Research

Original Investigation

Association of Metabolic Syndrome

and Hidradenitis Suppurativa
Iben Marie Miller, MD; Christina Ellervik, MD, PhD; Gabrielle Randskov Vinding, MD; Kian Zarchi, MD;
Kristina Sophie Ibler, MD; Kim Mark Knudsen, PhD; Gregor B. E. Jemec, DMSc, MD

Editorial page 1263

IMPORTANCE An association between the metabolic syndrome (MetS) and chronic Supplemental content at
inflammatory diseases, such as psoriasis or rheumatoid arthritis, has been suggested. jamadermatology.com
Hidradenitis suppurativa (HS), a more localized chronic inflammation of the skin, has been
speculated to have a similar association. Hidradenitis suppurativa is a substantial burden for
the individual and a socioeconomic burden globally. Information about the burden of possible
comorbidities is scarce.

OBJECTIVE To investigate the possibility of an association between HS and MetS.

DESIGN, SETTING, AND PARTICIPANTS Cross-sectional population- and hospital-based study of

HS and MetS. We identified 32 patients with physician-verified HS from the outpatient clinic
at the Department of Dermatology, Roskilde Hospital, and 326 patients with HS and 14 851
individuals without HS from the general population. Individuals with HS were younger,
predominantly female, and more often smokers compared with the non-HS group.

EXPOSURE Hidradenitis suppurativa.

MAIN OUTCOMES AND MEASURES Metabolic syndrome and its components of diabetes
mellitus, hypertension, dyslipidemia, and obesity.

RESULTS When compared with the non-HS group, the odds ratios (ORs) for the hospital HS
and population HS groups were 3.89 (95% CI, 1.90-7.98) and 2.08 (95% CI, 1.61-2.69),
respectively, for MetS; 5.74 (95% CI, 1.91-17.24) and 2.44 (95% CI, 1.55-3.83), respectively, for
diabetes mellitus; 6.38 (95% CI, 2.99-13.62) and 2.56 (95% CI, 2.00-3.28), respectively, for
general obesity; and 3.62 (95% CI, 1.73-7.60) and 2.24 (95% CI, 1.78-2.82), respectively, for
abdominal obesity. With regard to dyslipidemia, significant results were found for decreased
levels of high-density lipoprotein cholesterol, with ORs of 2.97 (95% CI, 1.45-6.08) and 1.94
(95% CI, 1.52-2.48) for the hospital HS and general population HS groups, respectively, when
compared with the non-HS group. With regard to increased triglyceride levels, only the result
for the population HS group compared with the non-HS group was significant, with an OR of
1.49 (95% CI, 1.18-1.87). The OR for hypertension, which was only significant for the hospital
HS group compared with the non-HS group, was 2.14 (95% CI, 1.01-4.53). Obesity and
inflammation acted as possible confounders. The ORs were higher for the hospital HS group
compared with the population HS group. The association between HS and MetS was not
influenced by the degree of HS severity.
Author Affiliations: Department of
Dermatology, Roskilde Hospital,
CONCLUSIONS AND RELEVANCE As with more systemic inflammatory diseases, HS appears to Roskilde, Denmark (Miller, Vinding,
be associated with MetS, indicating substantial comorbidities. Because this study is Zarchi, Ibler, Jemec); Department of
Health and Medical Sciences,
cross-sectional, causality remains to be explored.
University of Copenhagen,
Copenhagen, Denmark (Miller,
Jemec); Department of Clinical
Biochemistry, Nstved Hospital,
Nstved, Denmark (Ellervik); LEO
Pharma A/S, Ballerup, Denmark
(Knudsen).
Corresponding Author: Iben Marie
Miller, MD, Department of
Dermatology, Roskilde Hospital,
JAMA Dermatol. 2014;150(12):1273-1280. doi:10.1001/jamadermatol.2014.1165 Kgevej 7-13, 4000 Roskilde,
Published online September 17, 2014. Denmark (miller@dadlnet.dk).

