definitional criteria for chronic fatigue syn-
drome. J Eval Clin Pract. 2012;18(1):25–31.
[22] Friedberg F, Dechene L, McKenzie MJ, 2nd, Fontanetta R. Symptom patterns in long-duration
chronic fatigue syndrome. J Psychosom Res. 2000;48(1):59–68.
[23] Aaron LA, Herrell R, Ashton S, Belcourt M, Schmaling K, Goldberg J, Buchwald D. Comorbid
clinical conditions in chronic fatigue: a co-twin control study. J Gen Intern Med. 2001;16
(1):24–31.
[24] Wagner-Raphael LI, Jason LA, Ferrari JR. Chronic fatigue syndrome, chronic fatigue, and psy-
chiatric disorders: predictors of functional status in a national nursing sample. J Occup Health
Psychol. 1999;4(1):63–71.
[25] Fukuda K, Straus S, Hickie I, Sharpe M, Dobbins J, Komaroff A. The chronic fatigue syndrome:
a comprehensive approach to its definition and study. Ann Intern Med. 1994;121:953–959.
14 L. Bateman et al.
[26] Gunn PP, Fremont AM, Bottrell M, Shugarman LR, Galegher J, Bikson T. The Health Insurance
Portability and Accountability Act Privacy Rule: a practical guide for researchers. Med Care.
2004;42(4):321–327.
[27] Schmaling KB, Fiedelak JI, Katon WJ, Bader JO, Buchwald DS. Prospective study of the prog-
nosis of unexplained chronic fatigue in a clinic-based cohort. Psychosom Med. 2003;65
(6):1047–1054.
[28] Solomon L, Reeves WC. Factors influencing the diagnosis of chronic fatigue syndrome. Arch
Intern Med. 2004;164(20):2241–2245.
[29] Jason LA, Torres-Harding SR, Carrico AW, Taylor RR. Symptom occurrence in persons with
chronic fatigue syndrome. Biol Psychol. 2002;59(1):15–27.
[30] Nater UM, Lin JM, Maloney EM, Jones JF, Tian H, Boneva RS, Raison CL, Reeves WC, Heim
C. Psychiatric comorbidity in persons with chronic fatigue syndrome identified from the Georgia
population. Psychosom Med. 2009;71(5):557–565.
[31] Katon WJ. Clinical and health services relationships between major depression, depressive
symptoms, and general medical illness. Biol Psychiatry. 2003;54(3):216–226.
[32] Current depression among adults—United States, 2006 and 2008. MMWR Morbidity and mor-
tality weekly report. 2010;59(38):1229–35.
[33] Piccinelli M, Wilkinson G. Gender differences in depression. Critical review. Br J Psychiatry.
2000;177:486–492.
[34] Katon WJ. Epidemiology and treatment of depression in patients with chronic medical illness.
Dialogues Clin Neurosci. 2011;13(1):7–23.
[35] Reyes M, Nisenbaum R, Hoaglin DC, Unger ER, Emmons C, Randall B, Stewart JA, Abbey S,
Jones JF, Gantz N, Minden S, Reeves WC. Prevalence and incidence of chronic fatigue syn-
drome in Wichita, Kansas. Arch Intern Med. 2003;163(13):1530–1536.
[36] Chang CM, Warren JL, Engels EA. Chronic fatigue syndrome and subsequent risk of cancer
among elderly US adults. Cancer. 2012;118(23):5929–5936.
[37] Johnston S, Brenu EW, Staines D, Marshall-Gradisnik S. The prevalence of chronic fatigue syn-
drome/ myalgic encephalomyelitis: a meta-analysis. Clin Epidemiol. 2013;5:105–110.
Fatigue: Biomedicine, Health & Behavior 15
