Serum brain-derived neurotrophic factor (BDNF) in sleep-disordered patients: Relation to sleep stage N3 and rapid eye movement (REM) sleep across diagnostic entities

Abstract

Experimental and clinical evidence suggests an association between neuroplasticity, brain-derived neurotrophic factor and sleep. We aimed at testing the hypotheses that brain-derived neurotrophic factor is associated with specific aspects of sleep architecture or sleep stages in patients with sleep disorders. We included 35 patients with primary insomnia, 31 patients with restless legs syndrome, 17 patients with idiopathic hypersomnia, 10 patients with narcolepsy and 37 healthy controls. Morning serum brain-derived neurotrophic factor concentrations were measured in patients and controls. In patients, blood sampling was followed by polysomnographic sleep investigation. Low brain-derived neurotrophic factor levels were associated with a low percentage of sleep stage N3 and rapid eye movement sleep across diagnostic entities. However, there was no difference in brain-derived neurotrophic factor levels between diagnostic groups. Our data indicate that serum levels of brain-derived neurotrophic factor, independent of a specific sleep disorder, are related to the proportion of sleep stage N3 and REM sleep. This preliminary observation is in accordance with the assumption that sleep stage N3 is involved in the regulation of neuroplasticity.

Full text

Serum brain-derived neurotrophic factor (BDNF) in
sleep-disordered patients: relation to sleep stage N3 and rapid
eye movement (REM) sleep across diagnostic entities
MICHAEL DEUSCHLE
1
, MICHAEL SCHREDL
1
, CHRISTIAN WISCH
1
,
CLAUDIA SCHILLING
1
, MARIA GILLES
1
, OLGA GEISEL
2
and
RAINER HELLWEG
2
1
Central Institute of Mental Health, Department of Psychiatry and Psychotherapy, Medical Faculty Mannheim, University of Heidelberg,
Mannheim, Germany;
2
Department of Psychiatry and Psychotherapy, Charit
e, Berlin, Germany
Keywords
neuroplasticity, synaptic homeostasis theory,
neurotrophic factors
Correspondence
Michael Deuschle, MD, Central Institute of
Mental Health, J5, 68159 Mannheim, Germany.
Tel.: 0049 621 1703 2331;
fax: 0049 621 1703 2325
e-mail: michael.deuschle@zi-mannheim.de
Registration: German Clinical Trials
Registration: DRKS00008902
Accepted in revised form 18 May 2017; received
14 February 2017
DOI: 10.1111/jsr.12577
SUMMARY
Experimental and clinical evidence suggests an association between
neuroplasticity, brain-derived neurotrophic factor and sleep. We aimed at
testing the hypotheses that brain-derived neurotrophic factor is associ-
ated with specific aspects of sleep architecture or sleep stages in
patients with sleep disorders. We included 35 patients with primary
insomnia, 31 patients with restless legs syndrome, 17 patients with
idiopathic hypersomnia, 10 patients with narcolepsy and 37 healthy
controls. Morning serum brain-derived neurotrophic factor concentrations
were measured in patients and controls. In patients, blood sampling was
followed by polysomnographic sleep investigation. Low brain-derived
neurotrophic factor levels were associated with a low percentage of sleep
stage N3 and rapid eye movement sleep across diagnostic entities.
However, there was no difference in brain-derived neurotrophic factor
levels between diagnostic groups. Our data indicate that serum levels of
brain-derived neurotrophic factor, independent of a specific sleep
disorder, are related to the proportion of sleep stage N3 and REM
sleep. This preliminary observation is in accordance with the assumption
that sleep stage N3 is involved in the regulation of neuroplasticity.
INTRODUCTION
There is evidence from rodent research that neuroplasticity
and sleep are intertwined phenomena. Especially, sleep
stage N3, or slow-wave sleep, is assumed to be a sensitive
marker of cortical synaptic strength and network synchro-
nization (Esser et al., 2007). Moreover, it has been shown
that cortical brain-derived neurotrophic factor (BDNF), a
modulator of neuroplasticity, induces sleep stage N3 in the
subsequent sleep period (Faraguna et al., 2008). From
findings in adolescent mice it could be concluded that sleep
is associated with neuronal spine loss (Maret et al., 2012).