1273

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017

 Research Original Investigation
H
idradenitis suppurativa (HS) is a chronic, localized in-
flammatory skin disease producing inflamed nod-
ules in apocrine glandbearing skin.1,2 The treatment
of HS is often inadequate, and the disease inflicts a signifi-
cant burden on patients, in whom pain and suppuration from
lesions often lead to inactivity, depression, and significantly
impaired quality of life.3,4 Changes in patients socioprofes-
sional activity indicate the presence of important disease-
related impairment.5,6
Hidradenitis suppurativa may be more common than hith-
erto suggested. The disease is often misdiagnosed or under-
diagnosed, and prevalence estimates therefore range from
0.05% to 4%.7-10 An increasing incidence during the last 20
years has been suggested.11 The comorbidities of HS are poorly
described, and the available data are solely hospital based.12,13
The pandemic cluster of cardiovascular risk factors called
the metabolic syndrome (MetS) co-occurs more commonly
with chronic inflammatory diseases, such as rheumatoid ar-
thritis and psoriasis.14-18 In contrast to psoriasis, HS may be con-
sidered a more localized inflammation of the skin. An asso-
ciation between HS and MetS has been hypothesized to exist
and cause significant comorbidity, negatively influencing the
overall burden of disease. We therefore investigated a pos-
sible association using population- and hospital-based data in
a cross-sectional study to explore the association and assess
its possible clinical relevance.
Methods
Ethical Statement
This study was accepted by the ethics committee of the Zea-
land region (project numbers SJ-191, SJ-113, and SJ-114) in Den-
mark. Written informed consent was obtained from all study
participants.
Study Design
We performed a cross-sectional study of the association of HS
(referred to as the exposure) and MetS (referred to as the out-
come). We investigated 2 different groups of individuals with
HS. The first group was identified in a general population
sample (population HS group); the second was identified in a
hospital-based sample (hospital HS group).
Exposure
The population HS group was identified in the Danish Gen-
eral Suburban Population Study (GESUS). GESUS was initi-
ated in January 2010 with ongoing enrollment and is a cross-
sectional study of the adult Danish suburban general
population in the Nstved municipality (70 km south of
Copenhagen).19 All citizens 30 years and older and a random
selection of those aged 20 to 30 years were invited to partici-
pate. The population HS group was defined based on whether
participants reported boils within the previous 6 months and
a minimum of 2 boils (in the following 5 possible locations: ax-
illae, groin, genitals, mammae, or other [eg, perianal, neck, or
abdomen]) on a questionnaire. The validation of this HS di-
agnosis is discussed in a separate report5 showing a sensitiv-
1274 JAMA Dermatology December 2014 Volume 150, Number 12
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017
Hidradenitis Suppurativa
ity of 90% and a specificity of 97%. The overall participation
rate in GESUS was 49.3%. Further details about GESUS can be
found in Bergholdt et al.19
The hospital HS group was recruited from the outpatient
clinic at the Department of Dermatology at Roskilde Hospital
(serving the region of Zealand, which includes Nstved). In-
clusion criteria consisted of the diagnostic code for HS from
the International Classification of Diseases, 10th Revision (L73.2)
and systemic or laser treatment for HS, which indicated mod-
erate or severe disease. The diagnosis of HS was confirmed by
results of a physical examination by a physician from the De-
partment of Dermatology (I.M.M., G.R.V., K.Z., or K.S.I.). Eli-
gible patients were invited to undergo the same examination
as the population sample. The participation rate was 33 of 98
eligible patients (33.7%). The distribution of age and sex did
not differ between participants and nonparticipants (data not
shown).
Non-HS participants were defined as participants from
GESUS without HS. This population constituted the non-HS
group for this study.
Outcomes
Both HS groups and the population-based non-HS group
completed the same questionnaire and physical examina-
tions and contributed blood samples. The definitions of the
MetS outcome were based on the methods of GESUS involv-
ing the GESUS questionnaire (self-reporting), physical exami-
nation, and laboratory evaluation of nonfasting venous blood
samples.19
The MetS involves the following 4 key components: dia-
betes mellitus, hypertension, dyslipidemia, and obesity. The
methods of this study used to define the outcome are listed
in Box 1 and Box 2.20-24 We used a modified version of the cri-
teria of the National Cholesterol Education Program Adult
Treatment Panel III,20 which also accounts for the harmo-
nized definition of MetS.
Exploring Possible Confounders
We considered the following possible confounders:
1. General obesity defined by body mass index (BMI; calcu-
lated as weight in kilograms divided by height in meters
squared) measured at the physical examination;
2. Inflammatory load defined by high-sensitivity C-reactive-
protein level measured in venous blood samples;
3. Self-reported level of physical activity at work and in lei-
sure time;
4. Self-reported intake of atherogenic (ie, saturated) fat, fish,
fruit/vegetables, eggs, and alcohol.
Exploring the Severity of HS
The definition of severity of HS for the population HS group
was based on self-reported information on numbers and
locations of boils and subsequent scarring and was inspired
by the Hurley score, which is considered almost static.25
Mild HS was defined as a minimum of 2 boils and no subse-
quent scarring; moderate HS, a minimum of 2 boils and sub-
sequent scarring; and severe HS, a minimum of 2 boils in a
minimum of 2 locations and subsequent scarring. The
jamadermatology.com
 Hidradenitis Suppurativa
Box 1. Definition of Metabolic Syndrome
Metabolic syndrome according to a modified version of the US NCEP
ATP III includes 3 of the following criteria:
Abdominal Obesity
WC of >102 cm for men and >88 cm for women based on the
physical examination findings
Dyslipidemia
1. Decreased plasma HDL levels of <40 mg/dL for men and
<50 mg/dL for women based on analysis of the blood sample
2. Increased plasma TG levels of 150 mg/dL, based on analysis of
the blood sample
Hypertension
Blood pressure 130/85 mm Hg at the physical examination or self-
reported use of antihypertensive
Diabetes Mellitus
HbA1c level of 48 mmol/mol (IFCC) (6.5% [DCCT]), nonfasting
plasma glucose level of 220 mg/dL, or self-reported diagnosis of
diabetes mellitus
Abbreviations: DCCT, Diabetes Control and Complications Trials;
HbA1c, hemoglobin A1c; HDL, high-density lipoprotein cholesterol;
IFCC, International Federation of Clinical Chemistry; NCEP ATP III, National
Cholesterol Education Program Adult Treatment Panel III; TG, triglycerides;
WC, waist circumference.
SI conversion factors: To convert cholesterol to millimoles per liter, multiply
by 0.0259; glucose to millimoles per liter, multiply by 0.0555; and TG to
millimoles per liter, multiply by 0.0113.
severity of HS for the hospital HS group was assessed
by the Sartorius score based on results of the physical
examination.25 We also explored the severity of HS in the
population and hospital HS groups as the number of boils to
constitute a more dynamic scale.
Statistical Analysis
We compared the HS and non-HS groups using logistic regres-
sion adjusting for age, sex, and smoking status, yielding an ad-
justed odds ratio (OR) for binary outcomes, and linear regres-
sion of log-transformed outcomes adjusting for age, sex, and
smoking status, yielding an adjusted ratio of means (RM) for
continuous outcomes. The binary effects measure (OR) ex-
presses whether an association exists or not and the strength
of that association. The continuous effect measure (RM) ex-
presses the quantification of the association and can more eas-
ily be translated into clinical management than an OR be-
cause RM expresses the difference between the mean values
of the HS vs non-HS groups. Thus, the RM provides the ex-
pected higher value of an individual with HS compared with
a non-HS individual (eg, RM of 1.21 for BMI means that indi-
viduals with HS are expected to have 21% higher BMI than those
without HS). In comparison, the OR is based on cutoff values
and provides the expected higher odds (eg, OR of 2.00 for a
BMI > 30 means that individuals with HS have 2 times higher
odds of having a BMI > 30). P < .05 was considered to be sta-
tistically significant.
jamadermatology.com
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017
Original Investigation Research
Box 2. Metabolic Syndrome Component Outcomes
Diabetes Mellitus
Binary Outcome
1. Self-reported diagnosis or based on analysis of the blood sample:
HbA1c level of 48 mmol/mol (IFCC) (6.5% DCCT) or
nonfasting plasma glucose level of 220 mg/dL
2. Self-reported use of insulin or noninsulin antidiabetic
Continuous Outcome
1. Quantification of HbA1c level (mmol/mol) and nonfasting plasma
glucose level based on analysis of the blood sample
Hypertension
Binary Outcome
Self-reported use of antihypertensives or blood pressure
140/90 mm Hg according to definition of hypertension
Continuous Outcome
Quantification based on measurements of systolic and diastolic blood
pressure at physical examination
Dyslipidemia
Binary Outcome
1. Increased TG levels according to the US NCEP ATP III criteria of
150 mg/dL
2. Decreased level of HDL according to the US NCEP ATP III criteria
of <40 mg/dL for men and <50 mg/dL for women based on
analysis of the blood sample
Continuous Outcome
1. Quantification by measurement of TG, HDL, and TC levels, based
on analysis of blood sample
Obesity
Binary Outcome
1. General obesity defined as BMI of 30, based on physical
examination findings
2. Abdominal obesity defined as WC according to NCEP ATP III
criteria of >102 cm for men and >88 cm for women, based on
physical examination findings
Continuous Outcome
1. Quantification of obesity based on BMI and WC, based on
physical examination findings
Abbreviations: BMI, body mass index (calculated as weight in kilograms
divided by height in meters squared); DCCT, Diabetes Control and
Complications Trial; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein
cholesterol; IFCC, International Federation of Clinical Chemistry; NCEP ATP
III, National Cholesterol Education Program Adult Treatment Panel III;
TC, total cholesterol; TG, triglycerides; WC, waist circumference.
SI conversion factors: To convert cholesterol to millimoles per liter, multiply
by 0.0259; glucose to millimoles per liter, multiply by 0.0555; TG to
millimoles per liter, multiply by 0.0113.
The ORs and RMs for the population HS group are ex-
pressed as ORpop and RMpop. Similarly, ORs and RMs for the
hospital HS group are expressed as ORhos and RMhos.
We examined the influence of possible confounders on the
association between HS and MetS by including these in the re-
gression model and assessing the effect on the OR. We ex-
plored the relationship between the severity of HS and MetS
JAMA Dermatology December 2014 Volume 150, Number 12 1275
 Research Original Investigation Hidradenitis Suppurativa

