FSS begets FSS

1. The number of existing functional somatic syndromes (FSSs) is an important risk
factor for new, different FSSs
John W. Warren a,b,
⁎, Patricia Langenberg b
, Daniel J. Clauw c
a
Department of Medicine, University of Maryland School of Medicine, Baltimore, MD, USA
b
Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD, USA
c
Departments of Anesthesiology and Medicine, University of Michigan School of Medicine, Ann Arbor, MI, USA
a b s t r a c ta r t i c l e i n f o
Article history:
Received 30 April 2012
Received in revised form 5 September 2012
Accepted 6 September 2012
Keywords:
Functional somatic syndromes
Fibromyalgia
Chronic fatigue syndrome
Irritable bowel syndrome
Migraine
Chronic pelvic pain
Objective: The objective of this study is to test the hypothesis that the number of functional somatic syn-
dromes (FSSs) predicts new, additional FSSs.
Methods: In a recent case–control study of interstitial cystitis/painful bladder syndrome (IC/PBS), we used
symptom-based consensus definitions to identify these FSSs: fibromyalgia (FM), chronic fatigue syndrome
(CFS), irritable bowel syndrome (IBS), chronic pelvic pain, migraine, sicca syndrome and panic disorder. Those
present before the incidence year were called antecedent FSSs; those with onset during the incidence year
were called incident FSSs. In each of two groups, 312 IC/PBS cases and 313 controls, rates of incident FSSs
were compared among those with 0, 1, 2, or ≥3 antecedent FSSs. Confounding was assessed using logistic regres-
sion analyses that included the individual antecedent FSSs, published correlates of these FSSs, and demographic
variables.
Results: The incidence of a new FSS increased with the number of antecedent FSSs, as did that of incident FM,
CFS and IBS studied separately. These findings were not confounded by other variables. The presence of multiple
antecedent FSSs generally had the highest odds ratio for new, different, incident FSSs.
Conclusions: This study revealed that the number of antecedent FSSs was among the strongest risk factors for
other FSSs, especially incident FM, CFS and IBS. This suggests that the FSSs are linked through a polysyndromic
phenotype. If each FSS is heterogeneous, to seek a pathogenesis common to all FSSs, individuals with multiple
FSSs should be sought; to seek a pathogenesis unique to a specific FSS, mature persons who have only that FSS
should be studied.
© 2012 Elsevier Inc. All rights reserved.
Introduction
Functional somatic syndromes (FSSs) are often discussed as a group
because they have many similarities: absent or incidental local patholo-
gy, non-diagnostic laboratory tests, over-representation in women, as-
sociation with childhood adversities, exacerbation with stress and
menstruation, chronicity, co-morbidity and unknown etiology [1]. Fi-
bromyalgia (FM), chronic fatigue syndrome (CFS) and irritable bowel
syndrome (IBS) are almost always included as FSSs; chronic pelvic
pain (CPP) and migraine sometimes are; and sicca syndrome and
panic disorder have similar characteristics including association with
these syndromes [1].
Interstitial cystitis/painful bladder syndrome (IC/PBS) is a chronic
disorder diagnosed by bladder pain and urinary symptoms that has
many of the characteristics of an FSS and, indeed, some experts include
it in discussions of FSSs [1,2]. In a recent case–control study of over 600
women, Events Preceding Interstitial Cystitis (EPIC), we showed that 11
syndromes were risk factors for IC/PBS; these syndromes included the
seven above plus depression, allergies, asthma and vulvodynia [3].
These antecedent syndromes generally were associated with each
other, e.g. a woman who had FM was more likely to have IBS than if
she did not have FM [3]. The risk of IC/PBS increased as the number of
these antecedent syndromes increased [4]. Indeed, the presence of mul-
tiple syndromes had the largest odds ratio for IC/PBS of any variable
tested and this association was not confounded by other variables [5].
These findings generated the hypothesis that patients who had
IC/PBS and multiple syndromes shared a pathogenesis that was
common to IC/PBS and each of the other syndromes [4,6]. If such a
shared pathogenesis existed, then a corollary hypothesis would be
that the presence of multiple syndromes is a marker for the shared
pathogenesis and would emerge as a risk factor for each of the
remaining syndromes, just as it had for IC/PBS. The study design
allowed testing of this hypothesis.
Methods
Methods for EPIC have been presented in detail [3,4,7–10]. Briefly,
1177 women were screened throughout the United States through
Journal of Psychosomatic Research 74 (2013) 12–17
⁎ Corresponding author at: Department of Medicine, University of Maryland School
of Medicine, 10 South Pine Street, #900, Baltimore, MD 21201, USA.
E-mail address: jwarren@medicine.umaryland.edu (J.W. Warren).
0022-3999/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jpsychores.2012.09.002
Contents lists available at SciVerse ScienceDirect
Journal of Psychosomatic Research
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.

2. patient support and physician organizations to enroll 312 adult female
incident IC/PBS cases (≤12 months of IC/PBS symptoms). Inclusion
criteria were pain perceived to be from the bladder and two of urgency,
frequency and nocturia; exclusion criteria were known diseases with
similar symptoms. Iterative questions and medical record review iden-
tified the IC/PBS symptom onset date (index date). Recruitment of con-
trols was by national residential random digit telephone dialing; as
described [3], 34,889 telephone calls resulted in 756 eligible women
who were interviewed until each case was matched on gender, age,
and national region. The matched control was assigned an index date
at an equivalent interval before the interview. As suggested by the in-
terviewers, 78% of controls chose a personally important date such as
a birthday or anniversary that was within one month of the assigned
index date [3]. The mean age at the index date of women with IC/PBS
was 42.3 years and 42.9 years for controls; their other characteristics
have been well described [9,10].
