Arthritis Care & Research

Vol. 65, No. 5, May 2013, pp 777–785

DOI 10.1002/acr.21931
© 2013, American College of Rheumatology
ORIGINAL ARTICLE

Fibromyalgia Prevalence, Somatic Symptom

Reporting, and the Dimensionality of
Polysymptomatic Distress: Results
From a Survey of the General Population
FREDERICK WOLFE,1 ELMAR BRÄHLER,2 ANDREAS HINZ,2 AND WINFRIED HÄUSER3

Objective. To evaluate fibromyalgia in the general population with emphasis on prevalence, dimensionality, and somatic
symptom severity.
Methods. We studied 2,445 subjects randomly selected from the German general population in 2012 using the American
College of Rheumatology 2010 preliminary diagnostic criteria for fibromyalgia, as modified for survey research, and the
polysymptomatic distress scale (PSD). Anxiety, depression, and somatic symptom severity were assessed with the Patient
Health Questionnaire (PHQ) series, and measures of symptoms and quality of life were assessed with the European
Organization for Research and Treatment of Cancer questionnaire.
Results. The prevalence of fibromyalgia was 2.1% (95% confidence interval [95% CI] 1.6, 2.7), with 2.4% (95% CI 1.5, 3.2)
in women and 1.8% (95% CI 1.1, 2.6) in men, but the difference was not statistically significant. Prevalence rose with age.
Fibromyalgia subjects had markedly abnormal scores for all covariates. We found smooth, nondisordered relationships
between PSD and all predictors, providing additional evidence against the hypothesis that fibromyalgia is a discrete
disorder and in support of a dimensional or spectrum disorder. There was a strong correlation (r ⴝ 0.790) between the
PSD and the PHQ somatic symptom severity scale; 38.5% of persons with fibromyalgia satisfied the proposed Diagnostic
and Statistical Manual of Mental Disorders, Fifth Edition criteria for a physical symptom disorder.
Conclusion. The modified 2010 diagnostic criteria do not result in high levels of fibromyalgia. PSD and fibromyalgia are
strongly related to somatic symptom severity. There is evidence in support of fibromyalgia as a dimensional or continuum
disorder. This has important ramifications for neurobiologic and epidemiology research, and for clinical diagnosis,
treatment, and ascertainment of disability.

INTRODUCTION tender point count ascertainment required by the ACR

The development of the 2010 American College of
Rheumatology (ACR) fibromyalgia criteria (1) and their
modification for survey research (2) made it possible to
conduct population-based research relating to fibromyal-
gia because the high costs and difficulties surrounding the
1
Frederick Wolfe, MD: National Data Bank for Rheumatic
Diseases and University of Kansas School of Medicine,
Wichita; 2Elmar Brähler, PhD, Andreas Hinz, PhD: Univer-
sität Leipzig, Leipzig, Germany; 3Winfried Häuser, MD:
Technische Universität München, Munich, and Klinikum
Saarbrücken, Saarbrücken, Germany.
Dr. Häuser has received consultant fees, speaking fees,
and/or honoraria (less than $10,000 each) from Daiichi
Sankyo and Abbott Germany.
Address correspondence to Frederick Wolfe, MD, 1035
North Emporia, Suite 288, Wichita, KS 67214. E-mail:
fwolfe@arthritis-research.org.
Submitted for publication October 31, 2012; accepted in
revised form December 11, 2012.
1990 criteria (3) were eliminated (4).
Soon after the publication of the criteria, it was sug-
gested that the 2 components of the 2010 criteria, the 0 –19
widespread pain index (WPI) and the 0 –12 symptom se-
verity (SS) score, could be combined by addition into a
0 –31 index. Originally called the “fibromyalgianess scale”
(5), a term that was a little awkward and limiting, it has
subsequently been termed the “polysymptomatic distress”
scale (PSD), a term first suggested by Wessely and Hotopf
(6). Patients who satisfy the 2010 criteria, defined by either
1) WPI ⱖ7/19 pain sites and SS score ⱖ5/12 (Type A) or
2) WPI between 3– 6/19 and SS score ⱖ9/12 (Type B), will
always have a score on the PSD scale of at least 12 (7 ⫹ 5
or 3 ⫹ 9). Thus, fibromyalgia can be mapped out on a
dimensional or continuum scale, allowing further explo-
ration of the fibromyalgia concept (7). Fibromyalgia differs
from the frequently studied chronic widespread pain con-
cept (8,9) by its inclusion of nonpain symptoms, including
severity measures of fatigue, unrefreshed sleep, cognitive
problems, and somatic symptom reporting. In addition,