Table. Background Factors, Characteristics, and Distribution of Outcome in Study Groupsa

HS Groups
Hospital Population Non-HS Group
Characteristic (n = 32) (n = 326) (n = 14 851)
Age, mean (range), y 42 (22-64) 47 (22-78) 56 (20-96)
Sex, female vs male, No. (%) 25 (78) vs 7 (22) 218 (67) vs 108 (33) 8090 (54.5) vs 6761 (45.5)
Smoking status, No. (%)
Present 17 (53.1) 133 (40.8) 2683 (18.1)
Past 13 (40.6) 114 (35.0) 5633 (37.9)
Never 1 (3.1) 70 (21.5) 5684 (38.3)
Abbreviations: BMI, body mass index
Ethnicity, white, No. (%) 31 (96.9) 315 (96.6) 14 624 (98.5)
(calculated as weight in kilograms
CRP level, median (range), mg/L 5.1 (0.2-119.0) 2.1 (0.1-38.0) 1.4 (0.1-136.0) divided by height in meters squared);
HS severity distribution, % CRP, C-reactive protein;
HDL, high-density lipoprotein
Mild 4 (12.5) 165 (50.6) NA
cholesterol; HS, hidradenitis
Moderate 5 (15.6) 90 (27.6) NA suppurativa; MetS, metabolic
Severe 23 (71.9) 71 (21.8) NA syndrome; NA, not applicable;
TG, triglycerides; WC, waist
Sartorius score, median (range) 29 (5-176) NA NA
circumference.
No. of boils, median (range) 12 (1-171) 3 (2-106) NA
SI conversion factor: To convert CRP
MetS, No. (%)b 17 (53.1) 105 (32.2) 3192 (21.5) to nanomoles per liter, multiply by
Diabetes mellitus, No. (%)c 4 (12.5) 23 (7.1) 728 (4.9) 9.524.
a
Hypertension, No. (%)d 18 (56.3) 157 (48.2) 9007 (60.6) Missing data for groups are reported
in eTable 5 in the Supplement.
Blood sample, No. (%) b
Indicates a modified Adult
Decreased HDL level 15 (46.9) 108 (33.1) 2687 (18.1) Treatment Panel III definition.20
Increased TG level 16 (50.0) 159 (48.8) 6428 (43.3) c
Indicates self-reported or based on
Obesity, No. (%) results of laboratory analysis of
venous blood sample.
General (BMI) 16 (50.0) 107 (32.8) 2799 (18.8)
d
Indicates self-reported or measured
Abdominal (WC) 20 (62.5) 168 (51.5) 5524 (37.2)
during the physical examination.

