Journal of Affective Disorders 168 (2014) 373379

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research report

Serum Brain-derived neurotrophic factor levels in

post-stroke depression
Jie Li a, Yan-Dong Zhao b, Jun-Wei Zeng c, Xiao-Yan Chen a, Ruo-Dan Wang a, Sai-Yu Cheng a,n
a

Department of Neurology, Second Afliated Hospital and Xin Qiao Hospital, Third Military Medical University, Chongqing 400037, China
Department of Neurobiology, College of Basic Medical Sciences, Chongqing Key Laboratory of Neurobiology, Third Military Medical University, Chongqing
400038, China
c
Department of Physiology, Zunyi Medical College, Zunyi, Guizhou province 563000, China
b

art ic l e i nf o

a b s t r a c t

Article history:
Received 1 June 2014
Received in revised form
6 July 2014
Accepted 7 July 2014
Available online 18 July 2014

Background: Depression is a frequent mood disorder that affects around a third of stroke patients and
has been associated with poorer outcome. Our aim was to determine whether there is a relationship
between serum Brain-derived neurotrophic factor (BDNF) levels and post-stroke depression (PSD).
Methods: Two hundred and sixteen ischemic stroke patients admitted to the hospital within the rst
24 h after stroke onset were consecutively recruited and followed up for 3 months. Based on the
symptoms, diagnoses of depression were made in accordance with DSM-IV criteria for post-stroke
depression at day 90. Enzyme-linked immunosorbent assay (ELISA) was used to measure serum levels of
BDNF at admission. Multivariate analyses were performed using logistic regression models.
Results: In our study, 59 patients (27.3%) were diagnosed as having major depression at 3 months.
Patients with major depression showed lower levels of serum BDNF [8.1 (5.69.4) vs. 13.7 (10.416.5)ng/
ml, Po 0.0001] at admission. In multivariate analyses, serum BDNF was an independent predictor of PSD
at 3 months [odds ratio (OR): 0.79(0.720.87), P 0.003]. Serum levels of BDNF r10.2 ng/ml were
independently associated with post-stroke (OR, 11.5; 95% CI, 5.623.4, P o0.0001), after adjustment for
possible variables.
Conclusion: The present study demonstrates a strong relationship between serum BDNF levels at
admission and the development of PSD within 3 months. Further studies are necessary to conrm this
association, which may open the way to the proposal of new therapeutic options.
& 2014 Elsevier B.V. All rights reserved.

Keywords:
Brain-derived neurotrophic factor
Depression
Acute ischemic stroke
Chinese

1. Introduction
Depression is particularly prevalent among stroke survivors,
affecting approximately a third of individuals (Lindn et al., 2007).
Patients with depression experience worse stroke-related outcomes
in the form of greater functional disability and higher mortality
(Ellis et al., 2010), and, nally, with worse rehabilitation outcome.
Early recognition of depression symptoms and introduction of
pharmacological treatment could lead to better functional outcome
(Zavoreo et al., 2009), making the prevention and management of
post-stroke depression an important area of research.
Neurotrophins are an important class of signaling molecules in
the brain responsible for axon targeting, neuron growth, maturation of synapses during development, and synaptic plasticity
(Autry and Monteggia., 2012). Brain-derived neurotrophic factor
(BDNF) is a neurotrophin that has been linked to the viability of

Corresponding author. Tel.: 86 23 68755613.

E-mail address: cqpek1949@163.com (S.-Y. Cheng).

http://dx.doi.org/10.1016/j.jad.2014.07.011
0165-0327/& 2014 Elsevier B.V. All rights reserved.

neurons in brain circuits (Molendijk et al., 2011). In addition to its

importance in learning, studies have revealed BDNF's involvement
in cognition as well as mood-related behaviors (Autry and
Monteggia., 2012).
One study found that some BDNF gene polymorphisms may be
contributing factors in the pathogenesis of bipolar disorder (Sears
et al., 2011), and several studies reported that blood levels of BDNF
were reduced in patients with schizophrenia (Green et al., 2011).
Recent evidence supports the neurotrophin hypothesis of depression in its prediction that BDNF is involved in depression (Taliaz et
al., 2010). Several works have demonstrated decreased levels in
depressed patients and a recovery after antidepressants treatment
(Gazal et al., 2012; Zhou et al., 2011).
It was reported that BDNF could cross the bloodbrain barrier, and
that BDNF levels in the brain and serum underwent similar changes
during the maturation and aging processes in rats, suggesting that
serum BDNF levels may reect BDNF levels in the brain (Hashimoto,
2010). Pikula et al. (2013) found that lower serum BDNF were
associated with increased risk of incident stroke/TIA, and higher levels
of BDNF were also associated with less white matter hyperintensity