777
778
Significance & Innovations
● Using a modification for surveys of the American
College of Rheumatology 2010 fibromyalgia crite-
ria, the prevalence of fibromyalgia in the general
German population was 2.1%.
● There is evidence that fibromyalgia is a continuum
disorder.
● Forty percent of fibromyalgia cases in the general
population satisfy proposed Diagnostic and Statis-
tical Manual of Mental Disorders, Fifth Edition
criteria for physical symptom disorder.
fibromyalgia is a medical diagnosis that is included in the
international list of diseases of the World Health Organi-
zation, in which chronic widespread pain is not. Chronic
widespread pain may be related to fibromyalgia by an
increase in somatic symptoms (10) and psychological dis-
tress (8).
In the study that follows, we examined several issues of
importance regarding fibromyalgia in the general popula-
tion using data from a 2012 German epidemiology survey.
In the first part of the report, we studied the prevalence of
fibromyalgia, examining the effect of age and sex on prev-
alence to determine whether prevalence was much greater
in women, as was the case in the clinic when tender points
were in use. We also considered whether the use of self-
report symptom criteria would lead to an inflation of di-
agnosis rates, and we considered whether the use of
Type B criteria would disturb the uniformity and homo-
geneity of the fibromyalgia diagnosis. The second major
issue of the study was to seek evidence about the “contin-
uum hypothesis,” whether fibromyalgia was best under-
stood as part of a continuum of polysymptomatic distress
or should be understood as a categorical disorder. Finally,
we evaluated somatic symptom reporting and intensity as
a central part of the fibromyalgia concept.
PATIENTS AND METHODS
Design and subjects. A representative sample of the
German general population was selected with the assis-
tance of a demographic consulting company (USUMA,
Berlin, Germany). The random selection was based on
multistage sampling. First, 320 sample point regions were
randomly drawn from the last political election register,
covering rural and urban areas from all regions in Ger-
many. The second stage was a random selection of house-
holds using the random route procedure (based on a start-
ing address). The third stage was a random selection of
household respondents with the Kish selection grid. The
sample was aimed to be representative in terms of age, sex,
and education for the general German population. The
inclusion criteria for the study were age ⱖ14 years and the
ability to read and understand the German language.
All subjects were visited by a study assistant and in-
Wolfe et al
formed about the investigation. Subjects were presented
with self-rating questionnaires. The survey included sev-
eral questionnaires on somatic and psychological symp-
toms (health survey) as well as questionnaires on eating,
political attitudes, and media use. The assistant waited
until the participants answered all questionnaires, and
offered help if persons did not understand the meaning of
questions.
Data collection took place between May and June 2012.
A first attempt was made at 4,448 addresses, and 2,515
persons (56.7%) participated fully. Reasons for nonpar-
ticipation included the following: 3 unsuccessful attempts
to contact the household or selected household member
(12.9%), the household or selected household member
declined to participate (13.7%), or the household mem-
ber was on a holiday break (1.1%). Furthermore, 0.5% of
the participants were excluded because they were not able
to follow the interview due to illness, as were 3.3% who
refused to finish the interview.
To study the adult population, we excluded subjects
who were ⬍18 years of age. Three subjects who had miss-
ing data for the fibromyalgia diagnosis variables were also
excluded. After these exclusions, a total of 2,445 subjects
were included in the analysis.
Ethics. All participants were informed about the study
procedures and signed an informed consent form. The
study was approved by the Institutional Ethics Review
Board of the University of Leipzig (Az 092-12-05032012).
Questionnaires. We assessed marital status, educa-
tional status, current professional status, and family in-
come by a sociodemographic questionnaire. We used a
slightly modified social class index that is used in rehabil-
itation care and surveys in Germany (11).
Fibromyalgia Survey Questionnaire (FSQ). The modi-
fied ACR 2010 diagnostic criteria were developed in a
longitudinal study of patients of the National Data Bank
for Rheumatic Diseases by substituting a count of 3 symp-
toms for the physician’s (range 0 –3) evaluation of the
extent of somatic symptom intensity by a questionnaire
assessing the number of pain sites and somatic symptom
severity (2). The FSQ included the SS score with 3 major
symptoms (fatigue, trouble thinking or remembering,
waking up tired [unrefreshed]), which can be coded 0 –3
(0 ⫽ no problem; 1 ⫽ slight or mild problems, generally
mild or intermittent; 2 ⫽ moderate, considerable prob-
lems, often present and/or at a moderate level; 3⫽ severe,
continuous, life-disturbing problems), and 3 additional
symptoms (pain or cramps in lower abdomen, depression,
headache), which can be coded as present (1) or absent (0)
(total subscore 0 –3). These 3 items are surrogates for the
somatic symptom burden item of the ACR 2010 criteria (1).
The SS score ranges from 0 –12. The WPI includes 19 pain
sites (both jaws, both shoulders, both upper arms, both
lower arms, both hips, both upper legs, both lower legs,
neck, chest, upper back, lower back, abdomen). It is also
possible to determine from the WPI pain sites if a patient
satisfies the ACR 1990 widespread pain criterion (3). The
sum of the WPI and the SS score constitutes the fibromy-
Prevalence, Dimensionality, and Somatic Symptom Severity of FM 779