using the same regression method as above analyzing only the
HS groups. All statistical analyses were performed using com-
mercially available software (SAS, version 9.3; SAS, Inc).
Owing to differences in background factors between ex-
posed and unexposed groups and knowledge from previous
studies that age, sex, and smoking status act as confounders
in cardiovascular risk, the effect measures were adjusted ac-
cordingly. In eTables 1 and 2 in the Supplement, crude ORs and
RMs are displayed.
Results
The data used from GESUS were collected from January 1, 2010,
through August 2, 2012. A total of 32 individuals with HS from
the hospital, 326 with HS from the general population, and
14 851 non-HS individuals from the general population were
identified. The background factors and characteristics of the
HS and non-HS groups showed that individuals with HS were
predominately younger, female, and smokers (Table).
We investigated the association of HS with the condi-
tions involved in the MetSdiabetes mellitus, hypertension,
dyslipidemia, and obesityand with MetS defined by the cri-
teria of the National Cholesterol Education Program Adult
Treatment Panel III.20 The methods used to define the out-
come are described in Box 1 and Box 2. The ORs and RMs ad-
justed for age, sex, and smoking status are illustrated in Figure 1
and Figure 2.
Association of MetS and MetS Components With HS
When adjusting for age, sex, and smoking status, the associa-
tion between HS and MetS was significant for the hospital and
population HS groups (ORhos, 3.89 [95% CI, 1.90-7.98]; ORpop,
2.08 [95% CI, 1.61-2.69]). Findings for the associations of MetS
components with HS are discussed individually. We found a
uniform pattern of the hospital HS group having higher ORs
than the population HS group.
Diabetes Mellitus
When adjusting for age, sex, and smoking status, a signifi-
cant association between HS and diabetes mellitus was found
for the hospital and population HS groups compared with the
non-HS group (ORhos, 5.74 [95% CI, 1.91-17.24]; ORpop, 2.44 [95%
CI, 1.55-3.83]). When we explored the type of antidiabetic used
(insulin or noninsulin), only the association of noninsulin drugs
was significant (ORhos, 7.93 [95% CI, 1.75-36.03]; ORpop, 3.50
[95% CI, 2.05-5.98]). When quantifying the association, only
the RMhos was significant, with levels 8% (95% CI, 2%-15%) and
10% (95% CI, 5%-15%) higher for the HS groups compared with
the non-HS group with regard to glucose and hemoglobin A1c
levels, respectively.
Hypertension
When adjusting for age, sex, and smoking status, a signifi-
cant association between HS and hypertension was found only
for the hospital HS group (ORhos, 2.14 [95% CI, 1.01-4.53]). When
quantifying this association, only the RMhos for diastolic blood

1276 JAMA Dermatology December 2014 Volume 150, Number 12 jamadermatology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017

 Hidradenitis Suppurativa Original Investigation Research

pressure was significant, with a level 5% (95% CI, 1%-10%)

Figure 1. Hidradenitis Suppurativa (HS) and the Metabolic Syndrome
higher in the HS hospital group when compared with the
non-HS group. Hospital HS group vs non-HS group
Population HS group vs non-HS group
Dyslipidemia
OR Negative Positive
When adjusting for age, sex, and smoking status, an associa-
Outcome (95% CI) Association Association
tion between HS and high triglyceride (TG) levels was found,
Metabolic syndrome 3.89 (1.90-7.98)
but the association was statistically significant for the popu- 2.08 (1.61-2.69)
lation HS group only (ORpop, 1.49 [95% CI, 1.18-1.87]). A sig- Diabetes mellitus 5.74 (1.91-17.24)
nificant association between HS and low levels of high- 2.44 (1.55-3.83)
density lipoprotein cholesterol (HDL) was found (ORhos, 2.97 Hypertension 2.14 (1.01-4.53)
[95% CI, 1.45-6.08]; ORpop, 1.94 [95% CI, 1.52-2.48]). When 1.08 (0.85-1.38)
quantifying the association, the RMpop for TG level was 21% Increased TG level 1.67 (0.81-3.44)
(95% CI, 0%-46%) and the RMhos was 11% (95% CI, 4%-17%), 1.49 (1.18-1.87)