374

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

and better visual memory. Kim et al. (2008) reported that the BDNF
val66met polymorphism may modify the association between stroke
and depression. Thus, the role of BDNF in patients with stroke and
depression excited our interest. In a large cohort, Kim et al. (2012)
found evidence for serotonin and BDNF polymorphisms as susceptibility factors and genegene interactions between these systems for
depression at 2 weeks post-stroke. Interestingly, there is rare study on
serum BDNF levels in Chinese patients with post-stroke depression
(PSD). One study reported that serum concentrations of BDNF
decrease in Chinese PSD patients and BDNF may play an important
role in the pathogenesis of PSD. However, only 93 patients were
included (Zhou et al. 2011). Therefore, our aim was to determine
whether there is a relationship between serum BDNF and PSD in a
large cohort.

examination. The presence of anhedonia and depressive mood

was essential for the diagnosis.
2.4. Laboratory tests
Fasting venous blood was collected from all participants in
vacutainer tubes and quickly centrifuged to avoid glycolysis.
Serum samples were kept at  80 1C until assay. Biomarker
concentrations were measured in a central laboratory by investigators blinded to the clinical outcome and neuroimaging ndings.
BDNF serum levels were measured with sandwich-ELISA, using a
commercial kit according to the manufacturer instructions
(DuoSet ELISA Development, R&D Systems, Inc., USA). The lower
detection limit was 1.6 ng/ml and the line range was 1.650 ng/ml.
The intra-assay coefcient of variation [CV] and inter-assay
CV were 3.56.8% and 4.4%7.5%, respectively.

2. Methods
2.5. Statistical analyses
2.1. Study population
Two hundred and ninety-ve patients with a rst episode of
acute ischemic stroke admitted to our hospital within the rst 24 h
of stroke onset were prospectively included in the study. Patients
with subarachnoid or intracranial hemorrhage, decreased level of
consciousness, severe aphasia or dysarthria, or psychiatric illness,
severe infectious or inammatory diseases, and life expectancy o 3
month were excluded. One hundred and sixty out of 295 patients
(54.2%) were male, with a mean age of 68.9 711.3 years. Seventynine patients were not evaluated at 3 month (38 patients died and
12 refused to attend the follow-up, 10 patients had difculty in
being transported to hospital, and 19 patients were lost to followup); the remaining 216 patients were valid for analysis.
Informed consent was obtained after having provided verbal
and written information to participants or nearest relatives when
relevant. Ethics approval was granted by The Ethics Committee for
Medical Research at the Xin Qiao Hospital, Third Military Medical
University.
2.2. Clinical variables
At baseline, age, sex, body mass index and history of risk factors
were obtained. Stroke subtype was classied according to TOAST
(Trial of ORG 10172 in Acute Stroke Treatment) criteria (Adams et
al., 1993). Routine blood and biochemical tests, brain CT/MRI scan
were performed in all patients at admission. MRI with diffusionweighted imaging (DWI) was available in some patients. The
infarct volume was calculated by using the formula 0.5  a  b  c
(Sims et al., 2009). Stroke severity was evaluated by trained
neurologists using the NIHSS at admission (Brott et al., 1989).
Functional outcome was evaluated by the modied Rankin Scale
(mRS) at 3 month (Bonita, 1988). A favorable functional outcome
was dened as an mRS score of 0 to 2 points, while an unfavorable
functional outcome was dened as an mRS score of 3 to 6 points.
2.3. Psychological measurement
Depression assessments were conducted by a neurologist/
psychiatrist who was unaware of the type, size and location of
the index stroke at the time of 3 months after stroke onset.
Previous history of psychiatric disease and depression, educational
level and people living with the patient were recorded at admission. Patients should nish the Hamilton Rating Scale for Depression
(HAM-D) at 3 months follow-up (Hamilton., 1960). Clinical depression was diagnosed according to DSM-III-R criteria using algorithms based on psychiatric interview and neuropsychiatric

The results are expressed as percentages for categorical variables and as mean (standard deviation, S.D.) or median (interquartile range, IQR) for the continuous variables depending on
their normal distribution. ShapiroWilk tests were used for
normal distribution test. Proportions were compared using the
Chi-square test. Two-group comparison of not normally distributed data was performed using MannWhitney U test, and a twotailed Student's unpaired t-test was used for normally distributed
continuous variables. Spearman's Rank correlation was used
for bivariate correlations. Associations between the severity
of depression evaluated by HAM-D scale and the serum levels of
BDNF were also assessed by using ordered logistic regression
models with multivariate adjustment for possible confounders,
for instance, age, sex, body mass index, stroke syndrome, stroke
etiology, the NIHSS score, infarct volume, vascular risk factors and
a history of depression. The inuence of serum BDNF levels on PSD
was performed by binary logistic regression analysis, which allows
adjustment for above confounding factors. The results are
expressed as adjusted odds ratios (ORs) with the corresponding
95% condence intervals (CIs). Receiver operating characteristic
(ROC) curves were utilized to evaluate the accuracy of serum BDNF
to predict PSD. Area under the curve (AUC) was calculated as
measurements of the accuracy of the test. All statistical analysis
was performed with SPSS for Windows, version 19.0 (SPSS Inc.,
Chicago, IL, USA). Statistical signicance was dened as P o0.05.

