Psychopharmacology (2013) 230:1521

DOI 10.1007/s00213-013-3085-x

ORIGINAL INVESTIGATION

The lack of association between components of metabolic

syndrome and treatment resistance in depression
Marina Sagud & Alma Mihaljevic-Peles & Suzana Uzun &
Bjanka Vuksan Cusa & Oliver Kozumplik &
Suzan Kudlek-Mikulic & Maja Mustapic & Ivan Barisic &
Dorotea Muck-Seler & Nela Pivac

Received: 25 July 2012 / Accepted: 25 March 2013 / Published online: 12 April 2013
# Springer-Verlag Berlin Heidelberg 2013

Abstract
Rationale Although a number of studies investigated the
link between major depressive disorder (MDD) and metabolic syndrome (MetS), the association between MetS and
treatment-resistant depression (TRD) is still not clear.
Objectives The aim of the study was to investigate the
relationship between TRD and MetS and/or components of
MetS and cardiovascular risk factors. Given the high prevalence of both conditions, the hypothesis was that TRD
would be significantly associated with MetS.
Methods This cross-sectional study included 203 inpatients
with MDD, assessed for the treatment resistance, MetS and
its components, and severity of MDD. Diagnoses and evaluations were made with SCID based on DSM-IV, National
Cholesterol Education Program Adult Treatment Panel III
criteria, and the Hamilton Depression Rating Scale.

Results TRD prior to study entry was found in 26.1 % of

patients, while MetS was observed in 33.5 % of patients.
The prevalence of MetS did not differ significantly between
TRD and non-TRD patients. In addition, the frequency of
the altered values of particular components of the MetS or
cardiovascular risk factors was not associated with treatment
resistance in depressed patients. Patients with TRD were
older, had a higher number of lifetime episodes of depression and suicide attempts, and longer duration of MDD
compared to non-TRD patients.
Conclusions The occurrence of either MetS or the particular
components of the MetS and other cardiovascular risk factors was similar between TRD and non-TRD patients.
Although there is a bidirectional relationship between depression and MetS, neither MetS nor its components appear
to influence treatment resistance to antidepressants.

M. Sagud : A. Mihaljevic-Peles
School of Medicine, University of Zagreb and Clinical Hospital
Center Zagreb, Zagreb, Croatia

Keywords Major depressive disorder (MDD) . Treatmentresistant depression (TRD) . Metabolic syndrome (MetS) .
Components of the metabolic syndrome . Cardiovascular risk
factors

S. Uzun : O. Kozumplik
Department of General Psychiatry, Clinic for Psychiatry Vrapce,
Zagreb, Croatia
B. V. Cusa : S. Kudlek-Mikulic
Department of Psychiatry, Clinical Hospital Center Zagreb,
Zagreb, Croatia
M. Mustapic : D. Muck-Seler : N. Pivac (*)
Division of Molecular Medicine, Rudjer Boskovic Institute,
PO Box 180, HR-10002, Zagreb, Croatia
e-mail: [email protected]
I. Barisic
Department of Nephrology, Dialysis Unit, Clinical Hospital Center
Zagreb, Zagreb, Croatia

Introduction
Major depressive disorder (MDD) is a complex and highly
heterogeneous disorder with significant morbidity and mortality. In spite of the growing number of different antidepressant drugs, treatment-resistant depression (TRD) remains an
important public health issue, with an estimated 12-month
prevalence of 2 % in general population, for a stage 2 of
treatment resistance (Nemeroff 2007). Treatment resistance
has been related to a greater number of psychiatric hospitalizations (Shah et al. 2002), and higher medical costs compared