significantly higher than in the non-HS group. The RMhos for
HDL level was 14% (95% CI, 5%-22%) and the RMpop was 10%
(95% CI, 7%-13%), significantly lower than in the non-HS
group.
Obesity
When adjusting for age, sex, and smoking status, we found a
significant association between general obesity by BMI and ab-
dominal obesity by waist circumference (WC) for the popula-
tion and hospital HS groups. With regard to general obesity,
the ORhos was 6.38 (95% CI, 2.99-13.62) and the ORpop was 2.56
(95% CI, 2.00-3.28). With regards to abdominal obesity, the
ORhos was 3.62 (95% CI, 1.73-7.60) and the ORpop was 2.24 (95%
CI, 1.78-2.82).
When we quantified the association, the RMhos for BMI was
21% (95% CI, 14%-29%) and the RMpop was 9% (95% CI, 7%-
11%), significantly higher for the HS groups than the non-HS
group. The RMhos for WC was 14% (95% CI, 9%-19%) and the
RMpop was 7% (95% CI, 6%-9%), significantly higher for the HS
groups than the non-HS group.
Role of Obesity, Inflammatory Load, Physical Activity,
Diet, and HS Severity
When exploring the possible confounders, we found that ad-
justing for obesity or inflammatory load reduced the strengths
of the associations; however, the associations remained. Thus,
obesity and inflammation were identified confounders,
whereas level of physical activity and diet were not (eTable 3
in the Supplement). The association between HS and MetS or
between HS and the MetS components was not influenced by
the degree of HS severity with the exception of general obe-
sity (eTable 4 in the Supplement).
Discussion
This broad population- and hospital-based study suggests
an association between HS and MetS and the individual
MetS components of diabetes mellitus, low levels of HDL,
and general and abdominal obesity. Positive associations
were also found with regard to hypertension and dyslipid-
emia. Hypertension, however, was only statistically signifi-
cant in the hospital HS group. When we used binary data
(with acknowledged cutoff values) for increased levels of
Decreased HDL level 2.97 (1.45-6.08)
1.94 (1.52-2.48)
General obesity 6.38 (2.99-13.62)
2.58 (2.00-3.23)
Abdominal obesity 3.62 (1.73-7.60)
2.24 (1.78-2.82)
0.50 1.00 2.00 5.00 10.00 20.00
Odds Ratio
Odds ratios (ORs) for the binary outcomes of the metabolic syndrome and its
component outcomes (diabetes mellitus, hypertension, increased triglyceride
[TG] levels and decreased high-density lipoprotein cholesterol [HDL] levels, and
general and abdominal obesity). The ORs are adjusted for age, sex, and smoking
status. Whiskers indicate 95% confidence intervals.
TG, only the association for the population HS group was
statistically significant. However, statistically significant dif-
ferences in the TG levels between the hospital HS group and
the non-HS group were found when we examined the con-
tinuous data.
The hospital HS group had uniformly higher ORs than
the population HS group. This result could indicate that dif-
ferences in HS severity or the recent suggestion of different
HS subtypes may play a part.26 This difference might also be
an expression of a dilution of the population-based HS
sample due to misclassification bias. The stronger ORs for the
hospital HS group may also indicate detection bias; that is,
HS patients within the hospital system are more likely to
have been diagnosed with the outcome, which influences the
self-reported diagnoses. However, detection bias was mini-
mized by the inclusion of self-reported diagnosis and results
of the physical examination and laboratory analysis of blood
samples.
When we examined the MetS components individually, the
ORs for diabetes mellitus indicated a positive association,
mainly due to type 2 diabetes mellitus. Quantification analy-
sis demonstrated that nonfasting glucose level was 1% to 8%
higher, and hemoglobin A1c level was 1% to 10% higher in the
HS groups compared with the non-HS group. Previous stud-
ies have suggested that an approximately 1% reduction in he-
moglobin A1c level may decrease the risk for myocardial in-
farction by 14% to 16%, indicating that the abnormalities seen
in the HS patients have clinical significance.27 Further indi-
rect support for this association is provided by the observa-
tion that metformin hydrochloride treatment may amelio-
rate HS.28

jamadermatology.com JAMA Dermatology December 2014 Volume 150, Number 12 1277

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017

 Research Original Investigation Hidradenitis Suppurativa

Figure 2. Hidradenitis Suppurativa (HS) and the Components of the Metabolic Syndrome

Hospital HS vs non-HS groups

Population HS vs non-HS groups

RM Negative Positive
Component (95% CI) Association Association
Nonfasting blood glucose level 1.08 (1.02-1.15)
1.01 (0.99-1.03)
HbA1c level (IFCC) 1.10 (1.05-1.15)
1.01 (1.00-1.03)
Diastolic blood pressure 1.05 (1.01-1.10)
1.01 (1.00-1.02)
Systolic blood pressure 1.02 (0.97-1.07) Ratio of means (RMs) for continuous
0.99 (0.98-1.01) outcomes of the components of the
TG level 1.21 (1.00-1.46) metabolic syndrome (nonfasting
1.11 (1.04-1.17) blood glucose level, hemoglobin A1c
[HbA1c] level, diastolic and systolic
HDL level 0.88 (0.78-0.95) blood pressure, triglyceride [TG]
0.90 (0.87-0.93) blood level, high-density lipoprotein
LDL level 0.94 (0.84-1.06) cholesterol [HDL] level, low-density
0.99 (0.96-1.03) lipoprotein cholesterol [LDL] level,
total cholesterol [TC] level, general
TC level 0.96 (0.90-1.06)
obesity, and abdominal obesity).
0.98 (0.96-1.00)
The RMs are adjusted for age, sex,
General obesity, BMI 1.21 (1.14-1.29) and smoking status. Whiskers
1.09 (1.07-1.11) indicate 95% confidence intervals.
Abdominal obesity, WC 1.14 (1.09-1.19) BMI indicates body mass index
1.07 (1.06-1.09) (calculated as weight in kilograms
divided by height in meters squared);
0.80 1.00 1.20 1.40 1.60 IFCC, International Federation of
Ratio of Means Clinical Chemistry; WC, waist
circumference.