algianess scale or PSD, a measure of physical and psycho- Table 1. The prevalence of fibromyalgia and mean
logical symptom intensity (distress) that can be applied to values of related variables in the general population*
subjects regardless of disease (12). We used the validated
German versions of the FSQ (13). Category Cases N Value (95% CI)
Subjects satisfy the fibromyalgia survey criteria if they
Fibromyalgia prevalence
meet the following 3 conditions: 1) WPI ⱖ7/19 pain sites All 52 2,445 2.1 (1.6, 2.7)
and SS score ⱖ5/12 or WPI between 3– 6/19 and SS score Women 31 1,308 2.4 (1.5, 3.2)
ⱖ9/12; 2) symptoms have been present at a similar level Men 21 1,137 1.8 (1.1, 2.6)
for at least 3 months; and 3) the patient does not have All cases
another disorder that would otherwise sufficiently explain Ages ⬍40 years 6 741 0.8 (0.2, 1.5)
Ages 40–59.9 years 22 895 2.5 (1.4, 3.5)
the pain (2). The conditions 1 and 2 can be assessed by the Ages ⱖ60 years 24 809 3.0 (1.8, 4.1)
FSQ in epidemiologic studies. Women
Ages ⬍40 years 6 392 1.5 (0.3, 2.8)
Patient Health Questionnaire (PHQ). The PHQ Somatic Ages 40–59.9 years 13 491 2.6 (1.2, 4.1)
Symptom Short Form (PHQ-SSS) is the short form of the Ages ⱖ60 years 12 425 2.8 (1.2, 4.4)
Men
PHQ-15 (14) and comprises 8 items (stomach or problems
Ages ⬍40 years 0 349 0.0 (0.0, 0.0)
going to the toilet; pain in the back; pain in the arms, legs, Ages 40–59.9 years 9 404 2.2 (0.8, 3.7)
or joints; headaches; chest pain or getting out of breath; Ages ⱖ60 years 12 384 3.1 (1.4, 4.9)
dizziness; feeling tired or having low energy; trouble PDS
sleeping), with each symptom scored from 1 (“not both- All 2,445 3.0 (2.9, 3.1)
ered at all”) to 5 (“bothered very strongly”) within the last Women 1,308 3.2 (3.0, 3.4)
Men 1,137 2.8 (2.6, 3.0)
7 days (14). The PHQ-15 is best characterized as a measure
All cases
of somatic symptom severity (14). Ages ⬍40 years 741 1.8 (1.6, 2.0)
The PHQ-SSS (not yet validated) had been proposed for Ages 40–59.9 years 895 2.9 (2.7, 3.2)
the research definition of criterion A (somatic symptoms: Ages ⱖ60 years 809 4.2 (3.9, 4.4)
one or more somatic symptoms that are distressing and/or Women
result in significant disruption in daily life) of the physical Ages ⬍40 years 392 2.1 (1.8, 2.4)
Ages 40–59.9 years 491 3.1 (2.8, 3.5)
symptom disorder of the Diagnostic and Statistical Manual Ages ⱖ60 years 425 4.3 (3.9, 4.7)
of Mental Disorders, Fifth Edition (DSM-V). Criterion A is Men
met if a person reports being bothered strongly or very Ages ⬍40 years 349 1.5 (1.2, 1.7)
strongly by a symptom (15). We assumed that criterion B Ages 40–59.9 years 404 2.7 (2.4, 3.0)
(disproportionate and persistent thoughts about the seri- Ages ⱖ60 years 384 4.0 (3.6, 4.4)
WPI
ousness of one’s symptoms) was met if a person reported at
All 2,445 1.3 (1.2, 1.4)
least one severe, continuous, life-disturbing problem in Women 1,308 1.4 (1.3, 1.5)
the 2010 criteria SS score. Men 1,137 1.2 (1.1, 1.4)
The PHQ-4 (16) is an ultra-brief self-report question- All cases
naire that consists of a 2-item depression scale (PHQ-2) Ages ⬍40 years 741 0.5 (0.4, 0.6)
and a 2-item anxiety scale (Generalized Anxiety Disorder). Ages 40–59.9 years 895 1.2 (1.1, 1.4)
Ages ⱖ60 years 809 2.2 (2.0, 2.4)
A score of 3 or greater on the depression subscale repre-
Women
sents a reasonable cut point for identifying potential cases Ages ⬍40 years 392 0.6 (0.4, 0.7)
of major depression or other depressive disorders, and a Ages 40–59.9 years 491 1.3 (1.1, 1.5)
score of 3 or greater on the anxiety subscale represents a Ages ⱖ60 years 425 2.3 (2.0, 2.6)
reasonable cut point for generalized anxiety, panic, social Men
anxiety, and posttraumatic stress disorders in primary care Ages ⬍40 years 349 0.4 (0.3, 0.5)
Ages 40–59.9 years 404 1.1 (0.9, 1.3)
settings and in the general population (17). The PHQ-4 Ages ⱖ60 years 384 2.2 (1.9, 2.4)
total score can serve as a measure for psychological dis- SS score
tress (14). We used the validated German version of the All 2,445 1.7 (1.6, 1.7)
PHQ (17). Women 1,308 1.8 (1.7, 1.9)
Health-related quality of life was assessed by the vali- Men 1,137 1.5 (1.4, 1.6)
All cases
dated German version of the European Organization for
Ages ⬍40 years 741 1.3 (1.2, 1.5)
Research and Treatment of Cancer (EORTC) QLQ-C30, Ages 40–59.9 years 895 1.7 (1.6, 1.8)
version 3.0 (18). The EORTC QLQ-C30 includes 5 func- Ages ⱖ60 years 809 1.9 (1.8, 2.0)
tioning scales, which cover the dimensions physical func- Women
tioning (5 items), role functioning (2 items), emotional Ages ⬍40 years 392 1.5 (1.4, 1.7)
functioning (4 items), cognitive functioning (6 items), and Ages 40–59.9 years 491 1.8 (1.7, 2.0)
Ages ⱖ60 years 425 2.0 (1.8, 2.2)
social functioning (2 items). In addition, the questionnaire
Men
contains a scale for global health status (2 items), and Ages ⬍40 years 349 1.1 (0.9, 1.3)
symptom scales that are relevant to cancer patients (e.g., Ages 40–59.9 years 404 1.6 (1.4, 1.8)
pain, nausea, fatigue, dyspnea). For functional and global Ages ⱖ60 years 384 1.8 (1.6, 2.0)
quality of life scales, higher scores mean a better level of
* 95% CI ⫽ 95% confidence interval; PDS ⫽ polysymptomatic distress
functioning. For symptom-oriented scales, a higher score scale; WPI ⫽ widespread pain index; SS ⫽ symptom severity.
means more severe symptoms (19).
780 Wolfe et al

Table 2. Clinical and demographic variables for full population and fibromyalgia groups*