16

to non-TRD patients (Gibson et al. 2010). Although TRD

appears to be a common clinical condition, remarkably little
is known about the underlying biology (Carney and Freedland
2009). Patients with TRD exhibit right superior, mediofrontal,
and striatal atrophy, as well as hippocampal and rostral anterior cingulate cortex changes, compared to non-TRD and
healthy controls (Shah et al. 2002). Recent articles suggest
differences in neural activity (Guo et al. 2012) and specific
brain microstructural white matter abnormalities
(Hoogenboom et al. 2012) in patients with TRD.
Interestingly, the study which included TRD patients, revealed
that those who were remitted at follow-up had an increase in
total brain volume, compared to non-remitters who showed
decreased white matter volume in the left anterior limb of the
internal capsule (Phillips et al. 2012).
Only a few studies investigated clinical features associated with TRD, with inconsistent findings. Anxiety comorbidity, comorbid panic disorder and social phobia,
personality disorder, suicidal risk, severity of symptoms,
melancholic features, more than one hospitalization, recurrent episodes, early age at onset, and nonresponse to the first
antidepressant received lifetime, were all associated with
TRD (Souery et al. 2007). However, there is no reliable
predictor of clinical response to antidepressants in general.
Among many others, genetic factors, like genetic variations
of serotonin transporter and P glycoprotein, could play a
role in the antidepressant response (Bozina et al. 2008;
Mihaljevic-Peles et al. 2008). The ongoing study performed
by the European Group for the Study of Resistant
Depression (Schosser et al. 2012) has shown that response
phenotype could be associated with genes variants of
catchol-O-methyltransferase, brain-derived neurotrophic
factor, and serotonergic 5-HT2A receptor, but not to cytochrome P450. Given the severe consequences of depression,
a better understanding of factors predicting response to
antidepressants is needed.
Previous research suggested a bidirectional link between
metabolic syndrome (MetS) and depression. In general,
MetS may be an important predisposing factor for the later
development of depression (Koponen et al. 2008), or cardiovascular disease (Pannier et al. 2006) in healthy subjects,
while severe depressive symptoms or extremely stressful
life event(s) may predict the risk for later development of
the MetS (Raikkonen et al. 2007). Furthermore, depressed
compared to nondepressed subjects have greater risk of
developing MetS, elevated waist circumference, and increased glucose levels (Toker et al. 2008), while depression
scores were significantly higher in healthy subjects with
than without MetS (Pannier et al. 2006; Skilton et al.
2007; Toker et al. 2008).
Although a number of studies (Koponen et al. 2008;
Pannier et al. 2006; Richter et al. 2010; Zeugmann et al.
2010) investigated the link between depression and MetS,

Psychopharmacology (2013) 230:1521

and both MetS and treatment resistance are common in

MDD patients, the association between MetS and TRD is
still not clear (Chokka et al., 2006). The aim of our study
was to determine whether MetS or its components, or other
cardiovascular risk factors, are related to treatment resistance in patients with MDD. The hypothesis of this study
was that the frequency of MetS (or its components) would
be higher in patients with TRD compared to non-TRD
patients.

Methods
This cross-sectional study was conducted at the University
Hospital Center Zagreb, Department of Psychiatry, and
University Clinics for Psychiatry Vrapce, Zagreb, Croatia,
and included 203 inpatients older than 18 years and hospitalized with the primary diagnosis of MDD. Patients were
hospitalized according to the discretion of the attending
physician, mostly due to severe depression which did not
respond to the current treatment. In addition, some patients
were hospitalized due to suicidal risk. Clinical diagnosis of
MDD on admission was further confirmed by the psychiatrists, using a SCID according to DSM-IV criteria (APA
1994). The severity of depression was evaluated using the
17-item Hamilton Depression Rating Scale (HDRS)
(Hamilton 1960). The retrospective medical record review
including previously diagnosed psychiatric and medical
comorbidities, and psychiatric and medical drug treatments,
both current and prescribed during the previous 12 months,
was also used. Demographic data, family, medical and psychiatric histories, and smoking habits, were recorded from
the clinical interview with the patient and from medical
records.
Treatment resistance prior to study entry was defined
according to Souery et al. (1999) as a failure to achieve as
a minimum 50 % reduction from baseline scores on HDRS
after no less than two courses of antidepressant monotherapy for at least 8 weeks (Little 2009), with full antidepressant dose (i.e., 2040 mg/day for fluoxetine or its
equivalent). Antidepressant drugs mostly prescribed were
selective serotonin reuptake inhibitors (SSRIs), noradrenaline reuptake inhibitors, followed by tianeptine, mirtazapine,
bupropion, reboxetine, and maprotiline. Excluded were
patients diagnosed with schizophrenia or schizoaffective
disorder, bipolar disorder, or organic mental disorder. We
also excluded patients who had MDD with psychotic features and/or have received any mood stabilizers and/or
antipsychotic drugs, in any dose, in the previous 12 months.
Patients who received less than two antidepressant drug
trials for the current depressive episode were also excluded.
Presence of MetS was diagnosed according to the
National Cholesterol Education Program Adult Treatment