Furthermore, the ORs for general and abdominal obesity
were significant. When quantified, BMI was 9% to 21% greater,
and WC was 7% to 14% larger in the HS groups than the non-HS
group. Trimming of the WC by 4.4 cm from 102 cm (ie, a 4.3%
reduction) may reduce the risk for diabetes mellitus by 58%,
similarly indicating clinical relevance and a substantial poten-
tial for prevention in this generally neglected disease.29
The association of an atherogenic lipid profile (ie, in-
creased TG levels and decreased HDL levels) was significant
with regard to the decrease in HDL levels in both HS groups,
but only significant with regard to the increase in TG levels in
the population HS group. When quantified, the HS groups had
an 11% to 21% higher TG level and a 10% to 14% lower HDL level
than the non-HS group. A follow-up study of the effects of stat-
ins in 17 802 healthy individuals showed a significant reduc-
tion of cardiovascular events when TG levels were reduced by
17% and HDL levels were increased by 4%.30
Having hypertension was only significant with regard to
the hospital HS group. When quantifying the hospital HS group,
only the diastolic blood pressure was significantly increased
by 5%. A recent meta-analysis of preventive treatment of hy-
pertension suggested that lowering diastolic blood pressure
from 90 to 85 mm Hg (ie, a 6% reduction) would reduce the
risk for cardiovascular heart disease and stroke by approxi-
mately 20% and 30%, respectively.31
Our findings are in concordance with and expand the find-
ings of 2 previous hospital-based studies.12,13 In aggregate, these
data therefore suggest that the comorbidities of HS are clini-
cally significant and that an increased clinical awareness of the
HS diagnosis and its comorbidities might be warranted in this
potentially substantial group of patients.
Obesity and inflammatory load were identified as possible
confounders, partly but not exclusively explaining the associa-
tions and indicating a complex and overlapping relationship.
Surprisingly, we found that physical activity level, diet and al-
cohol consumption, and the severity of HS did not influence the
associations. The latter is in contrast to our supposition that the
hospital HS group had higher ORs because of more severe dis-
ease, implying detection bias as previously discussed or an in-
sufficiently sensitive measure of disease severity. Therefore, one
can speculate that HS subtypes may influence the association
with MetS more strongly than HS severity.
The association of MetS has also been shown in the chronic
generalized inflammatory disease rheumatoid arthritis and the
skin disease psoriasis, for which the ORs for MetS are approxi-
mately 2.00 compared with almost 6.00 in HS.8 Furthermore,
the associations in psoriasis have been suggested to be signifi-
cant only with regard to a hospital-based HS cohort,8 imply-
ing that the burden of these comorbidities is greater for HS than
for psoriasis. Therefore, the common etiological factors be-
tween inflammatory skin disease and MetS are more readily
identified in HS than in psoriasis.
The major strengths of our study are the large population-
based HS group and the inclusion of individuals with HS from
hospital- and population-based groups. The broad recruit-
ment reduced selection bias with a broader range of disease se-
verities and thereby aided the generalization of the results. To
explore misclassification bias of individuals with HS, we vali-

1278 JAMA Dermatology December 2014 Volume 150, Number 12 jamadermatology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017

 Hidradenitis Suppurativa Original Investigation Research

dated the HS definition with a sensitivity of 90% and a speci-
ficity of 97%.9 Because the self-reported questions used to iden-
tify HS patients refer to symptoms (ie, boils) rather than the
actual diagnosis (ie, do you have HS?), we strove to optimize the
inclusion of individuals with undiagnosed and misdiagnosed
HS in the population, which is particularly pertinent for under-
diagnosed diseases. The diagnosis of HS among hospitalized par-
ticipants was physician verified. The combined methods of self-
reporting, laboratory analysis of blood samples, and physical
examination aimed to reduce false-negative findings with re-
gard to the outcome (MetS). Finally, essential possible con-
founders were recognized and explored systematically.
Potential limitations merit consideration. First, one must
recognize that, because this study is cross-sectional, we can-
not prove causality between HS and MetS. Furthermore, the
population is suburban, most of the participants were white,
and the group aged 20 to 30 years was underrepresented, which
may limit the generalizability of our findings. In addition, an
age bias was found between the HS and non-HS groups. How-
ever, we accommodated this age bias by age adjustments. The
low participation rate of the hospital HS group increased varia-
tion and reduced power. The low participation rate may be an
expression of the limited resources of HS patients due to the
physical and mental burden of the disease. Nonfasting blood
samples were used. However, differences in fasting and non-
fasting lipid levels have been shown to be minimal.32 Further-
more, we did not include information on HS medical treat-
ment, which might confound the results. Last, as with any
questionnaire survey, the risk for recall bias was present.
Conclusions
The data suggest an association between HS and the MetS in
the hospital and population HS groups. This allegedly in-
creased disease burden due to comorbidities indicates that HS
patients require general medical attention beyond the skin.
Future longitudinal studies with similar methods are needed
to explore the temporal relationship of these associations. These
studies should be large, include individuals with HS from the
general population and hospital, and explore the differences be-
tween these groups and additional possible confounders.