Tononi’s synaptic homeostasis hypothesis proposes that
“sleep is the price the brain pays for plasticity”: synaptic
potentiation may occur primarily in the awake stage, when
the individual interacts with the environment, while renormal-
ization of synaptic strength and neuronal spine loss may
happen mainly during sleep (Tononi and Cirelli, 2014). This
hypothesis is based on rodent research, but may provide a
framework for understanding the relationship between sleep,
neuroplasticity and learning in healthy subjects and patients
with neuropsychiatric disorders.
In humans, there are several sources of evidence asso-
ciating neurotrophic factors with sleep. First, the BDNF
Val66Met genotype is related to polysomnographic features,
with Met carriers showing decreased spectral power in the
alpha band in N1 stage and decreased theta power in N2 and
sleep stage N3 (Guindalini et al., 2014). In contrast, homozy-
gous Val carriers had higher sleep stage N3 intensity
compared with Val/Met carriers (Bachmann et al., 2012).
Second, in healthy controls and patients with a lifetime
diagnosis of restless legs syndrome (RLS) or periodic limb
movement (Giese et al., 2013), as well as in female patients
with disturbed sleep (Nishich et al., 2013), sleep distur-
bances are related to low BDNF. In contrast, in patients with
narcolepsy being characterized by daytime sleepiness and
increased rapid eye movement (REM) sleep, serum BDNF
was found to be increased (Klein et al., 2013).
ª2017 European Sleep Research Society 1
J Sleep Res. (2017) Regular Research Paper

Next to these epidemiological and clinical observations, the
association of sleep with BDNF has mainly been examined in
pharmacological studies in depressed patients, as it is widely
accepted that the expression of BDNF is reduced in the brain
and blood of patients with affective disorders (Lee et al.,
2007). First, it was shown that in depressed patients, sleep
disturbance is related to low plasma levels of BDNF
(Dell’Osso et al., 2010). In addition, ketamine has been
identified to regulate sleep stage N3 and brain BDNF levels in
depressed patients in a coordinated manner (Duncan et al.,
2014). Lastly, it has repeatedly been shown that antidepres-
sants acting on monoamines may increase BDNF concen-
trations in animals and depressed patients (Brunoni et al.,
2008; Nibuya et al., 1995). Within this context, however, a
considerable heterogeneity was observed with some antide-
pressants having strong effects, while others may hardly
change BDNF concentrations (Molendijk et al., 2011). Our
research showed the effect of various antidepressants on
serum BDNF to differ (amitriptyline > paroxetine; mirtazapine
> venlafaxine; Deuschle et al., 2013; Hellweg et al., 2008).
Based on these data, it may be hypothesized that antide-
pressants with sleep-promoting properties (amitriptyline,
mirtazapine) have stronger effects on serum BDNF than
antidepressants without major effects on sleep (paroxetine,
venlafaxine). These findings contributed to the neurotrophin
hypothesis of depression (Duman and Monteggia, 2006),
with stress and neuroplasticity being considered key ele-
ments in the pathophysiology of affective disorders (MacQu-
een and Frodl, 2011). In contrast to depression, BDNF levels
in sleep disorders received less attention.
Taken together, substantial experimental and clinical
evidence suggests an association between daytime neuro-
plasticity and BDNF on the one hand and nighttime sleep on
the other. However, it is not clear whether BDNF, as a
presumable marker of neuroplasticity, is related to sleep
efficiency or duration per se or rather to a specific sleep
stage. Our study tested the hypotheses that morning BDNF is
related to: (1) specific sleep disorders; or (2) sleep efficiency
or specific sleep stages in the following night. We investi-
gated a heterogeneous group of patients with sleep disorders
rather than a homogenous group of healthy controls in order
to cover more variance of sleep variables.