16

to non-TRD patients (Gibson et al. 2010). Although TRD

appears to be a common clinical condition, remarkably little
is known about the underlying biology (Carney and Freedland
2009). Patients with TRD exhibit right superior, mediofrontal,
and striatal atrophy, as well as hippocampal and rostral anterior cingulate cortex changes, compared to non-TRD and
healthy controls (Shah et al. 2002). Recent articles suggest
differences in neural activity (Guo et al. 2012) and specific
brain microstructural white matter abnormalities
(Hoogenboom et al. 2012) in patients with TRD.
Interestingly, the study which included TRD patients, revealed
that those who were remitted at follow-up had an increase in
total brain volume, compared to non-remitters who showed
decreased white matter volume in the left anterior limb of the
internal capsule (Phillips et al. 2012).
Only a few studies investigated clinical features associated with TRD, with inconsistent findings. Anxiety comorbidity, comorbid panic disorder and social phobia,
personality disorder, suicidal risk, severity of symptoms,
melancholic features, more than one hospitalization, recurrent episodes, early age at onset, and nonresponse to the first
antidepressant received lifetime, were all associated with
TRD (Souery et al. 2007). However, there is no reliable
predictor of clinical response to antidepressants in general.
Among many others, genetic factors, like genetic variations
of serotonin transporter and P glycoprotein, could play a
role in the antidepressant response (Bozina et al. 2008;
Mihaljevic-Peles et al. 2008). The ongoing study performed
by the European Group for the Study of Resistant
Depression (Schosser et al. 2012) has shown that response
phenotype could be associated with genes variants of
catchol-O-methyltransferase, brain-derived neurotrophic
factor, and serotonergic 5-HT2A receptor, but not to cytochrome P450. Given the severe consequences of depression,
a better understanding of factors predicting response to
antidepressants is needed.
Previous research suggested a bidirectional link between
metabolic syndrome (MetS) and depression. In general,
MetS may be an important predisposing factor for the later
development of depression (Koponen et al. 2008), or cardiovascular disease (Pannier et al. 2006) in healthy subjects,
while severe depressive symptoms or extremely stressful
life event(s) may predict the risk for later development of
the MetS (Raikkonen et al. 2007). Furthermore, depressed
compared to nondepressed subjects have greater risk of
developing MetS, elevated waist circumference, and increased glucose levels (Toker et al. 2008), while depression
scores were significantly higher in healthy subjects with
than without MetS (Pannier et al. 2006; Skilton et al.
2007; Toker et al. 2008).
Although a number of studies (Koponen et al. 2008;
Pannier et al. 2006; Richter et al. 2010; Zeugmann et al.
2010) investigated the link between depression and MetS,

Psychopharmacology (2013) 230:1521

and both MetS and treatment resistance are common in

MDD patients, the association between MetS and TRD is
still not clear (Chokka et al., 2006). The aim of our study
was to determine whether MetS or its components, or other
cardiovascular risk factors, are related to treatment resistance in patients with MDD. The hypothesis of this study
was that the frequency of MetS (or its components) would
be higher in patients with TRD compared to non-TRD
patients.

Methods
This cross-sectional study was conducted at the University
Hospital Center Zagreb, Department of Psychiatry, and
University Clinics for Psychiatry Vrapce, Zagreb, Croatia,
and included 203 inpatients older than 18 years and hospitalized with the primary diagnosis of MDD. Patients were
hospitalized according to the discretion of the attending
physician, mostly due to severe depression which did not
respond to the current treatment. In addition, some patients
were hospitalized due to suicidal risk. Clinical diagnosis of
MDD on admission was further confirmed by the psychiatrists, using a SCID according to DSM-IV criteria (APA
1994). The severity of depression was evaluated using the
17-item Hamilton Depression Rating Scale (HDRS)
(Hamilton 1960). The retrospective medical record review
including previously diagnosed psychiatric and medical
comorbidities, and psychiatric and medical drug treatments,
both current and prescribed during the previous 12 months,
was also used. Demographic data, family, medical and psychiatric histories, and smoking habits, were recorded from
the clinical interview with the patient and from medical
records.
Treatment resistance prior to study entry was defined
according to Souery et al. (1999) as a failure to achieve as
a minimum 50 % reduction from baseline scores on HDRS
after no less than two courses of antidepressant monotherapy for at least 8 weeks (Little 2009), with full antidepressant dose (i.e., 2040 mg/day for fluoxetine or its
equivalent). Antidepressant drugs mostly prescribed were
selective serotonin reuptake inhibitors (SSRIs), noradrenaline reuptake inhibitors, followed by tianeptine, mirtazapine,
bupropion, reboxetine, and maprotiline. Excluded were
patients diagnosed with schizophrenia or schizoaffective
disorder, bipolar disorder, or organic mental disorder. We
also excluded patients who had MDD with psychotic features and/or have received any mood stabilizers and/or
antipsychotic drugs, in any dose, in the previous 12 months.
Patients who received less than two antidepressant drug
trials for the current depressive episode were also excluded.
Presence of MetS was diagnosed according to the
National Cholesterol Education Program Adult Treatment