3. Results
3.1. Baseline characteristics of study samples
The study cohort consisted of 295 patients at baseline (stroke
admission). By the time of follow-up at 3 months, leaving 216
individuals were included in our study. However, these 216
patients were similar in terms of baseline characteristics [age
(P 0.632), gender (P 0.803), NIHSS (P 0.654) and weight
(P 0.723)] compared to the overall cohort. In the study population, 45.8% were females and the average age was 66.5 710.2
years. The median (quartiles) NIHSS score on admission was 6 (3,
12), and the median time from symptom recognition to admission
to hospital was 4.8 h (IQR, 2.47.5). The number of tissue plasminogen activator-treated patients was 65 (30.1%).
3.2. Main ndings
Ninety-four patients (43.5%) showed depression (major and
minor) at 3 months after admission and in 59 patients (27.3%) this

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

depression was classied as major. The baseline characteristics of

216 stroke patients presented with depression or not are described
in Table 1. Patients with depression were older and more frequently were female, living with offspring, widowhood, higher
admission stroke severity, higher serum levels of Hs-CRP and
lower BDNF. No association was found between etiological subtype
or infarct volume and the presence of depression. Similarly, if the
minor depression were included, we got the equal conclusion.
The results indicated that the median serum BDNF level was
12.4 (IQR, 8.715.5) ng/ml. The serum BDNF levels were signicantly decreased in PSD patients at the time of admission as
compared with stroke patients without depression [8.1 (IQR, 5.6
9.4) ng/ml and 13.7 (IQR, 10.416.5) ng/ml, respectively;
P o0.0001], Fig. 1a. Similarly, if the minor depression were
included, we also found that serum BDNF levels were signicantly
decreased in PSD patients [9.3 (IQR, 7.212.5) ng/ml and 14.5 (IQR,
11.217.4) ng/ml, respectively; P o0.0001], Fig. 1b. Serum BDNF
levels decreased with increasing severity of stroke as dened by
the NIHSS score. There was a negative correlation between levels
of BDNF and the NIHSS (r  0.286, P o0.0001; Fig. 2a.). Similarly,
the lower serum BDNF levels at admission corresponded to the
higher HAM-D score at 3 months (r  0.361, Po 0.0001; Fig. 2b).
BDNF was still signicantly associated with HAM-D score
(  0.304, P 0.009), after controlling for age, gender, body
mass index, stroke etiology, the NIHSS score, infarct volume,
vascular risk factors and a history of depression. In addition, there
was no correlation between level of BDNF and sex (P 0.211), and
age (P 0.326).
Based on ROC curves, the optimal cutoff value of serum BDNF
levels at admission which predicted the development of depression at 3 months was 10.2 ng/ml, which yielded the highest
sensitivity and specicity [80.3% and 81.8%, respectively; area
under the curve (AUC) 0.854, 95% CI: 0.7910.917; P o0.0001].
See Fig. 3a. BDNF levels had a higher prognostic accuracy as
compared to Hs-CRP [AUC 0.58 (0.470.65), P 0.013], HCY [AUC
0.69 (0.510.82), P 0.008] and NIHSS score at admission [AUC

375

0.66 (0.540.77), P 0.007]. In logistic regression analysis, BDNF

levels at admission were independently associated with depression (OR, 0.79; 95% CI, 0.720.87, P 0.003) after adjustment for
age, gender, widowhood, living with offspring, NIHSS on admission, serum levels of HS-CRP and HCY. See Table 2. Serum levels of
BDNF r10.2 ng/ml were independently associated with poststroke (OR, 11.5; 95% CI, 5.623.4, Po 0.0001), after adjustment
for above variables.
Again, if the minor depression were included, we have produced similar results. Based on ROC curves, the optimal cutoff
value of 11.5 ng/ml, which yielded the highest sensitivity and
specicity [73.2% and 70.7%, respectively; area under the curve
(AUC) 0.780, 95% CI: 0.7170.843; P o0.0001]. See Fig. 3b. In
logistic regression analysis, BDNF levels at admission were independently associated with depression (OR, 0.85; 95% CI, 0.760.93,
P 0.006) after adjustment for age, gender, widowhood, living
with offspring, NIHSS on admission, serum levels of HS-CRP and
HCY. See Table 3. Serum levels of BDNF r11.5 ng/ml were independently associated with post-stroke (OR, 6.93; 95% CI, 3.89
12.31, Po 0.0001), after adjustment for above variables.