All All FM (ⴙ) FM group A FM group B P,

(n ⴝ 2,445) (n ⴝ 52) (n ⴝ 43) (n ⴝ 9) A vs. B

Age, years (range 18–91 years) 50.2 ⫾ 17.4 59.8 ⫾ 16.4 60.7 ⫾ 16.9 55.7 ⫾ 13.7 0.410
Sex, % male 46.5 39.2 38.1 44.4 0.723
Social class, % lower 8.6 22.0 26.2 0.0 0.087
Depression score (PHQ-2) (range 0–6) 0.7 ⫾ 1.0 3.2 ⫾ 1.4 3.0 ⫾ 1.2 4.0 ⫾ 2.1 0.967
Depression (PHQ ⱖ3), % 6.0 63.5 61.9 77.8 0.366
Anxiety disorder score (GAD-2) (range 0–6) 0.5 ⫾ 0.9 2.6 ⫾ 1.5 2.4 ⫾ 1.4 3.7 ⫾ 1.7 0.015
Anxiety disorder (GAD-2 ⱖ3), % 2.4 50.0 42.9 88.9 0.012
Psychological distress (PHQ-4) (range 0–12) 1.2 ⫾ 1.7 5.8 ⫾ 2.5 5.4 ⫾ 2.1 7.7 ⫾ 3.6 0.014
Somatic symptom (PHQ-SSS) (range 0–26) 3.3 ⫾ 4.0 15.9 ⫾ 5.1 15.3 ⫾ 5.0 18.6 ⫾ 5.1 0.082
EORTC physical function (range 100–0) 92.2 ⫾ 15.0 53.1 ⫾ 25.5 56.3 ⫾ 24.0 37.8 ⫾ 27.9 0.046
EORTC cognitive function (range 100–0) 93.6 ⫾ 14.5 57.0 ⫾ 24.8 59.3 ⫾ 51.8 46.2 ⫾ 27.4 0.154
EORTC quality of life (range 100–0) 75.0 ⫾ 20.4 36.2 ⫾ 19.2 37.2 ⫾ 19.2 31.5 ⫾ 19.4 0.420
EORTC pain (range 0–100) 16.7 ⫾ 24.2 70.9 ⫾ 26.6 69.0 ⫾ 25.6 79.6 ⫾ 30.9 0.284
EORTC sleep status (range 0–100) 12.4 ⫾ 23.3 52.3 ⫾ 35.4 49.2 ⫾ 37.0 66.7 ⫾ 23.6 0.182
EORTC fatigue (range 0–100) 15.5 ⫾ 21.7 71.7 ⫾ 26.3 68.3 ⫾ 26.2 87.7 ⫾ 21.8 0.044
ACR widespread pain index (range 0–19) 1.3 ⫾ 2.3 8.9 ⫾ 3.5 9.6 ⫾ 3.0 4.3 ⫾ 1.3 0.000
ACR symptom severity score (range 0–12) 1.7 ⫾ 1.9 7.4 ⫾ 2.0 6.8 ⫾ 1.6 10.2 ⫾ 1.0 0.000
Polysymptomatic distress (range 0–27) 3.0 ⫾ 3.3 16.4 ⫾ 3.5 16.8 ⫾ 3.6 14.6 ⫾ 1.8 0.077

* Values are the mean ⫾ SD unless otherwise indicated. FM ⫽ fibromyalgia; PHQ-2 ⫽ 2-item Patient Health Questionnaire; GAD-2 ⫽ 2-item
Generalized Anxiety Disorder; PHQ-SSS ⫽ PHQ Somatic Symptom Short Form; EORTC ⫽ European Organization for Research and Treatment of
Cancer; ACR ⫽ American College of Rheumatology.

Statistical methods. We used predicted values from sec-
ond-degree fractional polynomial (FP) regression (20) to
describe the PSD and covariate relationships, after first
examining each covariate pair by lowess regression and
scatter plots to be sure the FP was an accurate and fair
representation of the underlying data. Comparisons in
Table 1 used t-tests and chi-square tests, as appropriate.
Data were analyzed using Stata, version 12.1.
RESULTS
Prevalence of fibromyalgia and component variable
levels. The study population was representative of the
German population in 2011 in terms of age and sex (21)
(Supplementary Appendix A, available in the online ver-
sion of this article at http://onlinelibrary.wiley.com/doi/
10.1002/acr.21931/abstract). The overall prevalence of fi-
bromyalgia was 2.1% (95% confidence interval [95% CI]
1.6, 2.7) (Table 1). Fibromyalgia increased with age: 0.8%
(95% CI 0.2, 1.5), 2.5% (95% CI 1.4, 3.5), and 3.0% (95%
CI 1.8, 4.1) in those ⬍40, 40 –59.9, and ⱖ60 years of age,
respectively (P ⫽ 0.004). Fibromyalgia was not statistically
more common in women than men (2.4% versus 1.8%;
P ⫽ 0.372 and 59.6% versus 40.4% among those with
fibromyalgia). We examined the 3 components of fibromy-
algia shown in Table 1. In regression analyses, the WPI
increased by 0.45 (95% CI 0.40, 0.50) units per 10-year
increase in age. Adjusted for age, the WPI for women was
0.17 (95% CI ⫺0.01, 0.34) units greater than in men (P ⫽
0.062). The SS score increased by 0.14 (95% CI 0.10, 0.19)
units per 10-year increase in age. Adjusted for age, the SS
score was 0.28 (95% CI 0.13, 0.43) units greater in women
than in men. The summary PSD score increased by 0.59
(95% CI 0.52, 0.67) units per 10 years and was 0.45 (95%
CI 0.17, 0.72) units greater in women than in men. Overall,
these data show that the prevalence of fibromyalgia and its
components increases with age, but that sex has a smaller
effect that is not significant for fibromyalgia or WPI, but is
significant for SS score and PSD.
Effect of Type A and B definitions. Of patients satisfy-
ing the modified ACR 2010 criteria, 82.7% satisfied it
using the Type A definition and 17.3% with the Type B
definition. As mandated by the Type A and Type B defi-
nitions that required WPI to be lower and SS scores to be
higher, SS scores were significantly higher for Type B
subjects (mean ⫾ SD 10.2 ⫾ 1.0 versus 6.8 ⫾ 1.6) (Table 2),
while WPI scores were greater in Type A subjects (mean ⫾
SD 9.6 ⫾ 3.0 versus 4.3 ⫾ 1.3). PSD scores were not
significantly different (P ⫽ 0.077). In addition, anxiety,
general psychological distress, physical function, and fa-
tigue were worse in the Type B group. Figure 1 shows the
distribution of the scores for both groups, with Type B
scores generally spread over the full ranges of fibromyalgia
patient scores. The 1990 widespread pain criterion was
noted in 82.7% overall, and in 93.0% of Type A and 33.3%
of Type B subjects.
Continuum hypothesis for fibromyalgia. By definition,
subjects with fibromyalgia (2.1%) had to have a PSD score
of at least 12, a level met by 3.4% of all subjects when
nonfibromyalgia subjects were included (Figure 2). How-
ever, a cut point of 13 better distinguished fibromyalgia
from nonfibromyalgia patients, for if a score of 12 (or 13)
were assumed to represent fibromyalgia for research pur-
poses, the respective kappa statistics for agreement would
be 0.758 (95% CI 0.678, 0.839) and 0.795 (95% CI 0.715,
0.875), respectively, indicating full agreement (22) with
Prevalence, Dimensionality, and Somatic Symptom Severity of FM 781