Psychopharmacology (2013) 230:1521

Panel III criteria (National Cholesterol Education Program

(US). Expert Panel on Detection Evaluation and Treatment
of High Blood Cholesterol in Adults. 2002). The criteria for
MetS include at least three of the following: waist circumference 102 cm (men) or 88 cm (women); triglyceride
(TG) 1.7 mmol/L; HDL-C 1.00 mmol/L (men),
1.30 mmol/L (women); fasting glucose 6.1 mmol/L or
use of medication for hyperglycemia and blood pressure
135/85 mmHg or use of medication for hypertension. All
included patients gave a written informed consent after the
procedure(s) were fully explained. The study protocol was
approved by the local Ethics Committees.
Blood samples were taken as a part of routine hospital
admission procedure, in the morning after an overnight fasting
for 12 h. Serum cholesterol, high density lipoprotein cholesterol (HDL), and TG levels were determined by the enzymatic
color test for clinical analyzers, with the linearity within
concentration range of 0.6418, 0.054.65, and 0.11
11.40 mmol/l, respectively. Serum low density lipoprotein
cholesterol (LDL) levels were determined by the enzymatic
clearance assay, with linearity up to 22.4 mmol/L. Blood
glucose was determined by the standard laboratory test.
Waist circumference was measured using a non-stretch tape
measure at the narrowest point between the lower costal (10th
rib) border and the iliac crest. The body mass index (BMI) was
calculated by dividing the weight (in kilograms) by squared
height (in meters). Arterial blood pressure was measured in a
sitting position using sphygmomanometer auscultatory
method.
Statistical analysis
The results were expressed as means standard deviations
(SD), number of cases, and adjusted odds ratios, and 95 %
confidence intervals (CI). The differences between groups
were evaluated using one-way analysis of variance
(ANOVA). The multivariate logistic regression analysis
was performed with TRD as dependent variable and several
independent variables (high glucose, high TG, low HDL,

17

high blood pressure, large waist circumference, MetS, high

BMI, high cholesterol) adjusted for age and sex. A 2 test
was used to evaluate the frequency of depressed patients
with or without TRD with present or absent MetS, and
subdivided according to borderline value for the particular
MetS components (elevated glucose, high TG and low
HDL, increased blood pressure, and waist circumference)
and variables of cardiovascular risk (BMI, blood LDL and
cholesterol levels, atherogenic index 1 and 2). Standardized
residuals (R) were used to determine what categories were
the major influence on the significant chi-square test statistic
(http://www.acastat.com/Statbook/chisqresid.htm), and
were evaluated using Microsoft Excel. Due to multiple
testing (Table 2), Bonferroni correction was used and the
level of significance was set to p 0.0045. The power of
calculation was set to be 0.800. A power calculation for
large (e.g., Cohens d=0.8) effect size for the MetS comparison between TRD and non-TRD groups was calculated
using the G*Power 3 program (Faul et al., 2007). In addition, we have also calculated the effect size (c) according to
Cramers c statistics (Daniel 1990). All results were evaluated using the statistical package Sigma Stat 3.5 (Jandel
Scientific Corp. San Raphael, CA, USA).

Results
The demographic data in patients with TRD and non-TRD are
presented in Table 1. There were significant differences in age,
number of depressive episodes, duration of illness, and number of previous suicide attempts between patients with or
without TRD (Table 1). TRD patients were significantly (p=
0.001) older, had significantly (p=0.002) higher number of
previous episodes of depression, significantly (p=0.001) longer duration of MDD, and significantly higher number (p=
0.005) of suicide attempts than non-TRD patients (Table 1).
The percentage of subjects in non-TRD group that had 0, 1, 2,
or 3 suicide attempt(s) was 75.3, 16.7, 6.0, and 2.0 %, respectively; while in TRD patients it was 47.2, 41.5, 9.4, and 1.9 %,

Table 1 The demographic and

clinical variables in patients with
treatment-resistant depression
(TRD) and without TRD
(non-TRD)

Results are expressed as

mean SD. The number of
subjects is given in parentheses

Patients with MDD (203)

Non-TRD
(150)

Age (years)
Number of episodes
Duration of disease (years)
Total HDRS scores
Number of suicidal attempts
Current suicidality (HDRS item 3 scores)

49.39.7
2.91.9
6.44.9
26.15.9
0.350.69
1.141.12

TRD
(53)

57.210.3
3.91.7
9.36.2
27.76.3
0.660.73
1.511.28

One-way ANOVA
df=1,201
F

p value

25.572
10.209
11.132
2.608
7.916
3.960

0.001
0.002
0.001
0.108
0.005
0.048

18

Psychopharmacology (2013) 230:1521

respectively. This difference in the frequency of the number of

suicide attempts was slightly (but not significantly, after
Bonferroni correction) different (2 =7.692; p=0.0055). To
determine which category contributed to this difference, the
absolute value of R was calculated, and since R=2.78 (greater
than 2.00) was detected in TRD patients with one suicide
attempt, this frequency had a major influence on a marginally
significant 2 test.
Overall, 53 (26.1 %) out of 203 patients had TRD at the
study entry, while 33.5 % of patients developed MetS
(Table 2). MetS occurred in 31.3 % of non-TRD patients
and in 39.6 % of patients with TRD, and this difference was
not significant (2 = 0.865; p = 0.352). For the expected
Table 2 The number of patients
with treatment-resistant depression (TRD) and without TRD
(non-TRD) with the presence of
altered variables of metabolic
syndrome (glucose, triglycerides
levels, HDL levels, blood pressure, and waist circumference)
and variables of cardiovascular
risk (BMI, LDL and cholesterol
levels; atherogenic index 1
and 2)