MATERIALS AND METHODS
Subjects
This study was approved by the local ethics committee of the
Medical Faculty Mannheim, University of Heidelberg, regis-
tered at German Clinical Trials Register (DRKS00008902),
and all subjects gave fully informed written consent prior to
the investigation. Thirty-five patients with primary insomnia,
31 patients with RLS, 17 patients with idiopathic hypersom-
nia, 10 patients with narcolepsy and 37 healthy controls were
included (Table 1). In our patient sample, 19 subjects were
smokers and 74 were non-smokers. Except in the RLS
group, we included only subjects with periodic limb move-
ment with arousal index (PLMI) <5h
1
(all subjects: PLMI
with arousals 0–4.8 h
1
). Also, we excluded all subjects with
an apnea–hypopnea index (AHI) of 5 or more per hour (all
subjects, except one narcolepsy patient: AHI: 0–4.5 h
1
). In
line with their rather young age, there were only a few
patients suffering from physical disorders, which were all
considered not to be related to the sleep disorder: hypothy-
roidism (three RLS, seven insomnia, one hypersomnia);
hypertension (six RLS, six insomnia); arthrosis; lumbago or
pain (six RLS, two insomnia, one narcolepsy); type 2
diabetes (one RLS, one insomnia, one narcolepsy); airway
disorders [one asthma bronchiale (insomnia); one chronic
obstructive pulmonary disease (RLS)]; mostly with adequate
treatments. Four patients had psychiatric diagnoses and
suffered from current mild to moderate depression (one
Table 1 BDNF serum concentrations as well as sleep parameters of patients with sleep disorders and healthy controls
Primary insomnia
(n=35)
RLS
(n=31)
Idiopathic
hypersomnia
(n=17)
Narcolepsy
(n=10)
Healthy
controls
(n=37)
ANCOVA: effect of
diagnosis (covariates:
age, nicotine)
Sex (f/m) 22/13 15/16 6/11 5/5 24/13
Age (years) 47.2 11.4 45.8 15.5 29.2 10.1 37.3 16.6 49.2 11.3 F
4,125
=8.66; P=0.001
BMI (kg m
2
) 24.8 3.5 25.1 4.8 24.9 4.0 26.2 2.3 24.8 3.5 n.s.
Polysomnography
Total sleep time (min) 365 57 342 66 385 38 362 52 n.a. F
3,89
=2.21; P=0.092
Sleep latency (min) 15.5 11.1 27.0 38.9 13.3 8.7 15.0 9.9 n.a. n.s.
WASO (min) 71 47 59 47 38 28 65 42 F
3,89
=2.26; P=0.087
Sleep efficiency (%) 80.0 11.7 78.0 12.6 87.6 6.8 76.9 20.4 n.a. F
3,89
=2.56; P=0.060
N1 stage (%) 10.1 4.3 12.1 7.6 9.0 4.3 18.0 11.0 n.a. F
3,89
=4.84; P=0.004
N2 stage (%) 51.0 9.7 47.0 10.3 52.8 4.7 40.4 15.0 n.a. F
3,89
=4.25; P=0.007
N3 stage (%) 7.6 7.5 10.9 11.8 12.3 7.8 5.1 8.5 n.a. n.s.
REM (%) 14.8 6.1 15.4 5.9 16.9 4.9 21.4 9.2 n.a. F
3,89
=3.22; P=0.026
BDNF (pg L
1
) 4352 1403 4217 1256 3804 1329 3651 1671 4139 1359 n.s.
BDNF, brain-derived neurotrophic factor; BMI, body mass index; REM, rapid eye movement; RLS, restless legs syndrome; WASO, wake after
sleep onset.
ª2017 European Sleep Research Society
2M. Deuschle et al.

narcolepsy) or major depressive disorder in remission (two
RLS) or obsessive compulsive disorder (one insomnia).
Some patients had been using Z-drugs, benzodiazepines or
sedating antidepressants that had been discontinued at least
6 days before polysomnography (18 insomnia, eight RLS).
All other drug treatments were continued.
Diagnostic and study procedures
Our sleep laboratory is a referral centre for patients with
probable neuropsychiatric sleep disorders. Patients were
recruited consecutively from our clinical outpatient depart-
ment for inclusion in the study. All diagnostics were
performed within routine diagnostic procedures according to
ICSD-2 criteria. Organic, substance-related or psychiatric
causes of sleep disorders were excluded by means of clinical
interview, physical examination, electrocardiogram (ECG)
and laboratory investigations. Blood was drawn after the
adaptation night at 08:30 hours, and serum was immediately
frozen and stored at 80°C. Similar to sleep laboratory
patients, healthy controls underwent physical examination
and clinical interview to exclude psychiatric disorders and
physical disorders that may affect sleep or BDNF in serum. In
healthy controls we found no clinical evidence for sleep
disorders by examination or interview, and blood was drawn
using the same procedures as in patients.