All participants were interviewed by telephone between 2004 and
2008 about pre‐index date medical history and exposures [3]. Previous
uncontrolled studies had shown that several syndromes or diseases
appeared to be correlates of IC/PBS [11]. To determine whether any
might be risk factors for IC/PBS, we asked for a self-report of physician
diagnosis before the index date of 23 medical conditions. Additionally,
we used symptom-based consensus definitions constructed by experts
[12–18] for seven of these syndromes: FM, CFS, IBS, sicca, CPP, migraine,
and panic disorder. For ease of discussion, these seven syndromes will
be termed FSSs. For each, a screening question that led to a diagnostic
algorithm was: “At any time in your life before your index date, have
you had (pertinent symptom) for (time or number of episodes)”. The
onset of each FSS was then queried as any time before the index date
and, if so, within 12 months before the index date. This 12 month peri-
od became the incidence year for each participant. Thus for each of
these seven syndromes, we were able to measure the incidence of
that syndrome that year, i.e., the number of women with onset during
the incidence year divided by the number of women who did not
have the syndrome at the beginning of the year. Syndromes present be-
fore the incident year were called antecedent FSSs. Those with onset
during the incidence year were called incident FSSs.
Non-FSS variables included demographics and published correlates of
these syndromes. Demographic variables were index date age, race, cur-
rently divorced, education (high school graduation v. ≥some college),
and household income (b v. ≥US$50,000/year). Published correlates of
FSSs that were queried as present prior to the incidence year included
surgery (total number, each of several pelvic surgeries, and any pelvic
surgery), childbirth (ever and number), and automobile accident that
required medical care. This study was approved by the University of
Maryland Baltimore Institutional Review Board.
Statistics
The number of participants was determined by power analysis for the
parent study, i.e., identification of risk factors for IC/PBS. Groupings of an-
tecedent FSSs into 0, 1, 2, and ≥3 were driven by the small numbers of
individuals in the latter group, particularly among controls. Rates of inci-
dent FSSs during the incidence year were compared among those with 0,
1, 2, or ≥3 antecedent FSSs by Mantel–Haenszel chi-square tests of trend
and Pearson chi-square. Individual antecedent FSSs as well as the demo-
graphic and other non-FSS antecedents were tested for association with
incident FSSs by McNemar's chi square, Fisher's exact, or t-test. Those
that distinguished participants with an incident FSS from those without
(p≤.05) were included in logistic regression analyses.
Results
During the incidence year, 25 (8%) of the 313 controls reported onset of 34 FSSs.
Twenty controls developed symptoms of a single FSS. Five reported onset of multiple
FSSs: two controls developed two FSSs each; two had onset of three FSSs, and one expe-
rienced four incident FSSs. Among the 312 women who would soon develop symptoms
of IC/PBS, 75 (24%) developed 96 FSSs. Sixty of these IC/PBS cases experienced onset of
a single incident FSS. Fifteen developed multiple FSSs: 11 IC/PBS cases reported symptoms
of two FSSs, two had onset of three FSSs, and two cases developed four incident FSSs. By
the end of the incidence year two women, both IC/PBS cases, reported all seven of these
FSSs: one was a woman with six antecedent FSSs that developed sicca during that year;
the other was a woman with antecedent IBS, CPP, and panic who experienced the onset
of FM, CFS, sicca, and migraine during the incidence year.
Table 1 shows the risk of an incident FSS by the number of antecedent FSSs. In both
controls and IC/PBS cases, the chance that a woman would experience onset of an FSS
during the incident year increased with the number of antecedent FSSs. For instance,
among controls 4.5% of the 178 women with no antecedent FSSs developed an incident
FSS compared to 20% with ≥3 antecedent FSSs.
But these rates did not adequately demonstrate the prominence of multiple anteced-
ent FSSs as a risk factor for a different, incident FSS. This is of course because those who
already had many FSSs were at risk for fewer incident FSSs than women who had no or
few antecedent FSSs. For instance, a woman that was diagnosed with 6 antecedent FSSs
was at risk for only one incident FSS while a woman with no antecedent FSSs theoretically
was at risk to develop all 7 incident FSSs. To more adequately reflect this phenomenon, we
manufactured an additional rate: the number of incident FSSs per number of “possible in-
cident FSSs”. The number of “possible incident FSSs” for each woman was 7 minus her
number of antecedent FSSs. For example, a woman with 2 antecedent FSSs was at risk
to develop the other 5 incident FSSs; i.e., her number of “possible incident FSSs” was 5.
In Table 1 in the last column, among controls the rate of incident FSSs per number of “pos-
sible incident FSSs” for those with ≥3 antecedent FSSs was about 10 times that of those
with no antecedent FSSs, i.e., a much greater difference than the factor of about 4 resulting
from the same comparison in the third column.
Table 2 shows the incidence of each individual FSS by the number of antecedent
FSSs. A test of trend is significant for incident CFS, IBS and sicca in controls and for in-
cident FM, CFS, IBS, and sicca in IC/PBS cases.
To determine whether the prediction of incident FSSs by the number of antecedent
FSSs was confounded by other factors, we sought bivariate associations with additional
variables: each of the individual antecedent FSSs as well as demographic, surgical, repro-
ductive and automobile accident antecedents, a total of 24 variables. Those associated
with each of the incident FSSs are shown in Table 3. For incident FSSs affecting six or
more women, we were able to perform logistic regression analyses. The variables in
Table 3 were entered into two sets of analyses. The first used the individual antecedent
FSS(s) as variables. The second set instead used the number of antecedent FSSs as a vari-
able. Table 4 shows that the presence of ≥3 antecedent FSSs predicted the incidence of
an FSS in both controls and cases, predicted incident FM and CFS in controls, and predicted
incident FM, CFS and IBS in cases. And the odds ratios (ORs) of ≥3 antecedent FSSs were
generally the largest of the associated variables.
These seven syndromes were studied because at the time the study began, these were
the ones that had been most strongly associated with IC/BPS. We acknowledge that some
of these are not generally thought to be FSSs. Thus, we replicated Table 1 using only the
best established FSSs, i.e., FM, CFS, and IBS. The findings were consistent. Among the
312 IC/PBS cases, 12 had all 3 as antecedent FSSs. This left 300 cases susceptible to one
or more of these as an incident FSS. Among cases with none of these as antecedents, the
incidence was 20/194 (10%), with one 11/75 (15%), and with two 7/31 (23%) (p=.05).