4. Discussion
Largely in accord with previous ndings and with the neurotrophin
hypothesis of depression (Autry and Monteggia., 2012; Molendijk et
al., 2011; Hashimoto., 2010; Shimizu et al., 2003), our data showed that
serum BDNF levels were low in PSD patients compared with stroke
patients without depression. Our results mainly suggested that serum
BDNF level was a powerful biological marker of risk of developing
post-stroke major depression at 3 month after adjustment by variables, and serum BDNF levelsr10.2 ng/ml were associated with 11.5fold increase in risk of post-stroke depression. Similarly, Yang et al.
(2011) reported that serum BDNF on day 1 after admission may
predict the risk of subsequent PSD, and serum BDNF o 5.86 ng/ml
was independently associated with incident PSD at the acute stage of

Table 1
Basal characteristic of stroke patients with depression and no depression.
Baseline characteristics

Depression patients (n 59)

No depression (n 157)

Pa

Age (years), mean(SD)

Female sex, %
BMI(kg m  2, IQR)
Hypertension, %
Diabetes at baseline, %
Days of hospitalization, median (IQR)
Admission median NIHSS score (IQR)
mRS at follow-up, median (IQR)
Infarct volume (ml), mean(SD)
Widowhood (%)
Living with offspring (%)
Family history of depression, %
TOAST classication (%)
a. Large artery
b. Small artery
c. Cardioembolism
d. Other cause
e. Unknown
Laboratory ndings (Median, IQR)
White cell count,  109/L
Glucose level, mmol/L
Hs-CRP, mg/dL
HCY, umol/L
BDNF, ng/ml

72.8 (11.2)
59.3
26.5 (22.828.5)
49.2
32.2
14 (718)
8 (414)
3 (13)
12.5 (1.6)
40.7
32.2
13.6

63.6 (9.1)
40.8
27.2 (23.029.2)
51.6
29.3
13 (618)
5 (28)
2 (13)
12.2 (1.5)
19.7
12.1
5.7

15.3
16.9
33.9
18.6
15.3

19.1
22.3
38.2
10.8
9.6

0.024
0.015
0.214
0.762
0.624
0.627
0.011
0.221
0.424
0.002
0.001
0.085
0.126

7.8 (5.98.6)
5.45 (4.796.52)
0.80 (0.351.88)
18.2 (14,323.4)
8.1 (5.69.4)

7.6 (5.58.4)
5.39 (4.856.55)
0.55 (0.261.26)
14.9 (11.817.8)
13.7 (10.416.5)

0.512
0.242
0.013
0.008
o 0.0001

Results are expressed as percentages or as medians (IQR) and means (SD).

IQR: interquartile range; SD: standard deviation; Hs-CRP: high-sensitivity C-reactive protein; HCY: homocysteine; BMI: body mass index; BDNF: brain-derived neurotrophic
factor; TOAST: Trial of ORG 10172 in Acute Stroke Treatment; mRS: modied Rankin Scale; NIHSS: National Institutes of Health Stroke Scale.
a

MannWhitney U test, student's t test or chi-square test were used.

376

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

Fig. 1. Serum BDNF levels in stroke patients with depression and no-depression group. MannWhitney U-test. All data are medians and in-terquartile ranges (IQR).
(a) Depression patients were dened as major depression; (b) patients with minor depression were also included.

Fig. 2. Correlation between serum BDNF levels and other predictors. (a) Correlation between serum BDNF levels and the National Institutes of Health Stroke Scale (NIHSS)
score; (b) Correlation between serum BDNF levels and HAM-D score.