Polysomnography
In patients, but not in controls, polysomnography was
performed using a standard polysomnography montage
according to the criteria of the American Academy of Sleep
Medicine (AASM). This included electroencephalography
(EEG) in seven derivations (F4-A1, C4-A1, O2-A1, Cz-A1,
F3-A2, C3-A2 and O1-A2), left and right electrooculography,
chin electromyography, surface electromyography of both
tibialis anterior muscles, and recording of ECG and respira-
tory variables. The EEG sampling rate was 256 s
1
. Sleep
stage scoring and detection of arousals for each 30-s epoch
was performed visually according to standard AASM proce-
dures (Berry et al., 2015). All patients were investigated by
polysomnography for two consecutive nights, with the first
night being considered an adaptation night. During the
second night, we determined sleep latency and efficiency
as well as percentage of sleep stages N1, N2 and N3 and
REM sleep.
BDNF
Blood was drawn, centrifuged (800 gfor 15 min) and serum
samples stored at 80°C until concentrations of BDNF were
determined. BDNF serum concentrations were quantified by
a modified enzyme immunoassay (Promega, Madison, WI,
USA), as described previously (Deuschle et al., 2013; Hell-
weg et al., 2008). This assay has a detection limit of
0.7 pg mL
1
serum BDNF, the coefficients of inter- and
intra-assay variation are 34.1% and 6.7%, respectively
(Hellweg et al., 2006, 2008; Ziegenhorn et al., 2007).
Statistics
First, we tested the association of age, body mass index
(BMI), sex and smoking status (Giese et al., 2014) with
BDNF using ANCOVA in order to identify confounders. Age
(F
1,87
=2.3; P=0.12; r=0.19; P=0.07) and nicotine use
(F
1,87
=3.7; P=0.059) were related, by trend, with BDNF
and were considered covariates in the next steps of analysis,
while BMI and sex were not related to BDNF. In the second
step, we used ANCOVA with age and nicotine use as covariates
to test the association of sleep disorder diagnoses with
BDNF. In the third step, we used univariate ANOVA and
multiple linear regression with sleep parameters (sleep
efficiency; percentage of REM, N3 and combined stage N1
and N2 sleep) as independent parameters, age and nicotine
use as covariates, and BDNF as dependent parameter. In a
fourth and explorative step, we added the latency of the first
REM period or arousal index in total sleep time or wakeful-
ness after sleep onset to the model. Because the Kol-
mogorov–Smirnov test rejected the hypothesis of normal
distribution for all relevant sleep variables (sleep-onset
latency, sleep efficiency, combined stage N1 and N2 sleep,
sleep stage N3 sleep, REM sleep), we used ln-transformed
variables. Statistical significance was assumed at the alpha
level of 0.05.
RESULTS
Sleep disorders and polysomnography
Controlling for age and nicotine use, we found significant
differences with regard to stage N1 and stage N2 and REM
sleep between the groups of patients with sleep disorders. All
polysomnography sleep parameters showed a pattern that
was in accordance with the clinical diagnoses (Table 1).
BDNF and sleep disorders
Age differed significantly between groups of patients with
sleep disorders and healthy controls (t-test: P=0.013), as
well as within diagnostic subgroups (ANOVA:F
3,89
=8.33;
P=0.001). Especially, patients with primary hypersomnia
were significantly younger than healthy controls. Also, age
was related to BDNF and, therefore, controlled for as a
potential confounder. Using ANCOVA, controlled for age and
nicotine use, we did not find significant group differences in
BDNF levels between sleep disorder patients and healthy
controls (Table 1).