Controls had fewer antecedent and incident FSSs. Only two had all three as antecedents.
Of the 311 remaining susceptible women, the incidence in those with none of antecedent
Table 1
The risk of an incident functional somatic syndrome (FSS) by the number of antecedent
FSSs
Number of
antecedent FSSs
Number of
women
No. of women
(%) with ≥1
incident FSS
No. of “possible”
incident FSSsa
No. of incident
FSSs (%)b
Cases
Zero 102 18 (17.6%) 714 22 (3.1%)
One 81 18 (22.2%) 486 23 (4.7%)
Two 63 17 (27.0%) 315 22 (7.0%)
≥Three 66 22 (33.3%)#
229 29 (12.7%)
p valuec
.016 Not applicable
Controls
Zero 178 8 (4.5%) 1246 9 (0.7%)
One 85 10 (11.8%)#
510 17 (3.3%)
Two 30 3 (10.0%) 150 3 (2.0%)
≥Three 20 4 (20.0%)#
72 5 (6.9%)
p value .007 Not applicable
#
p≤.05 compared to zero antecedent FSS.
a
The number of “possible” incident FSSs for each woman is seven (the number of
FSSs being studied) minus the number of her antecedent FSSs.
b
The number of incident FSSs is the total number experienced by each group of
women; some women had more than one incident FSS. The percentage is the number
of total incident FSSs divided by the number of total “possible” incident FSSs in each
group of women.
c
Test of trend.
13J.W. Warren et al. / Journal of Psychosomatic Research 74 (2013) 12–17
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.

3. FM, CFS, or IBS was 13/258 (5%), with one of these 4/41 (10%), and with two 1/12 (8%)
(p=.27). The proportion of “possible incident FSSs” was 23/582 (4%) for cases with no an-
tecedent FSSs, with one 13/150 (9%) and with two 7/31 (23%); and was 17/774 (2%) for
controls with no antecedent FSSs, with one 5/82 (6%), and with two 1/12 (8%).
Discussion
The incidence of another, different FSS, especially CFS, FM, and IBS,
increased with larger numbers of antecedent FSSs. These associations
were not confounded by and were generally stronger than the associ-
ations of other antecedent variables, many of which had already been
published as risk factors for these FSSs. This association was seen in
both IC/PBS cases and healthy controls. In fact, a control with ≥3 an-
tecedent FSSs had almost five times the odds of developing an FSS as
one with no antecedent FSS.
This finding has relevance to investigations of the pathogenesis of
the FSSs. In a provocative and widely-cited article, in 1999 Wessely et
al. stated that “the current practice of regarding patients with diagnoses
of different functional somatic syndromes as having clinically signifi-
cant differences is questionable. Rather, such patients may have similar
conditions or variants of a general functional somatic syndrome” [19].
This opinion piece prompted a controversy about FSSs that has evolved
into a debate between “lumpers” and “splitters” [1,20–25].
Recent observations have strengthened both camps. On the side of
the lumpers, most of the FSSs are now known to be widely associated
with each other, i.e., to occur as co-morbidities more often than expected
by chance. On the other side, although splitting was originally discussed
as distinguishing between FSSs, phenotypes have been revealed that split
even within an FSS. In a recent opinion paper in this journal, White
juxtaposed these findings as a paradox and concluded: “we need both
to split the FSS(s) and to lump them together, rather than to do one or
the other” [25].
At first glance, our observation that the number of FSSs is a risk fac-
tor for another FSS appears to be substrate for the lumpers. However,
we believe that this finding favors both insights and indeed provides
an explanation for how lumping and splitting of FSSs concur.
Lumpers: the FSSs are associated with each other
Increasingly over the last 30 years investigators have shown the FSSs
to be epidemiologically linked, i.e. if a person has one s/he is likely to
have others. These reports usually have focused on pairs of FSSs [summa-
rized in [2]]. Yunus et al. [26], Whorwell et al. [27], and Hudson et al. [28]
were among the first to demonstrate associations among numerous FSSs.
Others have confirmed and extended these findings [29–32]. For in-
stance, Schur et al. assessed the relationships among 9 FSSs: of 36 possi-
ble pair-wise associations, 31 exceeded chance expectations [31]. It
appears that FSSs, in some patients at least, may comprise a network of
inter-related disorders.
The important question is how the FSSs are related. One can begin
thinking about this with the simplest type of association, that of two
FSSs. This construct generates only two hypotheses: FSS1 initiates a pro-
cess that leads to FSS2, or a preceding process leads to each of FSS1 and
FSS2. Either process might be structural, physiological, environmental,
behavioral, or another type and with or without a genetic, epigenetic
or experiential underpinning. Investigations of individual prevalent
FSSs have revealed candidates for such processes: central augmentation
of pain processing via central sensitization and other spinal and
supraspinal processes, autonomic dysfunction, and anomalies of the hy-
pothalamus–pituitary–adrenal axis [33–42].
The question of which came first, the process or the syndrome, has
not been answered for any of these associations. One can reason that
as the number of associated FSSs increases, the number of possible com-
binations and permutations of associating processes, including conver-
gence and divergence of causative effects, becomes overwhelming.