Figure 1. The distribution of fibromyalgia syndrome (FMS) covariate scores for subjects meeting Type A and Type B FMS criteria. PHQ ⫽
Patient Health Questionnaire; GAD ⫽ Generalized Anxiety Disorder.

the ACR 2010 fibromyalgia criteria designation. As cutoff

values of 12 or 13 suggested a dimensional disorder, we
examined the relationship between various covariates and
PSD to observe for linear association or graphic disconti-
nuities. Figures 2 and 3 demonstrate continuous, generally
linear, relationships throughout the relevant ranges that
were all significant at P ⬍ 0.001. In addition, the figures
show no evidence of inflection points in the area of PSD ⫽
12 or 13. The strength of the relationships between PSD
and covariates is shown in Table 3.

Somatic symptoms and fibromyalgia variables. The

strongest correlation between PSD and non-PSD depen-
dent variables was with the PHQ-SSS (r ⫽ 0.790) (Figure 2
and Table 3); the correlation between PSD and other co-
variates is shown in the table. For diagnosis of fibromyal-
gia, the area under the receiver operating curve was 0.973
(95% CI 0.958, 0.988) for PHQ-SSS and 0.996 (95% CI
0.995, 0.998) for PSD. The difference was significant at P ⫽
0.002. Except for the WPI and SS scores, the association
between the PHQ-SSS was slightly stronger for all covari-
ates in Table 3 compared with PSD. Overall, these data
show the strong similarities as well as the differences
between PSD and PHQ-SSS.
Figure 2. The distribution of polysymptomatic distress scale
(PSD) scores (yellow line) and the relationship between the Pa-
tient Health Questionnaire (PHQ) Somatic Symptom Short Form
and PSD in the general population. Blue points are subjects neg-
ative for fibromyalgia syndrome (FMS) and red points are subjects
who satisfy the FMS criteria. The vertical line at 12 is the point at
which cases of FMS first appear. Scores ⱖ12 are found in 3.4% of
subjects and FMS criteria are met by 2.1% of the population. A
small amount of random noise is added to the FMS-negative cases
to aid visualization.
782 Wolfe et al

Figure 3. The relationship between polysymptomatic distress scale (PSD) and relevant covariates. The grey area represents 95%
confidence intervals. The P values for the fractional polynomial regression models were ⬍0.001 for all models. There is no evidence of
discontinuity or important nonlinearity over the relevant ranges. BMI ⫽ body mass index.