N number of patients

effect size of d=0.8; p=0.0045; power=0.800; df=1, the

needed sample was calculated to be N=22 out of included
203 patients (Faul et al., 2007), so the study was well
powered. However, with the expected sample size of d=
0.8; p=0.0045; df=1, and total sample size N=203, to detect
significant differences 2 should be 36.976 with a power=
0.9999. According to Cramers c statistics (Daniel 1990) c
in the present study was very small (c=0.06). In addition,
there was no significant (p >0.0045) difference in the frequency of patients with or without the presence of particular
MetS components (elevated glucose, high TG and low HDL
values, increased blood pressure, and waist circumferences),
or the values of the other variables of cardiovascular risk

Patients with MDD (203)

Glucose
<6.1 mmol/l
6.1 mmol/l
HDL
>1.3 mmol/l women; >1.0 mmol/l men
1.3 mmol/l women; 1.0 mmol/l men
Triglycerides
<1.7 mmol/l
1.7 mmol/l
Blood pressure
<135/85 mm Hg
135/85 mmHg
Waist circumference
<88 cm women; <102 cm men
88 cm women; 102 cm men
Metabolic syndrome
No
Yes
BMI
<27.50 kg/m2
27.50 kg/m2
Cholesterol
<5.2 mmol/l
5.2 mmol/l
LDL
<3.0 mmol/l
3.0 mmol/l
Atherogenic index 1 (Cholesterol/LDL ratio)
<4.0 women; <5.2 men
4.0 women; 5.2 men
Atherogenic index 2 (LDL/HDL ratio)
<2.3 women; <3.0men
2.3 women; 3.0 men

Non-TRD
(150)

TRD
(53)

111
39

32
21

2 =2.867; df=1; p=0.090

115
35

40
13

2 =0.115; df=1; p=0.735

58
92

25
18

2 =0.846; df=1; p=0.358

88
62

27
26

2 =0.663; df=1; p=0.119

103
47

33
20

2 =0.465; df=1; p=0.495

103

32

2 =0.865; df=1; p=0.142

47

21

79
71

29
24

2 =0.009; df=1; p=0.923

92
53

24
29

2 =0.464; df=1; p=0.496

57
93

28
25

2 =2.956; df=1; p=0.086

122
28

49
4

2 =2.857; df=1; p=0.091

90
60

36
17

2 =0735.; df=1; p=0.391

Psychopharmacology (2013) 230:1521

(BMI, LDL, cholesterol levels, atherogenic index 1 and 2)

between patients with and without TRD (Table 2).
Multivariate logistic regression analysis, adjusted for sex
and age, with TRD set as a dependent variable, showed a
nonsignificant trend (after Bonferroni correction) for the
marginally higher (p=0.029) glucose levels that predicted
development of treatment resistance in patients with MDD.
These patients had 2.39 times higher odds ratio (95 % CI
1.0905.245) to develop TRD (Table 3).

Discussion
The results of the present study showed that MetS occurred
similarly in patients with or without TRD, and that distribution
of patients with the presence of particular components of the
MetS (elevated glucose, elevated TG, low HDL, elevated
blood pressure, and increased waist circumference), or alteration in other variables of cardiovascular risk (BMI, LDL,
cholesterol, atherogenic index 1 and 2) did not differ significantly between patients with and without TRD. Marginally
increased glucose levels were found more frequently in
patients with TRD compared to non-TRD patients, but the
significance was lost after Bonferroni correction.
In line with the other studies (Carney and Freedland
2009; Souery et al. 2007), treatment resistance was present
in approximately one third of our patients with MDD.
Patients with TRD were older, had a longer duration of
disease, higher number of both depressive episodes and
suicide attempts, as well as increased current suicidality,
compared to non-TRD patients. MetS was observed in
33.5 % of patients which is in agreement with the 24
36 % prevalence of MetS in patients with depression
(Koponen et al. 2008; Pannier et al. 2006). To the best of
our knowledge, this is the first study investigating the association between TRD, MetS, and its components. Although
both TRD and MetS appear with high frequency in patients
with MDD, our results did not show that MetS occur more
frequently in patients with TRD.
Table 3 Multivariate logistic
regression analysis in patients
with MDD with TRD as dependent variable adjusted for age
and sex