Splitters: each FSS is heterogeneous
When White summarized the evidence for heterogeneity within an
FSS, he used CFS as an example because it was the most intensively stud-
ied [25]. Indeed, among CFS patients a variety of statistical methods have
Table 2
The risk of individual incident functional somatic syndromes (FSSs) by the number of antecedent FSSs
Number of antecedent FSSs Incident FSSs
FM CFS IBS Sicca CPP Migraine Panic
Cases
Zero 1/102 (1.0) 4/102 (3.9) 3/102 (2.9) 0/102 (0.0) 11/102 (10.8) 3/102 (2.9) 0/102 (0.0)
One 2/72 (2.8) 3/79 (3.8) 3/72 (4.2) 1/81 (1.2) 8/61 (13.1) 4/64 (6.3) 2/57 (3.5)
Two 4/52 (7.7)#
5/55 (9.0) 3/41 (7.3) 2/62 (3.2) 5/42 (11.9) 1/28 (3.6) 2/35 (5.7)#
≥Three 5/28 (17.9)#
6/32 (18.8)#
4/26 (15.4)#
4/61 (6.6)#
6/25 (24.0) 3/25 (12.0) 1/32 (3.1)
p value (trend) b.001 .005 .017 .006 .166 .115 .090
Controls
Zero 1/178 (0.6) 2/178 (1.1) 2/178 (1.1) 0/178 (0.0) 1/178 (0.6) 1/178 (0.6) 2/178 (1.1)
One 5/76 (6.6)#
5/83 (6.0)#
2/73 (2.7) 2/85 (2.4)#
0/72 (0.0) 1/53 (1.9) 2/68 (2.9)
Two 0/23 (0.0) 2/28 (7.1)#
1/23 (4.3) 0/29 (0.0) 0/21 (0.0) 0/11 (0.0) 0/15 (0.0)
≥Three 0/12 (0.0) 2/6 (33.3)#
1/10 (10.0)#
1/20 (5.0)#
0/9 (0.0) 0/7 (0.0) 1/8 (13.0)#
p value (trend) .382 b.001 .044 .048 .513 .817 .093
#
p≤.05 compared to zero antecedent FSS.
Table 3
Antecedent variables associated (p≤.05) with each incident functional somatic syn-
drome (FSS) on bivariable analyses
Incident
FSS
Antecedent variables
Other FSSs Non-FSS variables
Cases
FM CPP Age, ovarian surgery, divorced, low household income
CFS FM, IBS, sicca,
CPP
Low educational level, bilateral oophorectomy, low
household income, childbirth
IBS FM Low educational level, low household income
Sicca IBS, CPP None
CPP IBS (inverse) Ovarian surgery, bilateral oophorectomy, other
gynecologic surgery
Migraine None Age, bilateral oophorectomy, low household income
Panic CFS, sicca None
Controls
FM Migraine Automobile accident
CFS FM, migraine Age, hysterectomy, number of surgeries, divorced, low
household income
IBS CPP Low household income
Sicca IBS None
CPP None Race, divorced
Migraine IBS None
Panic Migraine None
14 J.W. Warren et al. / Journal of Psychosomatic Research 74 (2013) 12–17
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.

4. been applied to its symptoms, signs and laboratory findings to divide the
syndrome into up to five phenotypes [43–50]; additionally, a genomic
analysis yielded 7 genotypes [51]. Some of these studies assessed the va-
lidity of the subgroups by showing differences among them in variables
not used in the initial classification schemes [43,45,49–54]. Other FSSs
have been similarly explored, but less intensively. For instance, among
FM patients, pain profiles, quantitative sensory testing, tender points,
and psychological status have been the criteria used to determine pheno-
types [55–59]. IBS patients have been subgrouped by neurophysiologic
measures [60], migraineurs by migraine symptoms [61–63], and patients
with panic disorder by panic attack symptoms [64,65]. Implied but not
often stated in most of these studies is that different phenotypes may
have different pathogeneses.
Our observation: the risk of an incident FSS increased with the number of
antecedent FSSs
This finding predicts the existence of a distinctive subgroup of indi-
viduals with multiple FSSs. Such a polysyndromic phenotype has been
revealed in numerous studies of CFS heterogeneity [43–45,47,48,50]
and in large scale population-based studies investigating the prevalence
of FSSs [31,32]. Among the latter, in a study of almost 4000 American
twins, Schur et al. used latent class analysis to reveal 4 classes, grouped
by gender and syndrome. One, of 2% of the population, had high preva-
lence of each of the 9 syndromes queried; 89% of this subgroup were
women [31]. Kato et al. in a Swedish twin study of >28,000 participants
also used latent class analysis and grouped individuals by symptoms
into 5 classes [32]. The smallest, about 3% of the population, had multi-
ple FSSs, specifically FM, CFS, IBS, depression and anxiety; 76% were
women.
Our observation then is confirmatory and may help to understand
White's paradox [25]. One can rephrase the paradox as a question:
how can the FSSs each be heterogeneous yet associated with each
other? Parsing this question yields this logic: heterogeneity of an FSS
is defined by two or more phenotypes; our finding is that a given FSS
predicts additional FSSs; therefore, this phenotype links this FSS to
other FSSs. Interestingly, after their population-based latent class anal-
yses, Kato et al. similarly concluded that their small polysyndromic
group was “the source of high comorbidity among these FSSs” [32]. In-
deed, they explored the use of the terminology [1] that distinguishes be-
tween an uncomplicated FSS (i.e., a single FSS) and complicated FSS
(i.e., multiple FSSs).
Continuing this logic leads to a research strategy. If there is a patho-
genesis common to all the FSSs, investigators should search for it in in-
dividuals with multiple FSSs. If there is a pathogenesis unique to a
specific FSS, researchers should seek it in persons who have only that
FSS.
However, nature is elegant but complex. Cross sectional studies will
not answer the question which came first, the pathophysiology or the
FSS? In regard to the former, the prevailing opinion is that the FSSs
have a strong common, genetic contribution, perhaps best seen in
twin studies [24,32]. Like in most other chronic diseases, this is almost
certainly polygenic [66]. Just as other polygenic disorders differ widely
in their expression from one individual to another based on the interac-
tion of genes and the complex environmental factors that trigger and
exacerbate these illnesses, the same is likely to be true of the FSSs.