stroke (OR 28.992; 95% CI, 8.014104.891; po 0.001 after adjustment). Thus, it may open the way to the proposal of new therapeutic
options in patients with ischemic stroke. In addition, our results also
indicated a signicant negative correlation between HAM-D score, the
severity of depressive symptoms, and serum BDNF levels. Several
studies showed a negative correlation between BDNF levels and
severity of depressive symptoms (Shimizu et al., 2003).
The prevalence of PSD varies over time with an apparent peak 36
months after stroke with a range of 934% during this time-frame and
subsequently decline reaching about 50% of the initial rates at one year
(Whyte and Mulsant., 2002). In our study, we found that 27.3% of
stroke patients were classied as major depression at 3 month, while
the depression prevalence was reported to be ranging from 17 to 62.2%
among Chinese stroke patients (Zhang et al., 2010; Tang et al., 2004;
Cheng et al., 2014). In addition, low circulating BDNF concentrations
have been observed in patients with coronary artery disease, type
2 diabetes mellitus, metabolic syndrome, stroke and physical inactivity
(Autry and Monteggia., 2012; Pikula et al., 2013). Consistent with those
results, in our study, we found low serum BDNF levels in stroke
patients and depression patients. Depression had been widely documented to reduce the expression of BDNF in both animal and clinical
studies (Gazal et al., 2012).
One meta-analysis study demonstrated strong evidence that
BDNF levels were lower in depressed subjects than healthy control

subjects (P o6.8  10  8), and that BDNF levels were signicantly

(P 0.003) increased after antidepressant treatment (Sen et al.,
2008). The other meta-analysis similarly showed that BDNF levels
increased signicantly after antidepressant treatment (effect size:
0.62), and that there was a signicant (P 0.02) correlation
between changes in BDNF level and depression score changes
(Brunoni et al., 2008). Several lines of evidence suggest that the
expression of BDNF may be a downstream target of a variety of
antidepressant treatments; BDNF might therefore be an important
target for therapeutic recovery from depression, and it might
also provide protection against stress-induced neuronal damage
(Hashimoto., 2010).
The relationship between BDNF and stroke remains not completely understood. There is experimental evidence that neurons
and glial cells act as endogenous sources of BDNF after ischemic
and other brain injuries (Sandhofer et al., 2009). Dysfunction of
cerebral vascular BDNF signaling, therefore, may contribute to
disruption of the neurovascular unit, hence to an alteration of
tissue responses to vascular injury (Guo et al., 2008). A small
molecule BDNF mimetic (LM22A-4) when administered immediately after an ischemic stroke in adult mice lead to increased
neurogenesis and improved functional motor recovery (Han et al.,
2012). Therefore, BDNF could reduce stroke risk through its
neurotrophic or its vascular effect (Pikula et al., 2013).

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

377

Fig. 3. Receiver operating characteristic (ROC) curves were utilized to evaluate the accuracy of serum BDNF levels to predict PSD. (a) Depression patients were dened as
major depression; (b) patients with minor depression were also included.

Table 2
Adjusted OR of depression for BDNF levels in stroke patients.
Parameter

ORa

95% CI

Age
Females
Widowhood
Living with offspring
NIHSS on admission
Hs-CRP
HCY
BDNF levels at admission
BDNF levels at admission( r10.2 ng/ml)

1.74
1.22
1.83
1.33
1.11
1.76
1.16
0.79
11.50

1.102.79
1.041.55
1.183.09
1.091.78
1.041.18
1.252.89
1.031.29
0.720.87
5.6023.40

0.024
0.015
0.002
0.001
0.001
0.013
0.008
0.003
P o0.0001

OR: odds ratio; CI: condence interval; NIHSS: National Institutes of Health Stroke
Scale; mRS: modied Rankin Scale; Hs-CRP: high-sensitivity C-reactive protein;
HCY: homocysteine; BDNF: Brain-derived neurotrophic factor.
a

The odds ratio corresponds to a unit increase in the explanatory variable.

Table 3
Adjusted OR of depression (minor depression were included) for BDNF levels in the
stroke patients.
Parameter

ORa

95% CI

Age
Females
Widowhood
Living with offspring
NIHSS on admission
Hs-CRP
HCY
BDNF levels at admission
BDNF levels at admission( r11.5 ng/ml)

1.77
1.22
1.87
1.37
1.13
1.79
1.18
0.85
6.93

1.082.81
1.051.56
1.153.14
1.111.82
1.031.22
1.232.95
1.041.33
0.760.93
3.8912.31

0.019
0.016
0.002
0.001
0.001
0.011
0.007
0.006
P o0.0001

OR: odds ratio; CI: condence interval; NIHSS: National Institutes of Health Stroke
Scale; mRS: Modied Rankin Scale; Hs-CRP: high-sensitivity C-reactive protein;
HCY: homocysteine; BDNF: Brain-derived neurotrophic factor.
a

The odds ratio corresponds to a unit increase in the explanatory variable.

Many etiologies of PSD have been proposed but it is unlikely

that any single hypothesis can explain what appears to be heterogeneous. It is probable that complex interactions between

hormones, neurotransmitters, and environmental factors are involved.