19

There are several factors that should be discussed. In general, our sample was characterized by patients with relatively
late age of onset of MDD (both TRD and non-TRD patients
experienced their first episode after the age of 40 years). Given
that patients with TRD were older, vascular etiology could
also have contributed to MDD onset and the development of
treatment resistance. Preclinical vascular disease (Muldoon et
al. 2007) was associated with blunted central serotonergic
function, measured by prolactin response to citalopram in
healthy volunteers (Muldoon et al. 2006). Altered serotonergic function, such as decreased postsynaptic 5-HT-2A receptors in the dorsal regions of the prefrontal and the anterior
cingulate cortex, were observed in antidepressant-free TRD
patients (Baeken et al. 2012). However, the results of the
present study show no differences between patients with and
without TRD in the frequency of components of MetS, BMI,
and both atherogenic index 1 and 2. These findings strongly
argue against the differences in vascular pathology between
patients with and without TRD.
Few studies investigated the association between lipid levels and antidepressant response. In one study (Papakostas et
al. 2003), a higher TG level was found in patients with TRD
compared to non-TRD patients. Elevated cholesterol levels
were associated with poor treatment response to noradrenergic
antidepressants in TRD patients (Papakostas et al. 2003). A
poor treatment response and high cholesterol levels were also
observed in patients with MDD treated with SSRI fluoxetine
(Iosifescu et al. 2005) or paroxetine (Muck-Seler et al. 2011).
Moreover, nonresponders to paroxetine had increased baseline cholesterol, LDL, TG levels, and LDL/HDL ratio compared to responders to paroxetine (Muck-Seler et al. 2011).
The discrepancies with our findings could be due to different
methodological issues. In contrast to the inclusion criteria in
the present study, our earlier study (Muck-Seler et al. 2011)
included patients who were drug-free, nonsuicidal, had no
diabetes, and did not receive cholesterol-lowering drugs.
Furthermore, patients in the previous studies (Iosifescu et al.
2005; Muck-Seler et al. 2011) were younger compared to
patients included in the present study, suggesting that

Patients with MDD

High glucose levels (mmol/l)

High TG levels (mmol/l)
Low HDL (mmol/l)
High blood pressure (mmHg)
Large waist circumferences (cm)
Metabolic syndrome
High BMI (kg/m2)
High cholesterol (mmol/l)

Odds ratio

95 % CI

p value

2.391
0.768
1.002
1.231
0.983
1.282
0.935
0.539

1.0905.245
0.3741.579
0.4442.262
0.5662.675
0.4662.072
0.6062.710
0.4192.089
2.6011.114

0.029
0.473
0.997
0.6
0.963
0.516
0.871
0.095

20

increased cholesterol levels might influence poor response at

least to SSRIs, predominantly in younger patients. The influence of age in response to nonpharmacological treatment of
depression was also reported. Patients at age <60 years had
greater response to repetitive transcranial magnetic stimulation compared to older patients (Pallanti et al. 2012). On the
other hand, our results are in line with data from Amital et al.
(2012), who found no difference between TRD and non-TRD
patients in the prevalence of comorbid somatic disorders,
including diabetes, hypertension, and ischemic heart disease.
Although the prevalence of MetS increases with age, it is
possible that factors other than MetS, influence poorer response to antidepressants in older age.
Suicidal behavior is a complex clinical syndrome, related
to numerous biological, sociological, and psychological
factors (Olin et al. 2012). We have found that TRD was
associated with current suicidality and that TRD patients
more frequently had at least one prior suicide attempt, which
suggests more frequent suicidal behavior in TRD study
population. So far, only one study (Papakostas et al. 2003)
investigated the suicidality in TRD and found that more than
half of patients with TRD reported thoughts or wishes of
death. The higher number of previous suicide attempts in
our TRD compared to non-TRD patients is in line with the
results of Amital et al. (2008). In spite of the increased
suicidality, our patients with TRD and non-TRD had similar
severity of depression. This is in contrast with more severe
symptoms of depression observed in patients with TRD
(Amital et al. 2008) that were treated not only with antidepressant, like in the present study, but with combination of
antidepressants and antipsychotics.
Limitations of the study are cross-sectional design and
the fact that patients were treated with different antidepressants. Antidepressant drug concentration was not determined, and compliance with the antidepressant regimen
was not checked. Information about childhood exposure to
trauma was not available. Since the sample was crosssectional and included depressed inpatients, this sample is
not generalizable to the whole community sample. This
study had adequate power (0.800) to detect significant
differences in age, number of depressive episodes, duration
of illness, and number of previous suicide attempts between
patients with or without TRD. At the study entry, we had an
adequate number of patients, N=203 out of needed N=22,
expecting a high effect size. However, the number of
patients included in the study was less than adequate for
the small effect size (c=0.06) detected in our study, since a
similar frequency of MetS was detected between patients
with or without TRD. In addition, less than adequate power
was found in the frequency of patients with or without
presence of components of MetS or other variables of cardiovascular risk between patients with and without TRD, as
stated in the results, suggesting that these results should be

Psychopharmacology (2013) 230:1521

considered preliminary, and might have been significant if

sufficient sample size was available. Therefore, further research is needed on a larger number of patients with TRD.