The study has certain strengths. First, we used symptom-based
criteria for diagnosis of the FSSs, not the less sensitive method of
self-report of physician diagnosis. Second, similar findings were ob-
served in each of two groups of women, i.e., IC/PBS cases and their con-
trols, which suggests that these findings might be generalizable. We
anticipated that the incidence of each FSS would be higher in IC/PBS
cases than controls because we had already discovered that the preva-
lence of each FSS at the index date was higher in IC/PBS cases [3]. The dif-
ferences between IC/PBS cases and controls in specific antecedent FSSs
that were associated with incident FSSs (Tables 3 and 4) might have
been from random chance, that the year before onset of IC/PBS in some
way increased the incidence of certain FSSs, or that the IC/PBS cases
and controls were different samples of the population. Third, previous in-
vestigators have shown FSSs to be risk factors for other FSSs. Hamilton et
al. demonstrated that functional somatic syndromes were diagnosed in
the three years before the index (diagnosis) date in significantly more
IBS and CFS cases than in matched controls [67] (although they did not
report numbers of antecedent syndromes). Four, our other findings are
similar to those reported in prior investigations. For instance, the inci-
dences of individual FSSs in our controls were comparable to those
found by others in adult American women [68–72]. Additionally, the
non-FSS variables that were risk factors for specific FSSs in our partici-
pants (Table 3) have been reported previously. For FM, these were age,
divorce, and low household income [73,74], hysterectomy [75] and
physical trauma [76]. Those for CFS included gynecologic surgeries [77]
and childbirth [67,77]. Lower income has been correlated with migraine
[78].
But the study has several weaknesses as well. First, the syndromes
that could be studied were only those seven for which we could identify
Table 4
Odd ratios by logistic regression analyses for incident FSSs by antecedent FSSs (individually and by number) and other variables
Incident FSSs Antecedent FSSs and other variables
Cases Controls
Each FSS OR (CI) Number
of FSSs
OR (CI) Each FSS OR (CI) Number
of FSSs
OR (CI)
Any FSS FM 2.4 (1.3, 4.7) ≥3 FSSs 2.6 (1.2, 5.5) Migraine 3.8 (1.6, 9.1) ≥3 FSSs 4.6 (1.2, 17)
Income 1.8 (1.01, 3.1) Income 1.9 (1.1, 3.3) Income 3.2 (1.2, 8.5) Income 3.1 (1.2, 8.1)
FM CPP 9.7 (2.3, 41) ≥3 FSSs 30.6 (3.1, 305) Migraine 7.5(1.3, 45) 1 FSS 11.7 (1.3, 105)
Divorce 8.7 (1.5, 49) divorce 7.7 (1.3, 46) Autoacc 4.8 (.8, 29) Autoacc 6.2 (1.01, 37)
CFS FM 10.2 (2.8, 37) ≥3 FSSs 9.0 (1.6, 50) Migraine 10.7 (2.4, 47) ≥3 FSSs 20.1 (1.8, 222)
Income 5.3 (1.6, 17) Income 5.1 (1.6, 16) Income 11.0 (1.2, 97) Income 9.0 (1.1, 77)
IBS FM 4.8 (1.4, 16) ≥3 FSSs 5.9 (1.2, 28)
Sicca CPP 18.8(2.2, 159) None Not done
CPP Othergynsurg 4.5 (1.9, 11) Othergynsurg 5.0 (2.0, 12) Not done
Migraine Oophorex 12.4 (2.1, 73) Oophorex 9.6 (1.5, 62) Not done
Panic Not done Not done
Income = low household income.
Autoacc = auto accident.
Othergynsurg = any other surgery to your urinary or reproductive organs or pelvic region that I have not mentioned (those “mentioned” were uterine, ovarian or bladder
surgeries).
Oophorex = bilateral oophorectomy.
Not done = for fewer than 5 women with an incident FSS, logistic regression analyses were not done.
15J.W. Warren et al. / Journal of Psychosomatic Research 74 (2013) 12–17
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.

5. the year of onset. Three of these, i.e., FM, CFS, and IBS, are widely accept-
ed as FSSs; when analyses are limited to these three, the findings are
consistent. Some investigators may question the inclusion of some or
all of the other four syndromes. However, each of these latter syn-
dromes has all the characteristics of an FSS, including epidemiologic as-
sociation with the venerable FM, CFS and IBS. Furthermore, over the
years, designation of a syndrome as an FSS has been based on judgment,
rather than facts such as a known pathogenesis. Additionally, we did not
want to change the study design ex post facto. Second, we did not ex-
amine these participants, which may limit the generalizability of the
study. However, the diagnosis of each of these FSSs was based upon
the reporting of symptoms; whether that is done face to face or by tele-
phone may be immaterial but has not been studied. Another weakness
is the small numbers of each of the incident FSSs. These findings need to
be corroborated in larger studies, but the identification of statistically
significant antecedents to these few incident, specific FSSs suggests
that the risk factors identified here may have quite strong effects. A
fourth weakness is that the study was retrospective. We feel reasonably
comfortable that most participants could accurately place the end of the
incidence year: for cases, it was a fulcrum in their lives, i.e., the onset of a
painful bladder disease; for 78% of controls, it was a personally impor-
tant date in the prior year. But the beginning of the incidence year, i.e.,
12 months earlier, may have been less clear in the minds of both
cases and controls. However, we thought that they probably could re-
member whether one FSS preceded another; if they could not, we rea-
soned that bias would be towards the null hypothesis. Another
weakness is that this study was restricted to women; whether these
findings apply to men is unknown. A final limitation is that although
we discussed heterogeneity within each FSS to explain the observa-
tions, we did not study this phenomenon.
Conflict of interest
All authors have completed the unified competing interest form
and declare no support from any other organization than those in
the Acknowledgments for the submitted work; no financial relation-
ships with any organizations than those in the Acknowledgments
that might have an interest in the submitted work in the previous
three years; and have no other relationships or activities that could
appear to have influenced the submitted work.
Acknowledgments
This study was funded by the National Institutes of Health (NIH)
[National Institute of Diabetes, Digestive and Kidney Diseases, (R01
DK 064880 and U01 DK 066136)] and the Interstitial Cystitis Associa-
tion. The sponsors had no role in the design, execution, interpretation,
or decision to publish any aspect of the project.