With respect to somatic symptoms, 53.8% of fibromyal-

gia subjects had ⱖ1 severe somatic symptoms and 32.7% Table 3. Correlations between polysymptomatic distress,
had ⱖ2 such symptoms on the PHQ-SSS: criterion A of the PHQ-SSS, and study covariates*
proposed DSM-V physical symptoms disorder. Twenty-
Variable PDS PHQ-SSS
one subjects (40.4%) noted at least one severe problem in
the 2010 criteria SS score, an approximation of criterion B. PDS 1.000 0.790
Twenty subjects (38.5%) noted at least one severe problem Widespread pain index 0.880 0.635
in the 2010 criteria SS score and ⱖ1 severe somatic symp- Symptom severity score 0.810 0.714
toms on the PHQ-SSS. GAD-2 anxiety score 0.531 0.550
PHQ-SSS 0.790 1.000
EORTC pain score 0.629 0.691
DISCUSSION EORTC quality of life ⫺0.593 ⫺0.633
EORTC physical function ⫺0.575 ⫺0.629
In 2008, a similar German population study used a differ- EORTC emotional function ⫺0.556 ⫺0.621
ent version of the WPI and an earlier non-ACR version of PHQ-2 depression score 0.555 0.574
fibromyalgia criteria (23) that resulted in an estimated EORTC physical role ⫺0.541 ⫺0.579
fibromyalgia prevalence of 3.8% (24,25). In this (2012) EORTC social function ⫺0.534 ⫺0.611
nationwide population study, we used the modified 2010 EORTC financial difficulties 0.459 0.536
ACR fibromyalgia criteria and found that the prevalence of Social class ⫺0.165 ⫺0.159
Body mass index 0.130 0.146
fibromyalgia was 2.1% (95% CI 1.6, 2.7). The small num-
ber of Type B cases (17.3%) with their associated increases * PHQ-SSS ⫽ Patient Health Questionnaire Somatic Symptom
in distress covariates indicates that Type B was a minor Short Form; PDS ⫽ polysymptomatic distress scale; GAD-2 ⫽
2-item Generalized Anxiety Disorder; EORTC ⫽ European Organi-
component of the fibromyalgia definition that tapped into zation for Research and Treatment of Cancer.
the SS scores appropriately, as the criteria designers had
Prevalence, Dimensionality, and Somatic Symptom Severity of FM
anticipated. The Type B criterion was added to capture
those persons who had all of the symptoms of fibromyalgia
but not enough painful areas. Prior to the 1990 ACR crite-
ria, the requirement for painful regions and tender points
was less strict (26). The 2010 criteria recognized this and
the clinical need to slightly relax requirements imposed by
the 1990 criteria.
The current study noted that the mean level of WPI and
the prevalence of fibromyalgia did not differ significantly
by sex, although there was a nonsignificant female pre-
dominance (59.6% versus 40.4%) among those with fibro-
myalgia (P ⫽ 0.372); for women, fibromyalgia prevalence
was 2.4%, compared with 1.8% in men. Other population
studies have found that rates of widespread pain for the
sexes were similar, and that there was only a modest
increase of fibromyalgia in women compared with men
(25,27). By contrast, in clinical populations women con-
stitute as many as 90% of patients. Increased rates in
clinical populations have been attributed to the use of the
tender point count and sex differences in health-seeking
behavior. George Ehrlich noted, however, that “No one has
FM until it is diagnosed . . . it has to be named by a doctor
to exist” (28). This observation underscores the difference
between the number of people who satisfy criteria for
fibromyalgia and the number who actually seek care for
symptoms that are cast by physicians or patients as fibro-
myalgia.
We made use of the PSD scale to investigate whether
fibromyalgia was a categorical or dimensional disorder.
The PSD scale arose after the publication of the 2010
criteria, and was formed by adding together the 2 compo-
nents of the 2010 fibromyalgia diagnosis, the WPI and the
SS score. Persons with fibromyalgia always have mini-
mum levels of PSD of at least 12 regardless of whether they
satisfy the Type A or Type B criterion. This allows fibro-
myalgia to be marked as a point on the PSD continuum.
The PSD scale is linearly correlated with all measures of
physical and psychological distress over relevant ranges
(Figures 2 and 3).
In deciding whether a disorder should be considered a
categorical or a dimensional illness, Kessler indicated that
“The critical test should be whether the predictors are
consistently related to differences in symptom severity
across the full relevant range of the dimensional distribu-
tion. If they are, then analysis of the dimensional version
of the symptom scale makes most sense . . .” (29). We
found the following strongly supportive of dimensional
illness. Figure 2 shows that fibromyalgia onset (PSD ⫽ 12)
lies at the end of the PSD continuum; 96.6% of all persons
in the general population have values less than 12. PSD is
also correlated with dimensional measures of distress in
the current study and in other samples (30). Figure 3,
which examines Kessler’s dictum using fractional polyno-
mial regression that is sensitive to nonlinearity, demon-
strates a smooth relationship between PSD and all relevant
covariates.
In a comprehensive review of fibromyalgia, Wessely and
Hotopf wrote that they were “unaware of any study” that
supported the view that fibromyalgia was a categorical
disorder (6), and wrote that “fatigue and myalgia syn-
dromes are arbitrarily created syndromes that lie at the
extreme end of the spectrum of polysymptomatic distress.
783
Definitive evidence to support or refute this view will
come from primary care or community samples, not the
study of specialist populations.” Croft et al posed the
question: “More pain, more tender points: is fibromyalgia
just one end of a continuous spectrum?” (31). In a study of
clinical data, Wolfe concluded that “Tender points are
linearly related to fibromyalgia variables and distress, and
there is no discrete enhancement or perturbation of fibro-
myalgia or distress variables associated with very high
levels of tender points” (32). The current study provides
the type of evidence for fibromyalgia as the continuum that
Wessely and Hotopf stated was necessary.
Although fibromyalgia in clinical practice is treated as
a categorical disorder, the confirmation of fibromyalgia
as a spectrum disorder has important research, clinical,
and societal connotations: “A study that takes the extreme
end of the spectrum, represented by selected samples of