Conclusions
The results of the study did not confirm the hypothesis that
MetS, its components, or variables of cardiovascular risk
factors, would be found more frequently in patients with
TRD compared to non-TRD patients. Although methodological and sample size limitations do not allow definitive
conclusions, and given the long-term health consequences
of both MetS and TRD, the relationship between components of MetS and TRD deserves further investigation in
adequately powered studies.
Acknowledgments This study was supported by the Croatian Ministry of Science, Education and Sport, grants nos. 098-0982522-2455;
098-0982522-2457; 109-1083509-3513; and 108-1083509-3511.
Conflict of interest Authors declare no conflict of interest.

References
Amital D, Fostick L, Silberman A, Beckman M, Spivak B (2008)
Serious life events among resistant and non-resistant MDD
patients. J Affect Disord 110:260264
Amital D, Fostick L, Silberman A, Calati R, Spindelegger C, Serretti
A, Juven-Wetzler A, Souery D, Mendlewicz J, Montgomery S,
Kasper S, Zohar J (2012) Physical co-morbidity among treatment
resistant vs. treatment responsive patients with major depressive
d i s o r d e r. E u r N e u r o p s y c h o p h a r m a c o l . d o i : 1 0 . 1 0 1 6 /
j.euroneuro.2012.09.002
APA, American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders: DSM-IV, 4th edn. American
Psychiatric Association, Washington, DC
Baeken C, De Raedt R, Bossuyt A (2012) Is treatment-resistance in
unipolar melancholic depression characterized by decreased
serotonin-2A receptors in the dorsal prefrontal-anterior cingulate
cortex? Neuropharmacol 62:340346
Bozina N, Peles AM, Sagud M, Bilusic H, Jakovljevic M (2008)
Association study of paroxetine therapeutic response with SERT
gene polymorphisms in patients with major depressive disorder.
World J Biol Psychiat 9:190197
Carney RM, Freedland KE (2009) Treatment-resistant depression and
mortality after acute coronary syndrome. Am J Psychiat 166:410
417
Chokka P, Tancer M, Vikram K, Yeragani VK (2006) Metabolic
syndrome: relevance to antidepressant treatment. J Psychiatry
Neurosci 31:414
Faul F, Erdfelder E, Lang A-G, Buchner A (2007) G*Power 3: a
flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 39:175191
Daniel WW (1990) Applied nonparametric statistics, 2nd edn. PWS
Kent Publishing Company, Boston, MA
Gibson TB, Jing Y, Smith Carls G, Kim E, Bagalman JE, Burton WN,
Tran QV, Pikalov A, Goetzel RZ (2010) Cost burden of treatment

Psychopharmacology (2013) 230:1521

resistance in patients with depression. Am J Manag Care 16:370
377
Guo WB, Liu F, Chen JD, Gao K, Xue ZM, Xu XJ, Wu RR, Tan CL,
Sun XL, Liu ZN, Chen HF, Zhao JP (2012) Abnormal neural
activity of brain regions in treatment-resistant and treatmentsensitive major depressive disorder: a resting-state fMRI study. J
Psychiatr Res 46:13661373
Hamilton M (1960) A rating scale for depression. J Neurol Neurosurg
Psychiatry 23:5662
Hoogenboom WS, Perlis RH, Smoller JW, Zeng-Treitler Q, Gainer VS,
Murphy SN, Churchill SE, Kohane IS, Shenton ME, Iosifescu DV
(2012) Limbic system white matter microstructure and long-term
treatment outcome in major depressive disorder: a diffusion tensor
imaging study using legacy data. World J Biol Psychiat.
doi:10.3109/15622975.2012.669499
Iosifescu DV, Clementi-Craven N, Fraguas R, Papakostas GI, Petersen
T, Alpert JE, Nierenberg AA, Fava M (2005) Cardiovascular risk
factors may moderate pharmacological treatment effects in major
depressive disorder. Psychosom Med 67:703706
Koponen H, Jokelainen J, Keinanen-Kiukaanniemi S, Kumpusalo E,
Vanhala M (2008) Metabolic syndrome predisposes to depressive
symptoms: a population-based 7-year follow-up study. J Clin
Psychiat 69:178182
Little A (2009) Treatment-resistant depression. Am Fam Physician
80:167172
Mihaljevic Peles A, Bozina N, Sagud M, Rojnic Kuzman M, Lovric M
(2008) MDR1 gene polymorphism: therapeutic response to paroxetine among patients with major depression. Prog
Neuropsychopharmacol Biol Psychiatry 32:14391444
Muldoon MF, Mackey RH, Korytkowski MT, Flory JD, Pollock BG,
Manuck SB (2006) The metabolic syndrome is associated with
reduced central serotonergic responsivity in healthy community
volunteers. J Clin Endocrinol Metab 91:718721
Muldoon MF, Mackey RH, Sutton-Tyrrell K, Flory JD, Pollock BG,
Manuck SB (2007) Lower central serotonergic responsivity is
associated with preclinical carotid artery atherosclerosis. Stroke
38:22282233
Muck-Seler D, Sagud M, Mihaljevi-Peles A, Jakovljevi M, Pivac N
(2011) Baseline lipid levels and acute treatment response to
paroxetine and tianeptine in depressed women. J Clin
Psychopharmacol 31:387390
National Cholesterol Education Program (US). Expert panel on detection evaluation and treatment of high blood cholesterol in adults.
(2002) 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): final report. The Program, Washington, D.C
Nemeroff CB (2007) Prevalence and management of treatmentresistant depression. J Clin Psychiat 68(Suppl 8):1725
Olin B, Jayewardene AK, Bunker M, Moreno F (2012) Mortality and
suicide risk in treatment-resistant depression: an observational