References
[1] Henningsen P, Zipfel S, Herzog W. Management of functional somatic syndromes.
Lancet Mar 17 2007;369:946-55.
[2] Aaron LA, Buchwald D. A review of the evidence for overlap among unexplained
clinical conditions. Ann Intern Med May 1 2001;134:868-81.
[3] Warren JW, Howard FM, Cross RK, Good JL, Weissman MM, Wesselmann U,
Langengerg P, Greenberg P, Clauw DJ. Antecedent nonbladder syndromes in case–
control study of interstitial cystitis/painful bladder syndrome. Urology 2009;73:
52-7.
[4] Warren JW, Wesselmann U, Morozov V, Langenberg PW. Numbers and types of
nonbladder syndromes as risk factors for interstitial cystitis/painful bladder syn-
drome. Urology Feb 2011;77:313-9.
[5] Warren JW, Clauw DJ, Wesselmann U, Langenberg PW, Howard FM, Morozov V.
Sexuality and reproductive risk factors for interstitial cystitis/painful bladder syn-
drome in women. Urology Jan 6 2011;77:570-5.
[6] Warren JW, van de Merwe JP, Nickel JC. Interstitial cystitis/bladder pain syndrome
and nonbladder syndromes: facts and hypotheses. Urology Oct 2011;78:727-32.
[7] Warren JW, Diggs C, Horne L, Greenberg P. Interstitial cystitis/painful bladder
syndrome: what do patients mean by “perceived” bladder pain? Urology Dec 10
2011;77:309-12.
[8] Warren JW, Langenberg P, Greenberg P, Diggs C, Jacobs S, Wesselmann U. Sites of
pain from interstitial cystitis/painful bladder syndrome. J Urol Oct 2008;180:1373-7.
[9] Langenberg PW, Wallach EE, Clauw DJ, Howard FM, Diggs CM, Wesselmann U,
et al. Pelvic pain and surgeries in women before interstitial cystitis/painful blad-
der syndrome. Am J Obstet Gynecol Mar 2010;202:286.e1-6.
[10] Warren JW, Brown V, Jacobs S, Horne L, Langenberg P, Greenberg P. Urinary tract
infection and inflammation at onset of interstitial cystitis/painful bladder syn-
drome. Urology 2008;71:1085-90.
[11] Alagiri M, Chottiner S, Ratner V, Slade D, Hanno PM. Interstitial cystitis:
unexplained associations with other chronic disease and pain syndromes. Urolo-
gy May 1997;49:52-7.
[12] 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 fibromyal-
gia. Report of the Multicenter Criteria Committee. Arthritis Rheum Feb 1990;33:
160-72.
[13] Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. The chronic fatigue
syndrome: a comprehensive approach to its definition and study. International
Chronic Fatigue Syndrome Study Group. Ann Intern Med Dec 15 1994;121:953-9.
[14] Longstreth GF, Thompson WG, Chey WD, Houghton LA, Mearin F, Spiller RC. Func-
tional bowel disorders. Gastroenterology Apr 2006;130:1480-91.
[15] Vitali C, Bombardieri S, Jonsson R, Moutsopoulos HM, Alexander EL, Carsons SE,
et al. Classification criteria for Sjogren's syndrome: a revised version of the Euro-
pean criteria proposed by the American–European Consensus Group. Ann Rheum
Dis Jun 2002;61:554-8.
[16] Mathias SD, Kuppermann M, Liberman RF, Lipschutz RC, Steege JF. Chronic pelvic
pain: prevalence, health-related quality of life, and economic correlates. Obstet
Gynecol Mar 1996;87:321-7.
[17] Headache Classification Committee of the International Headache Society. Classi-
fication and diagnostic criteria headache disorder, cranial neuralgia, and facial
pain. Cephalalgia 1988;8:1–96.
[18] Diagnostic and statistical manual disorders. 4th revised ed.Washington DC: American
Psychiatric Association; 2000.
[19] Wessely S, Nimnuan C, Sharpe M. Functional somatic syndromes: one or many?
Lancet Sep 11 1999;354:936-9.
[20] Wessely S, White PD. There is only one functional somatic syndrome. Br J Psychi-
atry Aug 2004;185:95-6.
[21] Ciccone DS, Natelson BH. Comorbid illness in women with chronic fatigue syndrome:
a test of the single syndrome hypothesis. Psychosom Med Mar–Apr 2003;65:268-75.
[22] Moss-Morris R, Spence M. To “lump” or to “split” the functional somatic syndromes:
can infectious and emotional risk factors differentiate between the onset of chronic fa-
tigue syndrome and irritable bowel syndrome? Psychosom Med May–Jun 2006;68:
463-9.
[23] Kanaan RA, Lepine JP, Wessely SC. The association or otherwise of the functional
somatic syndromes. Psychosom Med Dec 2007;69:855-9.
[24] Kato K, Sullivan PF, Evengard B, Pedersen NL. A population-based twin study of
functional somatic syndromes. Psychol Med Mar 2009;39:497-505.
[25] White PD. Chronic fatigue syndrome: is it one discrete syndrome or many? Impli-
cations for the “one vs. many” functional somatic syndromes debate. J Psychosom
Res May 2010;68:455-9.
[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 Aug 1981;11:151-71.
[27] Whorwell PJ, McCallum M, Creed FH, Roberts CT. Non-colonic features of irritable
bowel syndrome. Gut Jan 1986;27:37-40.
[28] Hudson JI, Goldenberg DL, Pope HG, Keck PE, Schlesinger L. Comorbidity of fibro-
myalgia with medical and psychiatric disorders. Am J Med Apr 1992;92:363-7.
[29] Nimnuan C, Rabe-Hesketh S, Wessely S, Hotopf M. How many functional somatic
syndromes? J Psychosom Res Oct 2001;51:549-57.