BDNF and sleep parameters
Controlling for age and nicotine use, our model (independent
variables: sleep efficiency; percentage of REM, N3 and
ª2017 European Sleep Research Society
BDNF is related to sleep stage N3 and REM sleep 3

combined stage N1 and N2 sleep) was of significance with
regard to BDNF (F
6,79
=2.57; P=0.025): the covariate age
(F
1,79
=4.03, b=0.28; P=0.061), but not nicotine use, was
related by trend with BDNF. Regarding the sleep variables,
we found significant associations of sleep stage N3
(F
1,78
=5.37; P=0.023) and REM sleep (F
1,79
=5.31;
P=0.024) with BDNF. There was no association of stage
1 and 2 sleep or sleep efficiency (all F<2.1) with BDNF.
Accordingly, a multiple linear regression model with BDNF as
dependent variable and sleep stage N3 (b=0.40;
P=0.007), REM (b=0.31; P=0.020), sleep efficiency
(n.s.), stage N1 and N2 sleep (n.s.) and age (b=0.29;
P=0.021) as independent variables was of significance
(F
5,80
=2.88; P=0.019).
BDNF and latency of first REM episode, arousal index,
wakefulness after sleep onset
In an explorative approach we added other sleep variables to
the above-mentioned model. Adding the latency of the first
REM episode showed a significant association with BDNF
(F
1,75
=6.02; P=0.016) without changing the effects of N3
(F
1,75
=4.78; P=0.032) or REM sleep (F
1,75
=10.56;
P=0.002). Adding wake after sleep onset (WASO) to the
model revealed a trend association (F
1,78
=3.65; P=0.060)
and diminished the effects of sleep stage N3 (F
1,78
=2.25;
n.s.) and REM sleep (F
1,78
=3.95; P=0.050). Adding
arousal index in total sleep time (F
1,78
=0.005; n.s.) or total
sleep time (F
1,78
=0.041; n.s.) to the model did not reveal
additional effects.
DISCUSSION
We tested the hypotheses that morning BDNF is related to:
(1) specific sleep disorders; or (2) specific sleep stages in the
following night and, to the best of our knowledge, this is the
first study using polysomnography to investigate a potential
association between BDNF in serum and specific sleep
stages in patients with sleep disorders. First, our data
indicate that BDNF in serum is not significantly related to a
specific sleep disorder. Second, independent of the nature of
a specific sleep disorder, low percentage of sleep stage N3
sleep as well as low percentage of REM sleep are related to
low serum BDNF.
With regard to our first observation, there is evidence that
narcolepsy is related to increased BDNF (Klein et al., 2013)
and insomnia to low BDNF (Giese et al., 2014). However,
this is the first systematic study including and comparing
various sleep disorders. Our data do not confirm the
assumption that a specific sleep disorder or diagnosis is
related to BDNF and, thus, BDNF may not be considered a
‘diagnostic marker’for a specific sleep disorder.
Regarding our second observation of BDNF being posi-
tively associated to stage N3 sleep and REM sleep latency
and duration, we are not aware of other studies relating
BDNF to specific sleep stages. The association of stage N3
with BDNF lost significance after adding WASO to the model,
which might be due to the strong interaction of these
variables. Several psychiatric and sleep disorders are show-
ing both specific changes of sleep stages and BDNF.
Depression, for example, is related to low BDNF (Brunoni
et al., 2008) as well as impaired sleep stage N3 (Riemann
et al., 2001). Also, narcolepsy is related to both increased
REM sleep and BDNF (Klein et al., 2013). Moreover, there is
limited evidence that REM sleep deprivation inhibits BDNF
expression in the rat brain (Sei et al., 2000; Shaffery and
Lopez, 2013). Thus, our findings are in accordance with
independent clinical and experimental observations.
Of course, due to the non-interventional nature of our data,
we may only speculate about the direction of this association.
However, there is some evidence for BDNF to be involved in
the regulation of sleep stage N3 sleep. For example, slow-
wave activity in recovery sleep after sleep deprivation was
found to be higher in BDNF Val/Val compared with Val/Met
genotype (Bachmann et al., 2012). Moreover, BDNF was
shown to have direct effects on rodents’sleep stage N3
regulation: intracerebroventricular BDNF application during
waking state was found to increase slow-wave activity in
subsequent sleep in rats (Faraguna et al., 2008), but also
REM sleep in rabbits (Sei et al., 2000). Our findings are in
accordance with these reports and show BDNF to be
positively related to sleep stage N3 sleep and REM. With
regard to the clinical example of major depressive disorder,
some antidepressants may induce BDNF thereby potentially
leading to improved sleep (Deuschle et al., 2013).