21
study of the long-term impact of intervention. PLoS One 7:
e48002. doi:10.1371/journal.pone.0048002
Pallanti S, Cantisani A, Grassi G, Antonini S, Cecchelli C, Burian J,
Cauli G, Quercioli L (2012) rTMS age-dependent response in
treatment-resistant depressed subjects: a mini-review. CNS
Spectr 17:2430
Pannier B, Thomas F, Eschwege E, Bean K, Benetos A, Leocmach Y,
Danchin N, Guize L (2006) Cardiovascular risk markers associated with the metabolic syndrome in a large French population:
the "SYMFONIE" study. Diabetes Metab 32:467474
Papakostas GI, Petersen T, Sonawalla SB, Merens W, Iosifescu DV,
Alpert JE, Fava M, Nierenberg AA (2003) Serum cholesterol in
treatment-resistant depression. Neuropsychobiology 47:146151
Phillips JL, Batten LA, Aldosary F, Tremblay P, Blier P (2012) Brainvolume increase with sustained remission in patients with
treatment-resistant unipolar depression. J Clin Psychiat 73:625
631
Raikkonen K, Matthews KA, Kuller LH (2007) Depressive symptoms
and stressful life events predict metabolic syndrome among
middle-aged women: a comparison of world health organization,
adult treatment panel III, and international diabetes foundation
definitions. Diabetes Care 30:27612761
Richter N, Juckel G, Assion HJ (2010) Metabolic syndrome: a followup study of acute depressive inpatients. Eur Arch Psychiatry Clin
Neurosci 260:4149
Schosser A, Serretti A, Souery D, Mendlewicz J, Zohar J, Montgomery
S, Kasper S (2012) European Group for the Study of Resistant
Depression (GSRD)Where have we gone so far: review of
clinical and genetic findings. Eur Neuropsychopharmacol.
doi:10.1016/j.euroneuro.2012.02.006
Shah PJ, Glabus MF, Goodwin GM, Ebmeier KP (2002) Chronic,
treatment-resistant depression and right fronto-striatal atrophy.
Brit J Psychiat 180:434440
Skilton MR, Moulin P, Terra JL, Bonnet F (2007) Associations between anxiety, depression, and the metabolic syndrome. Biol
Psychiat 62:12511257
Souery D, Amsterdam J, de Montigny C, Lecrubier Y, Montgomery S,
Lipp O, Racagni G, Zohar J, Mendlewicz J (1999) Treatment
resistant depression: methodological overview and operational
criteria. Eur Neuropsychopharm 9:8391
Souery D, Oswald P, Massat I, Bailer U, Bollen J, Demyttenaere K,
Kasper S, Lecrubier Y, Montgomery S, Serretti A, Zohar J,
Mendlewicz J (2007) Clinical factors associated with treatment
resistance in major depressive disorder: results from a European
multicenter study. J Clin Psychiat 68:10621070
Toker S, Shirom A, Melamed S (2008) Depression and the metabolic
syndrome: gender-dependent associations. Depression and
Anxiety 25:661669
Zeugmann S, Quante A, Heuser I, Schwarzer R, Anghelescu I (2010)
Inflammatory biomarkers in 70 depressed inpatients with and
without the metabolic syndrome. J Clin Psychiat 71:10071016