patients referred to rheumatology or pain services, and
compares them with non-fatigued controls, will produce
a . . . categorical solution but for spurious reasons” (6).
Future research, whether clinical or neurobiologic, must
account for the continuum nature of fibromyalgia to be
valid.
The ACR 2010 criteria definition of fibromyalgia was
constructed based on specialist physicians’ (experts) diag-
nosis of fibromyalgia using the ACR 1990 criteria. How-
ever, there was no gold standard for patients designated as
having fibromyalgia in the 1990 criteria study, as the pres-
ence of a fixed high number of tender points was not
required by most previous definitions (33–35), including
the definition by Yunus et al in their 1981 criteria (26).
Although one might conclude that the 1990 definition
of fibromyalgia was reasonable, it was also arbitrary.
With that in mind, it is problematic to sustain the idea that
a specific cut point accurately and validly separates dis-
ease from nondisease or illness from nonillness, particu-
larly in view of the continuous nature of polysymptomatic
distress.
There is considerable uncertainty about the nature of
fibromyalgia. Virtually every type of mental illness has
been found to be more common in fibromyalgia (36). Neu-
robiologic abnormalities and central sensitization have
also been suggested as causes of fibromyalgia (37). At the
other end of the spectrum, many disbelieve in the idea of
fibromyalgia, citing flaws in the conceptualization of the
illness as well as the continuous nature of polysymptom-
atic distress (28,38 – 45). One contribution to the nature of
fibromyalgia based on this study, in addition to the dem-
onstration of the continuous nature of polysymptomatic
distress, is the exploration of the relationship of somatic
symptom reporting and fibromyalgia. We found that PSD
was strongly correlated with the PHQ-SSS (r ⫽ 0.790). In
addition, 20 subjects (38.5%) noted at least one severe
problem in the 2010 criteria SS score and ⱖ1 severe so-
matic symptoms on the PHQ-SSS. From these data, it
would appear that approximately 40% of persons with
fibromyalgia satisfy the proposed DSM-V criteria for phys-
ical symptom disorder. The idea that fibromyalgia is pri-
marily a somatic symptom disorder is well accepted in
the pain and psychological literature (46 – 48), but not in
the rheumatology community, where the dominant idea is
central sensitization (49).
784
An important strength of this study is the high-quality
survey in which 56.7% of the requested population agreed
to participate. Participation rates that high are difficult to
get, and a recent Mayo Clinic epidemiology study that
used the 2010 criteria had 27.7% participation (16.2% in
the age group 21–39 years) (50). A potential limitation of
our study was that we did not exclude subjects whose
symptoms were less than 3 months in duration. We made
this choice in order to be able to consider nonfibromyalgia
subjects for whom the 3-month criterion was inappropri-
ate. Even so, only one fibromyalgia patient would have
been excluded by the 3-month requirement, and the prev-
alence did not change at the first decimal point by keeping
this subject. Another limitation is that we did not know
if there were relevant clinical differences between study
participants and nonparticipants. We also made no exclu-
sions for the presence of “another disorder that would
otherwise sufficiently explain the pain.” This is a contro-
versial requirement because it is not certain how to define
such disorders. Even so, our epidemiology studies, as with
most fibromyalgia epidemiology studies, did not have suf-
ficient data to make such exclusions.
In summary, we found evidence to support the idea that
fibromyalgia is a dimensional disorder, not a categorical
disorder. Sex differences had limited effect on prevalence,
and approximately 40% of fibromyalgia patients satisfied
proposed DSM-V criteria for a physical symptom disorder.
The modified fibromyalgia criteria that rely on self-report
can be a useful research tool for epidemiology and clinical
research.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors ap-
proved the final version to be published. Dr. Wolfe had full access
to all of the data in the study and takes responsibility for the
integrity of the data and the accuracy of the data analysis.
Study conception and design. Brähler, Hinz, Häuser.
Acquisition of data. Brähler, Hinz.
Analysis and interpretation of data. Wolfe, Brähler, Häuser.
REFERENCES
1. Wolfe F, Clauw DJ, FitzCharles MA, Goldenberg DL, Katz RS,
Mease P, et al. The American College of Rheumatology pre-
liminary diagnostic criteria for fibromyalgia and measurement
of symptom severity. Arthritis Care Res (Hoboken) 2010;62:
600 –10.
2. Wolfe F, Clauw DJ, Fitzcharles MA, Goldenberg DL, Hauser
W, Katz RS, et al. Fibromyalgia criteria and severity scales for
clinical and epidemiological studies: a modification of the
ACR preliminary diagnostic criteria for fibromyalgia. J Rheu-
matol 2011;38:113–22.
3. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C,
Goldenberg DL, et al. The American College of Rheumatology
1990 criteria for the classification of fibromyalgia: report of
the Multicenter Criteria Committee. Arthritis Rheum 1990;33:
160 –72.
4. McBeth J, Mulvey MR. Fibromyalgia: mechanisms and poten-
tial impact of the ACR 2010 classification criteria. Nat Rev
Rheumatol 2012;8:108 –16.
5. Wolfe F. New American College of Rheumatology criteria for
fibromyalgia: a twenty-year journey [editorial]. Arthritis Care
Res (Hoboken) 2010;62:583– 4.
6. Wessely S, Hotopf M. Is fibromyalgia a distinct clinical entity?
Wolfe et al
Historical and epidemiological evidence. Baillieres Best Pract
Res Clin Rheumatol 1999;13:427–36.
7. Hauser W, Schmutzer G, Brahler E, Glaesmer H. A cluster
within the continuum of biopsychosocial distress can be la-
beled “fibromyalgia syndrome”: evidence from a represen-
tative German population survey. J Rheumatol 2009;36:
2806 –12.
8. Benjamin S, Morris S, McBeth J, Macfarlane GJ, Silman AJ.
The association between chronic widespread pain and mental
disorder: a population-based study. Arthritis Rheum 2000;43:
561–7.
9. McBeth J, Macfarlane GJ, Hunt IM, Silman AJ. Risk factors
for persistent chronic widespread pain: a community-based
study. Rheumatology (Oxford) 2001;40:95–101.