Finally, we consider it a limitation that our healthy controls
could not be investigated with polysomnography. However,
the inclusion of healthy controls did allow us to show that
patients with sleep disorders do not have a general deviation
of BDNF in serum. Also, the heterogeneity, especially with
regard to age, may be considered a limitation for our
analyses. Ideally, future studies should provide information
on power in the delta range.
Taken together, our data indicate that low sleep stage N3
and REM sleep, independent of a specific sleep disorder, are
related to low BDNF. These findings extend the increasingly
acknowledged impact of an interplay between stress and
sleep on BDNF levels (Schmitt et al., 2016).
ACKNOWLEDGEMENTS
The expert technical assistance of Mrs S. Saft, H. Stender
and S. Laubender is appreciated.
AUTHOR CONTRIBUTORSHIP
MD and MS designed the study; MS and CS were respon-
sible for polysomnography and sleep analysis; CW organized
the data bank and was (together with MD and MG) respon-
sible for the statistical analysis; OG and RH did the laboratory
work; MD wrote the first draft of the paper; and all authors
contributed to the discussion.
ª2017 European Sleep Research Society
4M. Deuschle et al.

CONFLICT OF INTEREST
None of the authors reports any conflict of interest.
REFERENCES
Bachmann, V., Klein, C., Bodenmann, S., et al. The BDNF Val66Met
polymorphism modulates sleep intensity: EEG frequency- and
state-specificity. Sleep, 2012, 35: 335–344.
Berry, R. B., Brooks, R. and Gamaldo, C. E. et al.; for the American
Academy of Sleep Medicine. The AASM Manual for the Scoring of
Sleep and Associated Events: Rules, Terminology and Technical
Specifications, Version 2.2. www.aasmnet.org. American Acad-
emy of Sleep Medicine, Darien, Illinois, 2015.
Brunoni, A. R., Lopes, M. and Fregni, F. 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., 2008, 11: 1169–1180.
Dell’Osso, L., Del Debbio, A., Veltri, A., et al. Associations between
brain-derived neurotrophic factor plasma levels and severity of the
illness, recurrence and symptoms in depressed patients. Neu-
ropsychobiology, 2010, 62: 207–212.
Deuschle, M., Gilles, M., Scharnholz, B., et al. Changes of serum
concentrations of brain-derived neurotrophic factor (BDNF) during
treatment with venlafaxine and mirtazapine: role of medication and
response to treatment. Pharmacopsychiatry, 2013, 46: 54–58.
Duman, R. S. and Monteggia, L. M. A neurotrophic model for stress-
related mood disorders. Biol. Psychiatry, 2006, 59: 1116–1127.
Duncan, W., Sarasso, S., Ferrarelli, F., et al. Concomitant BDNF and
sleep slow wave changes indicate ketamine-induced plasticity in
major depressive disorder. Int. J. Neuropsychopharmacol., 2014,
16: 301–311.
Esser, S. K., Hill, S. L. and Tononi, G. Sleep homeostasis and cortical
synchronization: I. Modeling the effects of synaptic strength on
sleep slow waves. Sleep, 2007, 30: 1617–1630.
Faraguna, U., Vyazovskiy, V. V., Nelson, A. B., Tononi, G. and Cirelli,
C. A causal role for brain-derived neurotrophic factor in the
homeostatic regulation of sleep. J. Neurosci., 2008, 28: 4088–
4095.
Giese, M., Untern€
ahrer, E., Brand, S., et al. The interplay of stress
and sleep impacts BDNF level. PLoS ONE, 2013, 8: e76050.
Giese, M., Untern€
ahrer, E., H€
uttig, H., et al. BDNF: an indicator of
insomnia ? Mol. Psychiatry, 2014, 19: 151–152.