ARTICLE INFORMATION
Accepted for Publication: May 1, 2014.
Published Online: September 17, 2014.
doi:10.1001/jamadermatol.2014.1165.
Author Contributions: Drs Miller and Jemec had
full access to all the data in the study and take
responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Miller, Ellervik, Vinding,
Ibler, Jemec.
Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Miller, Ibler, Jemec.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: Miller, Knudsen.
Obtained funding: Miller, Ellervik, Jemec.
Administrative, technical, or material support: Miller,
Vinding, Ibler, Jemec.
Study supervision: Ellervik, Ibler, Jemec.
Conflict of Interest Disclosures: Dr Jemec has
received consulting fees from Abbott Laboratories,
AstraZeneca, MSD, Novartis, Pfizer, and Dumex-
Alpharma; lecture fees from Abbott Laboratories,
Galderma, Pfizer, and Roche; grant support from
Abbott Laboratories, Pfizer, Photocure, and LEO
Pharma; equipment on loan from Michelson
Diagnostics; and reimbursement for travel
expenses from Abbott Laboratories, Galderma, and
Photocure. No other disclosures were reported.
Funding/Support: This study was supported by a
grant under the Industrial PhD Programme and
grants from the Danish Agency for Science,
Technology, and Innovation and LEO Pharma (Dr
Miller) and by the Jacob Madsen & Olga Madsens
Foundation. The GESUS study was supported by
the Johan and Lise Boserup Foundation,
Trygfonden, Det Kommunale Momsfond, Johannes
Fogs Foundation, Region Zealand Foundation,
Nstved Hospital, Nstved Hospital Foundation,
and the National Board of Health.
Role of the Funder/Sponsor: Dr Knudsen, who is
an employee of LEO Pharma, which provided part
of the grant supporting this study, was involved in
analysis and interpretation of data and manuscript
revision. The funding organizations had no other
role in the design and conduct of the study;
collection, management, analysis, and
interpretation of the data; preparation, review, or
approval of the manuscript; and decision to submit
the manuscript for publication.
Additional Contributions: Claus Bay, MSc Stat,
Jrgen Iversen, BSc, and Torsten Skov, MD, PhD,
Department of Biostatistics, LEO Pharma, provided
statistical, technical, and epidemiological support.
Birgitte Schnack Nielsen, clinical nutritionist,
Department of Pediatrics, Roskilde Hospital,
Denmark, provided information on diet. The
contributors did not receive compensation.
REFERENCES
1. Kurzen H, Kurokawa I, Jemec GB, et al. What
causes hidradenitis suppurativa? Exp Dermatol.
2008;17(5):455-472.
2. Jemec GBE. Clinical practice: hidradenitis
suppurativa. N Engl J Med. 2012;366(2):158-164.
3. Matusiak L, Bieniek A, Szepietowski JC.
Psychophysical aspects of hidradenitis suppurativa.
Acta Derm Venereol. 2010;90(3):264-268.
4. Esmann S, Jemec GBE. Psychosocial impact of
hidradenitis suppurativa: a qualitative study. Acta
Derm Venereol. 2011;91(3):328-332.
5. Jemec GBE, Heidenheim M, Nielsen NH.
Hidradenitis suppurativa: characteristics and
consequences. Clin Exp Dermatol. 1996;21(6):419-423.
6. von der Werth JM, Williams HC. The natural
history of hidradenitis suppurativa. J Eur Acad
Dermatol Venereol. 2000;14(5):389-392.
7. Cosmatos I, Matcho A, Weinstein R,
Montgomery MO, Stang P. Analysis of patient
claims data to determine the prevalence of
hidradenitis suppurativa in the United States. J Am
Acad Dermatol. 2013;68(3):412-419.
8. Revuz JE, Canoui-Poitrine F, Wolkenstein P, et al.
Prevalence and factors associated with hidradenitis
suppurativa: results from two case-control studies.
J Am Acad Dermatol. 2008;59(4):596-601.
9. Vinding GR, Miller IM, Zarchi K, et al. The
prevalence of inverse recurrent suppuration:
a population-based study of possible hidradenitis
suppurativa. Br J Dermatol. 2014;170(4):884-889.
10. Jemec GBE, Heidenheim M, Nielsen NH. The
prevalence of hidradenitis suppurativa and its
potential precursor lesions. J Am Acad Dermatol.
1996;35(2, pt 1):191-194.
11. Vazquez BG, Alikhan A, Weaver AL, Wetter DA,
Davis MD. Incidence of hidradenitis suppurativa and
associated factors: a population-based study of
Olmsted County, Minnesota. J Invest Dermatol. 2013;
133(1):97-103.
12. Sabat R, Chanwangpong A, Schneider-Burrus S,
et al. Increased prevalence of metabolic syndrome
in patients with acne inversa. PLoS One. 2012;7
(2):e31810. doi:10.1371/journal.pone.0031810.
13. Gold DA, Reeder VJ, Mahan MG, et al. The
prevalence of metabolic syndrome in patients with
hidradenitis suppurativa. J Am Acad Dermatol.
2014;70(4):699-703.
14. Miller IM, Ellervik C, Yazdanyar S, Jemec GB.
Meta-analysis of psoriasis, cardiovascular disease,
and associated risk factors. J Am Acad Dermatol.
2013;69(6):1014-1024.
15. Miller IM, Skaaby T, Ellervik C, Jemec GB.
Quantifying cardiovascular disease risk factors in
patients with psoriasis: a meta-analysis. Br J Dermatol.
2013;169(6):1180-1187.
16. Miller IM, Jemec GBE. Maturation of an idea:
a historical perspective on the association of psoriasis
with the metabolic syndrome and cardiovascular
disease [letter]. Arch Dermatol. 2012;148(1):112.
17. Ferraz-Amaro I, Gonzlez-Juanatey C,
Lpez-Mejias R, et al. Metabolic syndrome in
rheumatoid arthritis. Mediators Inflamm. 2013;
2013:710928. doi:10.1155/2013/710928.
18. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC
Jr, Lenfant C; National Heart, Lung, and Blood