[30] Aggarwal VR, McBeth J, Zakrzewska JM, Lunt M, Macfarlane GJ. The epidemiology
of chronic syndromes that are frequently unexplained: do they have common as-
sociated factors? Int J Epidemiol Apr 2006;35:468-76.
[31] Schur EA, Afari N, Furberg H, Olarte M, Goldberg J, Sullivan PF, et al. Feeling bad in
more ways than one: comorbidity patterns of medically unexplained and psychi-
atric conditions. J Gen Intern Med Jun 2007;22:818-21.
[32] Kato K, Sullivan PF, Pedersen NL. Latent class analysis of functional somatic symp-
toms in a population-based sample of twins. J Psychosom Res May 2010;68:447-53.
[33] Julien N, Goffaux P, Arsenault P, Marchand S. Widespread pain in fibromyalgia is
related to a deficit of endogenous pain inhibition. Pain Mar 2005;114:295-302.
[34] Geisser ME, Glass JM, Rajcevska LD, Clauw DJ, Williams DA, Kileny PR, et al. A psy-
chophysical study of auditory and pressure sensitivity in patients with fibromyal-
gia and healthy controls. J Pain May 2008;9:417-22.
[35] Gracely RH, Petzke F, Wolf JM, Clauw DJ. Functional magnetic resonance imaging ev-
idence of augmented pain processing in fibromyalgia. Arthritis Rheum May 2002;46:
1333-43.
[36] Geisser ME, Strader Donnell C, Petzke F, Gracely RH, Clauw DJ, Williams DA. Co-
morbid somatic symptoms and functional status in patients with fibromyalgia
and chronic fatigue syndrome: sensory amplification as a common mechanism.
Psychosomatics May–Jun 2008;49:235-42.
[37] Meeus M, Nijs J, Van de Wauwer N, Toeback L, Truijen S. Diffuse noxious inhibito-
ry control is delayed in chronic fatigue syndrome: an experimental study. Pain
2009;139:439-48.
[38] Chang L, Mayer EA, Johnson T, FitzGerald LZ, Naliboff B. Differences in somatic
perception in female patients with irritable bowel syndrome with and without fi-
bromyalgia. Pain Feb 2000;84:297-307.
[39] Clauw DJ. Perspectives on fatigue from the study of chronic fatigue syndrome and
related conditions. PM R May 2010;2:414-30.
16 J.W. Warren et al. / Journal of Psychosomatic Research 74 (2013) 12–17
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.

6. [40] Piche M, Arsenault M, Poitras P, Rainville P, Bouin M. Widespread hypersensitivity
is related to altered pain inhibition processes in irritable bowel syndrome. Pain
Jan 2010;148:49-58.
[41] Zhou Q, Fillingim RB, Riley JL, Malarkey WB, Verne GN. Central and peripheral hy-
persensitivity in the irritable bowel syndrome. Pain Mar 2010;148:454-61.
[42] Tak LM, Cleare AJ, Ormel J, Manoharan A, Kok IC, Wessely S, et al. Meta-analysis
and meta-regression of hypothalamic–pituitary–adrenal axis activity in function-
al somatic disorders. Biol Psychol Feb 16 2011;87:183-94.
[43] Hickie I, Lloyd A, Hadzi-Pavlovic D, Parker G, Bird K, Wakefield D. Can the chronic fa-
tigue syndrome be defined by distinct clinical features? Psychol Med Sep 1995;25:
925-35.
[44] Wilson A, Hickie I, Hadzi-Pavlovic D, Wakefield D, Parker G, Straus SE, et al. What
is chronic fatigue syndrome? Heterogeneity within an international multicentre
study. Aust N Z J Psychiatry Aug 2001;35:520-7.
[45] Sullivan PF, Smith W, Buchwald D. Latent class analysis of symptoms associated
with chronic fatigue syndrome and fibromyalgia. Psychol Med Jul 2002;32:881-8.
[46] Nisenbaum R, Reyes M, Unger ER, Reeves WC. Factor analysis of symptoms among
subjects with unexplained chronic fatigue: what can we learn about chronic fa-
tigue syndrome? J Psychosom Res Feb 2004;56:171-8.
[47] Sullivan PF, Pedersen NL, Jacks A, Evengard B. Chronic fatigue in a population
sample: definitions and heterogeneity. Psychol Med Sep 2005;35:1337-48.
[48] Vollmer-Conna U, Aslakson E, White PD. An empirical delineation of the heterogene-
ity of chronic unexplained fatigue in women. Pharmacogenomics Apr 2006;7:
355-64.
[49] Hickie I, Davenport T, Vernon SD, Nisenbaum R, Reeves WC, Hadzi-Pavlovic D, et al.
Are chronic fatigue and chronic fatigue syndrome valid clinical entities across coun-
tries and health-care settings? Aust N Z J Psychiatry Jan 2009;43:25-35.
[50] Aslakson E, Vollmer-Conna U, Reeves WC, White PD. Replication of an empirical
approach to delineate the heterogeneity of chronic unexplained fatigue. Popul
Health Metr Oct 5 2009;7:17.
[51] Kerr JR, Burke B, Petty R, Gough J, Fear D, Mattey DL, et al. Seven genomic subtypes
of chronic fatigue syndrome/myalgic encephalomyelitis: a detailed analysis of
gene networks and clinical phenotypes. J Clin Pathol Jun 2008;61:730-9.
[52] Smith AK, White PD, Aslakson E, Vollmer-Conna U, Rajeevan MS. Polymorphisms
in genes regulating the HPA axis associated with empirically delineated classes of
unexplained chronic fatigue. Pharmacogenomics Apr 2006;7:387-94.
[53] Carmel L, Efroni S, White PD, Aslakson E, Vollmer-Conna U, Rajeevan MS. Gene ex-
pression profile of empirically delineated classes of unexplained chronic fatigue.
Pharmacogenomics Apr 2006;7:375-86.