In our study, one found that decreased BDNF levels may be important
in the pathophysiology of depression. One hypothesis would be that
reduced BDNF might reect a genetic vulnerability in patients with
depression. Two studies using mice with a genetic deletion of the
BDNF gene have demonstrated that BDNF play a critical role in
neuronal differentiation and survival (Ernfors et al., 1994; Jones et
al., 1994). Monteggia et al. (2007) showed that conditional BDNF
knockout mice also display an increase in depression-like behavior in
the forced-swim and sucrose preference tests, suggesting that low
production of BDNF may precipitate depressive disorder. Another
possible explanation would be that stress-induced BDNF reductions
might cause neuronal damage, which would in turn lead to acquired
biological vulnerability. Stress, which can precipitate and exacerbate
depression, causes neuronal atrophy and death, especially in the
hippocampus (Shimizu et al., 2003) proposed that stress-induced
changes in the hippocampus may be central to the development of
depression in genetically vulnerable individuals (Rajkowska, 2000).
Levels in PSD may reect the collapse of the stress-adaptation system
and its failure to protect the brain from stress-induced neuronal
degeneration. Third, BDNF has been shown to have antidepressant
effects in animal models of depression (Hashimoto., 2010). It has been
reported that forced swimming decreased BDNF mRNA in particular
regions (CA1, CA3, and the dentate gyrus) of the hippocampus, and
that a combination of physical activity and antidepressant treatment
increased the level of hippocampal BDNF mRNA to well above the
baseline value as well as enhanced swimming time in an animal
model (Russo-Neustadt et al., 2001). BDNF signaling appears to be
sufcient for antidepressant effects, as direct infusion of BDNF into
midbrain areas or the hippocampus induces behavioral responses
that are similar to those produced by antidepressants (Rantamki
et al., 2006).
This study has a number of limitations. The major limitation of
our study was that we were not able to examine the risk factors for
depressive episodes including lack of social support, poverty,
family violence, and increased life stress. In addition, the study
subjects were few and not randomly selected. The study was
conducted in only one clinic. Therefore, our ndings may not be
generalizable to other Chinese stroke patients. Further research is

378

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

needed. Third, the serum levels of BDNF were only measured at

the acute stage of stroke in the patients and, hence, this study
yielded no data regarding when and how long biomarkers were
changed in these patients. Forth, depression assessment was made
only once, at the 3-month follow up, whereas the NIHSS was used
only at the acute stage. In addition, patients who had more severe
stroke died before the 3-month follow up were not included. Some
patients who died and had depression might be excluded. Lastly,
the depressive status might be inuenced by the severity of stroke
itself. Schbitz et al. (2007) found that BDNF may have negative
effects on the course and prognosis of stroke. However, in this
study, the stroke severity was not evaluated at 3 months.
In spite of these limitations, the ndings of this study remained
important and showed that serum BDNF at admission was signicantly reduced and suggested that these alterations might
participate in the pathophysiology of depression symptoms in
stroke patients. Serum BDNF levels at admission could be seen as
one powerful biological marker of risk for developing post-stroke
major depression at 3 month. Further studies are necessary to
conrm this association. Brunoni et al. (2008) found that that
BDNF levels increased signicantly after antidepressant treatment,
and suggested the applicability of BDNF as an efcient and novel
anti-depression tool against depression in patients with ischemic
stroke. Future clinical trials with BDNF should be driven.

Role of funding source

The funding agencies played no role in the design and conduct of the study.

Conict of interest
We wish to conrm that there are no known conicts of interest associated
with this publication and there has been no signicant nancial support for this
work that could have inuenced its outcome.
We conrm that the manuscript has been read and approved by all named
authors and that there are no other persons who satised the criteria for authorship but are not listed. We further conrm that the order of authors listed in the
manuscript has been approved by all of us.
We conrm that we have given due consideration to the protection of
intellectual property associated with this work and that there are no impediments
to publication, including the timing of publication, with respect to intellectual
property. In so doing we conrm that we have followed the regulations of our
institutions concerning intellectual property.
We further conrm that any aspect of work covered in this manuscript that has
involved either experimental animals or human patients has been conducted with
the ethical approval of all relevant bodies and that such approvals are acknowledged within the manuscript.
We understand that the corresponding author is the sole contact for the
editorial process (including Editorial Manager and direct communications with the
ofce). He/she is responsible for communicating with other authors about progress,
submissions of revisions and nal approval of proofs. We conrm that we have
provided a current, correct email address which is accessible by the corresponding
author and which has been congured to accept emails.

Acknowledgment
This research was supported by the fundamental and advanced research
projects of Chongqing (No: cstc2013jcyjA10147). We express our gratitude to all
the patients, the nurses and physicians who participated in this study, and thereby
made this work possible. Authors also acknowledge the contribution of the
reviewers who have helped us to improve the manuscript.