jamadermatology.com JAMA Dermatology December 2014 Volume 150, Number 12 1279

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017

 Research Original Investigation Hidradenitis Suppurativa

Institute; American Heart Association. Definition of
metabolic syndrome: report of the National Heart,
Lung, and Blood Institute/American Heart
Association conference on scientific issues related
to definition. Arterioscler Thromb Vasc Biol.
2004;24(2):e13-e18.
19. Bergholdt HK, Bathum L, Kvetny J, et al. Study
design, participation and characteristics of the
Danish General Suburban Population Study. Dan
Med J. 2013;60(9):A4693.
20. Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults.
Executive summary of the Third Report of the
National Cholesterol Education Program (NCEP)
Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults (Adult
Treatment Panel III). JAMA. 2001;285(19):2486-2497.
21. Alberti KG, Eckel RH, Grundy SM, et al;
International Diabetes Federation Task Force on
Epidemiology and Prevention; National Heart,
Lung, and Blood Institute; American Heart
Association; World Heart Federation; International
Atherosclerosis Society; International Association
for the Study of Obesity. Harmonizing the
metabolic syndrome: a joint interim statement of
the International Diabetes Federation Task Force on
Epidemiology and Prevention; National Heart,
Lung, and Blood Institute; American Heart
Association; World Heart Federation; International
Atherosclerosis Society; and International
Association for the Study of Obesity. Circulation.
2009;120(16):1640-1645.
22. International Expert Committee. International
Expert Committee report on the role of the A1c
assay in the diagnosis of diabetes. Diabetes Care.
2009;32(7):1327-1334.
23. Institute for Rational Pharmacotherapy in
Denmark. Guidelines on diabetes. Mnedsbladet
rationel farmakoterapi. 2006;8. www.irf.dk. Accessed
May 21, 2014.
24. Chobanian AV, Bakris GL, Black HR, et al;
National Heart, Lung, and Blood Institute Joint
National Committee on Prevention, Detection,
Evaluation, and Treatment of High Blood Pressure;
National High Blood Pressure Education Program
Coordinating Committee. The seventh report of the
Joint National Committee on Prevention, Detection,
Evaluation, and Treatment of High Blood Pressure:
the JNC 7 report. JAMA. 2003;289(19):2560-2572.
25. Sartorius K, Lapins J, Emtestam L, Jemec GB.
Suggestions for uniform outcome variables when
reporting treatment effects in hidradenitis
suppurativa. Br J Dermatol. 2003;149(1):211-213.
26. Canoui-Poitrine F, Le Thuaut A, Revuz JE, et al.
Identification of three hidradenitis suppurativa
phenotypes: latent class analysis of a cross-sectional
study. J Invest Dermatol. 2013;133(6):1506-1511.
27. Macisaac RJ, Jerums G. Intensive glucose
control and cardiovascular outcomes in type 2
diabetes. Heart Lung Circ. 2011;20(10):647-654.
28. Verdolini R, Clayton N, Smith A, Alwash N,
Mannello B. Metformin for the treatment of
hidradenitis suppurativa: a little help along the way.
J Eur Acad Dermatol Venereol. 2013;27(9):1101-1108.
29. Tuomilehto J, Lindstrm J, Eriksson JG, et al;
Finnish Diabetes Prevention Study Group.
Prevention of type 2 diabetes mellitus by changes
in lifestyle among subjects with impaired glucose
tolerance. N Engl J Med. 2001;344(18):1343-1350.
30. Ridker PM, Danielson E, Fonseca FA, et al;
JUPITER Study Group. Rosuvastatin to prevent
vascular events in men and women with elevated
C-reactive protein. N Engl J Med. 2008;359(21):
2195-2207.
31. Law MR, Morris JK, Wald NJ. Use of blood
pressure lowering drugs in the prevention of
cardiovascular disease: meta-analysis of 147
randomised trials in the context of expectations
from prospective epidemiological studies. BMJ.
2009;338:b1665. doi:10.1136/bmj.b1665.
32. Langsted A, Freiberg JJ, Nordestgaard BG.
Fasting and nonfasting lipid levels: influence of
normal food intake on lipids, lipoproteins,
apolipoproteins, and cardiovascular risk prediction.
Circulation. 2008;118(20):2047-2056.

NOTABLE NOTES

Apple of the Dermatologists Eye

Deshan F. Sebaratnam, MBBS(Hons)

Dermatology is an intrinsically visual specialty, with pattern recognition
a vital skill for clinical practice. It therefore follows that many of the clas-
sic descriptions of cutaneous pathologic abnormalities draw on paral-
lels from the natural world, particularly the realm of botany. Trees, leaves,
fruits, and flowers have all lent themselves to morphological descrip-
tions over the centuries.
The distribution of the papulosquamous plaques of pityriasis
rosea has been compared to a fir tree and the hair shaft changes of
trichorrhexis nodosa to bamboo. The figurate polycyclic plaques of
erythema gyratum repens have been described as resembling wood
grain and the jagged forked lesions of livedo racemosa tree branches.1
The induration associated with scleroderma is classically described as
woody, and the neurofibromas of neurofibromatosis are said to have
a rubbery consistency. Lichenification derives its name from lichen (a
composite organism of fungi and algae living symbiotically), and the
tumors of mycosis fungoides were initially thought to be fungal
growths (though strictly speaking, the kingdom of Fungi is separate
from that of plants).
The ovate shape of the hypopigmented macules of tuberous
sclerosis has been compared to ash leaves, and the color of Zoon
balanitis has been described as resembling that of fallen autumn
leaves.
Fruits have inspired a range of dermatological descriptions, includ-
ing peau dorange, tapioca vesicles in pompholyx, and cayenne pepper
spots in pigmented purpuric dermatoses. The berries are particularly well
represented through strawberry birthmarks, mulberry xanthomas, and
cherry angiomas.
Flowers have also given rise to dermatological descriptions, such as
the heliotrope rash of dermatomyositis, which originates from the lilac
color of flowers of the genus Heliotropium.2 Perhaps the plant that has
inspired the greatest number of wordsmiths within dermatology and be-
yond is the rose. Varicella zoster virus gives rise to dew drops on a rose
petal, human herpes virus is associated with roseola infantum, and Sal-
monella typhi produces rose spots. The rosette is a well-recognized sign
reported in a range of descriptions: clinically, as in linear IgA disease; der-
moscopically, as in squamous cell carcinoma3; and pathologically, as in
cylindromas.
The visual nature of dermatology lends itself to comparisons be-
tween cutaneous pathological abnormalities and those seen in the bo-
tanical world, with several classical descriptions firmly implanted into
the dermatological canon.
Author Affiliation: Skin and Cancer Foundation Australia, Department of
Dermatology, Westmead, NSW, Australia.
Corresponding Author: Deshan F. Sebaratnam, Skin and Cancer Foundation
Australia, 7 Ashley Lane, Westmead, NSW 2154, Australia
(d.sebaratnam@hotmail.com).
1. Parsi K, Partsch H, Rabe E, Ramelet AA. Reticulate eruptions, part 2: historical
perspectives, morphology, terminology and classification. Australas J Dermatol.
2011;52(4):237-244.
2. Russo T, Piccolo V, Ruocco E, Baroni A. The heliotrope sign of
dermatomyositis: the correct meaning of the term heliotrope. Arch Dermatol.
2012;148(10):1178.
3. Rubegni P, Tataranno DR, Nami N, Fimiani M. Rosettes: optical effects and
not dermoscopic patterns related to skin neoplasms. Australas J Dermatol.
2013;54(4):271-272.

1280 JAMA Dermatology December 2014 Volume 150, Number 12 jamadermatology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/931783/ on 06/19/2017