[54] Aslakson E, Vollmer-Conna U, White PD. The validity of an empirical delineation of
heterogeneity in chronic unexplained fatigue. Pharmacogenomics Apr 2006;7:
365-73.
[55] Turk DC, Okifuji A, Sinclair JD, Starz TW. Pain, disability, and physical functioning
in subgroups of patients with fibromyalgia. J Rheumatol Jul 1996;23:1255-62.
[56] Turk DC, Okifuji A, Sinclair JD, Starz TW. Differential responses by psychosocial
subgroups of fibromyalgia syndrome patients to an interdisciplinary treatment.
Arthritis Care Res Oct 1998;11:397-404.
[57] Giesecke T, Williams DA, Harris RE, Cupps TR, Tian X, Tian TX, et al. Subgrouping
of fibromyalgia patients on the basis of pressure-pain thresholds and psycholog-
ical factors. Arthritis Rheum Oct 2003;48:2916-22.
[58] Wilson HD, Starz TW, Robinson JP, Turk DC. Heterogeneity within the fibromy-
algia population: theoretical implications of variable tender point severity rat-
ings. J Rheumatol Dec 2009;36:2795-801.
[59] Rehm SE, Koroschetz J, Gockel U, Brosz M, Freynhagen R, Tolle TR, et al. A
cross-sectional survey of 3035 patients with fibromyalgia: subgroups of patients
with typical comorbidities and sensory symptom profiles. Rheumatology (Oxford)
Jun 2010;49:1146-52.
[60] Arebi N, Bullas DC, Dukes GE, Gurmany S, Hicks KJ, Kamm MA, et al. Distinct neu-
rophysiological profiles in irritable bowel syndrome. Am J Physiol Gastrointest
Liver Physiol Jun 2011;300:G1086-93.
[61] Nyholt DR, Gillespie NG, Heath AC, Merikangas KR, Duffy DL, Martin NG. Latent
class and genetic analysis does not support migraine with aura and migraine
without aura as separate entities. Genet Epidemiol Apr 2004;26:231-44.
[62] Ligthart L, Boomsma DI, Martin NG, Stubbe JH, Nyholt DR. Migraine with aura and
migraine without aura are not distinct entities: further evidence from a large
Dutch population study. Twin Res Hum Genet Feb 2006;9:54-63.
[63] Schurks M, Buring JE, Kurth T. Migraine features, associated symptoms and triggers:
a principal component analysis in the women's health study. Cephalalgia May
2011;31:861-9.
[64] Briggs AC, Stretch DD, Brandon S. Subtyping of panic disorder by symptom profile.
Br J Psychiatry Aug 1993;163:201-9.
[65] Roberson-Nay R, Kendler KS. Panic disorder and its subtypes: a comprehensive
analysis of panic symptom heterogeneity using epidemiological and treatment
seeking samples. Psychol Med Nov 2011;41:2411-21.
[66] Young EE, Lariviere WR, Belfer I. Genetic basis of pain variability: recent advances.
J Med Genet 2012;49:1-9.
[67] Hamilton WT, Gallagher AM, Thomas JM, White PD. Risk markers for both chronic
fatigue and irritable bowel syndromes: a prospective case–control study in pri-
mary care. Psychol Med Nov 2009;39:1913-21.
[68] Forseth KO, Gran JT, Husby G. A population study of the incidence of fibromyalgia
among women aged 26–55 yr. Br J Rheumatol Dec 1997;36:1318-23.
[69] Reyes M, Nisenbaum R, Hoaglin DC, Unger ER, Emmons C, Randall B, et al. Preva-
lence and incidence of chronic fatigue syndrome in Wichita, Kansas. Arch Intern
Med Jul 14 2003;163:1530-6.
[70] Talley NF, Weaver AL, Sinsmeister AR, Melton LF. Onset and disappearance of gas-
trointestinal symptoms and functional gastrointestinal disorders. Am J Epidemiol
1992;136:165-77.
[71] Nicholl BI, Halder SL, Macfarlane GJ, Thompson DG, O'Brien S, Musleh M, et al.
Psychosocial risk markers for new onset irritable bowel syndrome—results of a
large prospective population-based study. Pain Jul 2008;137:147-55.
[72] Stewart WF, Linet MS, Celentano DD, Van Natta M, Ziegler D. Age- and sex-specific
incidence rates of migraine with and without visual aura. Am J Epidemiol Nov 15
1991;134:1111-20.
[73] Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L. The prevalence and characteristics
of fibromyalgia in the general population. Arthritis Rheum Jan 1995;38:19-28.
[74] White KP, Speechley M, Harth M, Ostbye T. The London fibromyalgia epidemiology
study: the prevalence of fibromyalgia syndrome in London, Ontario. J Rheumatol
Jul 1999;26:1570-6.
[75] ter Borg EJ, Gerards-Rociu E, Haanen HC, Westers P. High frequency of hysterecto-
mies and appendectomies in fibromyalgia compared with rheumatoid arthritis: a
pilot study. Clin Rheumatol 1999;18:1-3.
[76] Al-Allaf AW, Dunbar KL, Hallum NS, Nosratzadeh B, Templeton KD, Pullar T. A
case–control study examining the role of physical trauma in the onset of fibromy-
algia syndrome. Rheumatology (Oxford) Apr 2002;41:450-3.
[77] Boneva RS, Maloney EM, Lin JM, Jones JF, Wieser F, Nater UM, et al. Gynecological his-
tory in chronic fatigue syndrome: a population-based case–control study. J Womens
Health (Larchmt) Jan 2011;20:21-8.
[78] Jette N, Patten S, Williams J, Becker W, Wiebe S. Comorbidity of migraine and psychi-
atric disorders—a national population-based study. Headache Apr 2008;48:501-16.
17J.W. Warren et al. / Journal of Psychosomatic Research 74 (2013) 12–17
Downloaded from ClinicalKey.com at PeaceHealth March 31, 2016.
For personal use only. No other uses without permission. Copyright ©2016. Elsevier Inc. All rights reserved.