10. McBeth J, Macfarlane GJ, Benjamin S, Silman AJ. Features of
somatization predict the onset of chronic widespread pain:
results of a large population-based study. Arthritis Rheum
2001;44:940 – 6.
11. Winkler J. Measurement of social status by an index of
the health surveys of DHP. Hamburg: RKI-Schriften I. 1998:
69 –74.
12. Wolfe F. Fibromyalgianess [editorial]. Arthritis Rheum 2009;
61:715– 6.
13. Hauser W, Jung E, Erbsloh-Moller B, Gesmann M, Kohn-
Becker H, Petermann F, et al. Validation of the Fibromyalgia
Survey Questionnaire within a cross-sectional survey. PLoS
One 2012;7:e37504.
14. Kroenke K, Spitzer RL, Williams JB. The PHQ-15: validity of
a new measure for evaluating the severity of somatic symp-
toms. Psychosom Med 2002;64:258 – 66.
15. DSM-5 development. J00 somatic symptom disorder. Wash-
ington, DC: American Psychiatric Association; 2012. URL:
http://www.dsm5.org/ProposedRevisions/Pages/proposed
revision.aspx?rid⫽368.
16. Kroenke K, Spitzer RL, Williams JB, Lowe B. An ultra-brief
screening scale for anxiety and depression: the PHQ-4. Psy-
chosomatics 2009;50:613–21.
17. Lowe B, Wahl I, Rose M, Spitzer C, Glaesmer H, Wingenfeld
K, et al. A 4-item measure of depression and anxiety: valida-
tion and standardization of the Patient Health Questionnaire-
4 (PHQ-4) in the general population. J Affect Disord 2010;122:
86 –95.
18. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A,
Duez NJ, et al. The European Organization for Research and
Treatment of Cancer QLQ-C30: a quality-of-life instrument for
use in international clinical trials in oncology. J Natl Cancer
Inst 1993;85:365–76.
19. Favers PM, Aaronson NK, Bjordal K, Groenvold M, Curran D,
Bottomly A, et al. The EORTC QLQ-C30 scoring manual. 3rd
ed. Brussels: EORTC; 2001.
20. Royston P, Sauerbrei W. Multivariable model-building: a
pragmatic approach to regression analysis based on fractional
polynomials for modelling continuous variables. Chichester:
Wiley; 2008.
21. Federal Statistical Office. Statistical yearbook. 2012. URL:
https://www.destatis.de/DE/Publikationen/Statistisches
Jahrbuch/StatistischesJahrbuch2012.pdf;jsessionid⫽2CD0AD
FB63B227342A9DAD900C72E9D5.cae2?_blob⫽publicationFile.
22. Landis JR, Koch GG. The measurement of observer agreement
for categorical data. Biometrics 1977;33:159 –74.
23. Katz RS, Wolfe F, Michaud K. Fibromyalgia diagnosis: a com-
parison of clinical, survey, and American College of Rheuma-
tology criteria. Arthritis Rheum 2006;54:169 –76.
24. Hauser W, Schmutzer G, Glaesmer H, Brahler E. Prevalence
and predictors of pain in several body regions: results of a
representative German population survey. Schmerz 2009;23:
461–70. In German.
25. Branco JC, Bannwarth B, Failde I, Abello Carbonell J, Blotman
F, Spaeth M, et al. Prevalence of fibromyalgia: a survey in five
European countries. Semin Arthritis Rheum 2010;39:448 –53.
26. Yunus M, Masi AT, Calabro JJ, Miller KA, Feigenbaum SL.
Primary fibromyalgia (fibrositis): clinical study of 50 patients
with matched normal controls. Semin Arthritis Rheum 1981;
11:151–71.
27. Coster L, Kendall S, Gerdle B, Henriksson C, Henriksson KG,
Prevalence, Dimensionality, and Somatic Symptom Severity of FM
Bengtsson A. Chronic widespread musculoskeletal pain: a
comparison of those who meet criteria for fibromyalgia and
those who do not. Eur J Pain 2008;12:600 –10.
28. Ehrlich GE. Pain is real; fibromyalgia isn’t [editorial]. J Rheu-
matol 2003;30:1666 –7.
29. Kessler RC. The categorical versus dimensional assessment
controversy in the sociology of mental illness. J Health Soc
Behav 2002;43:171– 88.
30. Wolfe F, Rasker JJ. The Symptom Intensity Scale, fibromyal-
gia, and the meaning of fibromyalgia-like symptoms. J Rheu-
matol 2006;33:2291–9.
31. Croft P, Burt J, Schollum J, Thomas E, Macfarlane G, Silman
A. More pain, more tender points: is fibromyalgia just one end
of a continuous spectrum? Ann Rheum Dis 1996;55:482–5.
32. Wolfe F. The relation between tender points and fibromyalgia
symptom variables: evidence that fibromyalgia is not a dis-
crete disorder in the clinic. Ann Rheum Dis 1997;56:268 –71.
33. Freyberg RH. Non-articular rheumatism. Bull N Y Acad Med
1951;27:245–58.
34. Graham W. The fibrositis syndrome. Bull Rheum Dis 1953;3:
33– 4.
35. Traut EF. Fibrositis. J Am Geriatr Soc 1968;16:531– 8.
36. Fietta P, Manganelli P. Fibromyalgia and psychiatric disor-
ders. Acta Biomed 2007;78:88 –95.
37. Dadabhoy D, Clauw DJ. Therapy insight: fibromyalgia—a dif-
ferent type of pain needing a different type of treatment. Nat
Clin Pract Rheumatol 2006;2:364 –72.
38. Block SR. On the nature of rheumatism. Arthritis Care Res
1999;12:129 –38.
39. Hadler NM. “Fibromyalgia” and the medicalization of misery.
J Rheumatol 2003;30:1668 –70.
785
40. Hadler NM, Greenhalgh S. Labeling woefulness: the social
construction of fibromyalgia. Spine 2005;30:1– 4.
41. Cohen ML, Quintner JL. Fibromyalgia syndrome, a problem of
tautology. Lancet 1993;342:906 –9.
42. Cohen ML, Quintner JL. The derailment of railway spine: a
timely lesson for post-traumatic fibromyalgia syndrome. Pain
Rev 1996;3:181–202.
43. Cohen ML, Quintner JL. Fibromyalgia syndrome and dis-
ability: a failed construct fails those in pain. Med J Aust
1998;168:402– 4.
44. Ehrlich GE. Fibromyalgia, a virtual disease. Clin Rheumatol
2003;22:8 –11.
45. Wolfe F. Fibromyalgia wars. J Rheumatol 2009;36:671– 8.
46. Kroenke K. Somatoform disorders and recent diagnostic con-
troversies. Psychiatr Clin North Am 2007;30:593– 619.
47. Fink P, Schroder A. One single diagnosis, bodily distress
syndrome, succeeded to capture 10 diagnostic categories of
functional somatic syndromes and somatoform disorders.
J Psychosom Res 2010;68:415–26.
48. Smith RC, Dwamena FC. Classification and diagnosis of pa-
tients with medically unexplained symptoms. J Gen Intern
Med 2007;22:685–91.
49. Yunus MB. Central sensitivity syndromes: a new paradigm
and group nosology for fibromyalgia and overlapping condi-
tions, and the related issue of disease versus illness. Semin
Arthritis Rheum 2008;37:339 –52.
50. Vincent A, Lahr BD, Wolfe F, Clauw DJ, Whipple MO, Oh TH,
et al. Prevalence of fibromyalgia: a population-based study in
Olmsted County, Minnesota, utilizing the Rochester Epidemi-
ology Project. Arthritis Care Res (Hoboken) 2013;65:786 –92.