References
Adams, H.P., Bendixen, B.H., Kappelle, L.J., Biller, J., Love, B.B., Gordon, D.L., Marsh, E.,
1993. Classication of subtype of acute ischemic stroke. Denitions for use in a
multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment.
Stroke 24 (1), 3541.
Autry, A.E., Monteggia, L.M., 2012. Brain-derived neurotrophic factor and neuropsychiatric disorders. Pharmacol. Rev. 64 (2), 238258.
Brott, T., Marler, J.R., Olinger, C.P., Adams, H.P., Tomsick, T., Barsan, W.G., Walker, M.,
1989. Measurements of acute cerebral infarction: lesion size by computed
tomography. Stroke 20 (7), 871875.

Bonita, R, Beaglehole, R, 1988. Modication of rankin scale: recovery of motor

function after stroke. Stroke 19, 14971500.
Brunoni, A.R., Lopes, M., Fregni, F., 2008. A systematic review and meta-analysis of
clinical studies on major depression and BDNF levels: implications for the role
of neuroplasticity in depression. Int. J. Neuropsychopharmacol. 11, 11691180.
Cheng, S.Y., Zhao, Y.D., Li, J., et al., 2014. Plasma levels of glutamate during stroke is
associated with development of post-stroke depression. Psychoneuroendocrinology 47, 126135.
Ellis, C., Zhao, Y., Egede, L.E., 2010. Depression and increased risk of death in adults
with stroke. J. Psychosom. Res. 68, 545551.
Ernfors, P., Lee, K.F., Jaenisch, R., 1994. Mice lacking brain-derived neurotrophic
factor develop with sensory decits. Nature 368, 147150.
Gazal, M., Motta, L.S., Wiener, C.D., Fernandes, J.C., Quevedo, L.., Jansen, K.,
Oses, J.P., 2012. Brain-derived neurotrophic factor in post-partum depressive
mothers. Neurochem. Res. 37 (3), 583587.
Green, M.J., Matheson, S.L., Shepherd, A., Weickert, C.S., Carr, V.J., 2011. Brainderived neurotrophic factor levels in schizophrenia: a systematic review with
meta-analysis. Molecular psychiatry 16 (9), 960972.
Guo, S., Kim, W.J., Lok, J., Lee, S.R., Besancon, E., Luo, B.H., Lo, E.H., 2008.
Neuroprotection via matrix-trophic coupling between cerebral endothelial cells
and neurons. Proc. Natl. Acad. Sci. 105 (21), 75827587.
Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry
23 (1), 5662.
Han, J., Pollak, J., Yang, T., Siddiqui, M.R., Doyle, K.P., Taravosh-Lahn, K., Buckwalter,
M.S., 2012. Delayed administration of a small molecule tropomyosin-related
kinase B ligand promotes recovery after hypoxicischemic stroke. Stroke 43 (7),
19181924.
Hashimoto, K., 2010. Brainderived neurotrophic factor as a biomarker for mood
disorders: an historical overview and future directions. Psychiatry Clin.
Neurosci. 64 (4), 341357.
Jones, K.R., Farias, I., Backus, C., Reichardt, L.F., 1994. Targeted disruption of the
BDNF gene perturbs brain and sensory neuron development but not motor
neuron development. Cell 76 (6), 989999.
Kim, J.M., Stewart, R., Bae, K.Y., Kim, S.W., Kang, H.J., Shin, I.S., Yoon, J.S., 2012.
Serotonergic and BDNF genes and risk of depression after stroke. J. Affect.
Disord. 136 (3), 833840.
Kim, J.M., Stewart, R., Kim, S.W., Yang, S.J., Shin, I.S., Kim, Y.H., Yoon, J.S., 2008. BDNF
genotype potentially modifying the association between incident stroke and
depression. Neurobiol. Aging 29 (5), 789792.
Lindn, T., Blomstrand, C., Skoog, I., 2007. Depressive disorders after 20 months in
elderly stroke patients a case-control study. Stroke 38 (6), 18601863.
Molendijk, M.L., Bus, B.A., Spinhoven, P., Penninx, B.W., Kenis, G., Prickaerts, J.,
Elzinga, B.M., 2011. Serum levels of brain-derived neurotrophic factor in major
depressive disorder: statetrait issues, clinical features and pharmacological
treatment. Mol. Psychiatry 16 (11), 10881095.
Monteggia, L.M., Luikart, B., Barrot, M., Theobold, D., Malkovska, I., Nef, S., Nestler, E.J.,
2007. Brain-derived neurotrophic factor conditional knockouts show gender
differences in depression-related behaviors. Biol. Psychiatry 61 (2), 187197.
Pikula, A., Beiser, A.S., Chen, T.C., Preis, S.R., Vorgias, D., DeCarli, C., Seshadri, S.,
2013. Serum brainderived neurotrophic factor and vascular endothelial
growth factor levels are associated with risk of stroke and vascular brain injury
framingham study. Stroke 44 (10), 27682775.
Rajkowska, G., 2000. Postmortem studies in mood disorders indicate altered
numbers of neurons and glial cells. Biol. Psychiatry 48 (8), 766777.
Rantamki, T., Knuuttila, J.E., Hokkanen, M.E., Castrn, E., 2006. The effects of acute
and long-term lithium treatments on trkB neurotrophin receptor activation in
the mouse hippocampus and anterior cingulate cortex. Neuropharmacology 50
(4), 421427.
Russo-Neustadt, A., Ha, T., Ramirez, R., Kesslak, J.P., 2001. Physical activity
antidepressant treatment combination: impact on brain-derived neurotrophic
factor and behavior in an animal model. Behav. Brain Res. 120 (1), 8795.
Sears, C., Markie, D., Olds, R., Fitches, A., 2011. Evidence of associations between
bipolar disorder and the brainderived neurotrophic factor (BDNF) gene.
Bipolar Disord. 13 (78), 630637.
Sandhofer, A., Tatarczyk, T., Kirchmair, R., Iglseder, B., Paulweber, B., Patsch, J.R.,
Schratzberger, P., 2009. Are plasma VEGF and its soluble receptor sFlt-1
atherogenic risk factors? Cross-sectional data from the SAPHIR study. Atherosclerosis 206 (1), 265269.
Schbitz, W.R., Steigleder, T., Cooper-Kuhn, C.M., Schwab, S., Sommer, C., Schneider,
A., Kuhn, H.G., 2007. Intravenous brain-derived neurotrophic factor enhances
poststroke sensorimotor recovery and stimulates neurogenesis. Stroke 38 (7),
21652172.
Sen, S., Duman, R., Sanacora, G., 2008. Serum brain-derived neurotrophic factor,
depression, and antidepressant medications: meta-analyses and implications.
Biol. Psychiatry 64 (6), 527532.
Shimizu, E., Hashimoto, K., Okamura, N., Koike, K., Komatsu, N., Kumakiri, C., Iyo, M.,
2003. Alterations of serum levels of brain-derived neurotrophic factor (BDNF)
in depressed patients with or without antidepressants. Biol. Psychiatry 54 (1),
7075.
Sims, J.R., Gharai, L.R., Schaefer, P.W., Vangel, M., Rosenthal, E.S., Lev, M.H.,
Schwamm, L.H., 2009. ABC/2 for rapid clinical estimate of infarct, perfusion,
and mismatch volumes. Neurology 72 (24), 21042110.
Taliaz, D., Stall, N., Dar, D.E., Zangen, A., 2010. Knockdown of brain-derived
neurotrophic factor in specic brain sites precipitates behaviors associated
with depression and reduces neurogenesis. Mol. Psychiatry 15 (1), 8092.