Guindalini, C., Mazzotti, D. R., Castro, L. S., et al. Brain-derived
neurotrophic factor gene polymorphism predicts interindividual
variation in the sleep electroencephalogram. J. Neurosci. Res.,
2014, 92: 1018–1023.
Hellweg, R., Lohmann, P., Huber, R., K€
uhl, A. and Riepe, M. W.
Spatial navigation in complex and radial mazes in APP23 animals
and neurotrophin signaling as a biological marker of early
impairment. Learn. Mem., 2006, 13: 63–71.
Hellweg, R., Ziegenhorn, A., Heuser, I. and Deuschle, M. Serum
concentrations of nerve growth factor and brain-derived
neurotrophic factor in depressed patients before and after antide-
pressant treatment. Pharmacopsychiatry, 2008, 41: 66–71.
Klein, C., Bodenmann, S., Berger, W., Brugger, P. and Landolt, H.
The BDNF Val66Met polymorphism modulates sleep intensity:
EEG frequency- and state-specificity. Sleep, 2012, 35: 335–344.
Klein, A. B., Jennum, P., Knudsen, S., Gammeltoft, S. and Mikkelsen,
J. D. Increased serum brain-derived neurotrophic factor (BDNF)
levels in patients with narcolepsy. Neurosci. Lett., 2013, 544: 31–35.
Lee, B.-H., Kim, H., Park, S.-H. and Kim, Y.-K. Decreased plasma
BDNF level in depressive patients. J. Affect. Disord., 2007, 101:
239–244.
MacQueen, G. and Frodl, T. The hippocampus in major depression:
evidence for the convergence of the bench and bedside in
psychiatric research? Mol. Psychiatry, 2011, 16: 252–264.
Maret, S., Faraguna, U., Nelson, A. B., Cirelli, C. and Tononi, G.
Sleep and wake modulate spine turnover in the adolescent mouse
cortex. Nat. Neurosci., 2012, 14: 1418–1420.
Molendijk, M. L., Bus, B., Spinhoven, P., et al. Serum levels of brain-
derived neurotrophic factor in major depressive disorder: state-trait
issues, clinical features and pharmacological treatment. Mol.
Psychiatry, 2011, 16: 1088–1095.
Nibuya, M., Morinobu, S. and Duman, R. S. Regulation of BDNF and
trkB mRNA in rat brain by chronic electroconvulsive seizure and
antidepressant drug treatments. J. Neurosci., 1995, 15: 7539–
7547.
Nishich, R., Nufuji, Y., Washio, M. and Kumagai, S. Serum brain-
derived neurotrophic factor levels are associated with dyssomnia
in females, but not males, among japanese workers. J. Clin. Sleep
Med., 2013, 9: 649–654.
Riemann, D., Berger, M. and Voderholzer, U. Sleep and depression –
results from psychobiological studies: an overview. Biol. Psychol.,
2001, 57: 67–103.
Schmitt, K., Holsboer-Trachsler, E. and Eckert, A. BDNF in sleep,
insomnia, and sleep deprivation. Ann Med., 2016, 48(1–2): 42–51.
doi:10.3109/07853890.2015.1131327.
Sei, H., Saitoh, D., Yamamoto, K., Morita, K. and Morita, Y.
Differential effect of short-term REM sleep deprivation on NGF
and BDNF protein levels in the rat brain. Brain Res., 2000, 877:
387–390.
Shaffery, J. P. and Lopez, J. R. H. Brain-derived neurotrophic factor
(BDNF) reverses the effects of rapid eye movement sleep
deprivation (REMSD) on developmentally regulated, long-term
potentiation (LTP) in visual cortex slices. Neurosci. Lett., 2013,
513: 84–88.
Tononi, G. and Cirelli, C. Sleep and the price of plasticity: from
synaptic and cellular homeostasis to memory consolidation and
integration. Neuron, 2014, 81: 12–34.
Ziegenhorn, A., Schulte-Herbr€
uggen, O., Danker-Hopfe, H., et al.
Serum neurotrophins-A study on the time course and influencing
factors in a large old age sample. Neurobiol. Aging, 2007, 28:
1436–1445.
ª2017 European Sleep Research Society
BDNF is related to sleep stage N3 and REM sleep 5