J. Li et al. / Journal of Affective Disorders 168 (2014) 373379

Tang, W.K., Chan, S.S., Chiu, H.F., Wong, K.S., Kwok, T.C., Mok, V., Ungvari, G.S., 2004.
Can the geriatric depression scale detect poststroke depression in Chinese
elderly? J. Affect. Disord. 81 (2), 153156.
Whyte, E.M., Mulsant, B.H., 2002. Post stroke depression: epidemiology, pathophysiology, and biological treatment. Biol. Psychiatry 52, 253264.
Yang, L., Zhang, Z., Sun, D, Xu, Z., Yuan, Y., Zhang, X., Li, L., 2011. Low serum BDNF
may indicate the development of PSD in patients with acute ischemic stroke.
Int. J. Geriatr. Psychiatry 26 (5), 495502.

379

Zavoreo, I., Bai-Kes, V., Bosnar-Pureti, M., Demarin, V., 2009. Post-stroke
depression. Acta Clin. Croat. 48 (3), 329333.
Zhang, T., Wang, C., Liu, L., Zhao, X., Xue, J., Zhou, Y., Wang, Y., 2010. A prospective
cohort study of the incidence and determinants of post-stroke depression
among the mainland Chinese patients. Neurol. Res. 32 (4), 347352.
Zhou, Z., Lu, T., Xu, G., Yue, X., Zhu, W., Ma, M., Liu, X., 2011. Decreased serum brainderived neurotrophic factor (BDNF) is associated with post-stroke depression
but not with BDNF gene Val66Met polymorphism. Clin. Chem. Lab. Med. 49 (2),
185189.