Author’s Accepted Manuscript

Efficacy of bright light therapy in bipolar

depression

Nese Yorguner Kupeli, Serkut Bulut, Gresa

Carkaxhiu Bulut, Emel Kurt, Kaan Kora

www.elsevier.com/locate/psychres

PII: S0165-1781(17)31113-7
DOI: https://doi.org/10.1016/j.psychres.2017.12.020
Reference: PSY11056
To appear in: Psychiatry Research
Received date: 17 June 2017
Revised date: 23 October 2017
Accepted date: 9 December 2017
Cite this article as: Nese Yorguner Kupeli, Serkut Bulut, Gresa Carkaxhiu Bulut,
Emel Kurt and Kaan Kora, Efficacy of bright light therapy in bipolar depression,
Psychiatry Research, https://doi.org/10.1016/j.psychres.2017.12.020
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
 Efficacy of bright light therapy in bipolar depression

Nese Yorguner Kupeli*1, Serkut Bulut2, Gresa Carkaxhiu Bulut3, Emel Kurt4, Kaan Kora5

1
Department of Psychiatry, Marmara University Pendik Training and Research Hospital,
Istanbul, Turkey,
2
Department of Psychiatry, Mus State Hospital, Mus, Turkey,
3
Department of Child and Adolescent Psychiatry, Mus State Hospital, Mus, Turkey,
4
Department of Psychiatry, Hisar Intercontinental Hospital, Istanbul, Turkey,
5
Private Practice, Vizyon Psychiatry, Istanbul, Turkey,

neseyorguner@gmail.com
serkutbulut@gmail.com
gresacarkaxhiu@gmail.com
emelbuyrazkurt@gmail.com
kaankora@gmail.com

*
Correspondence: Nese Yorguner Kupeli , Address: Department of Psychiatry, Marmara
University Pendik Training and Research Hospital, Ustkaynarca, Pendik, Istanbul, Turkey. Fax:
+90 0216 625 46 39

ABSTRACT
For 30 years, bright light therapy (BLT) has been considered as an effective, well-tolerated
treatment for seasonal affective disorder (SAD). Because of low response rates, new treatment
strategies are needed for bipolar depression (BD), which resembles SAD in certain respects. Few
placebo-controlled studies of BLT efficacy have been carried out for BD. Accordingly, this study
evaluates the efficacy and safety of BLT as an add-on treatment for BD. Thirty-two BD outpatients
were randomly assigned to BLT (10000 lux) or dim light (DL, <500 lux). During a two-week
period, light was administered each morning for 30 min. The Hamilton Rating Scale for Depression
and the Montgomery-Ǻsberg Depression Rating Scale assessed clinical outcome, and the UKU Side
Effects Rating Scale evaluated side effects. No significant difference was observed in baseline
depression scores in the two groups. Response rates for BLT and DL were 81% and 19%, and
remission rates were 44% and 12.5%, respectively. Analyses showed statistically significant
reductions in depression scores for the BLT group compared with the DL group on all scales. Side
effects were similar in both groups, with headache as the most common side effect. The results
suggest that BLT is an effective and safe add-on treatment for BD.

ABBREVIATIONS
ANOVA: Mixed modal repeated measure analysis of variance; BLT: Bright Light Therapy; SAD:
Seasonal Affective Disorder; BD: Bipolar depression; DL: Dim Light; BPD: Bipolar disorder;
SCID-I: Structured Clinical Interview for DSM-IV Axis I Disorders; SIGH-SAD: Structured
Interview Guide for the Hamilton Depression Rating Scale, Seasonal Affective Disorders Version;
MADRS: The Montgomery-Asberg Depression Rating Scale ; HAM-D: The Hamilton Depression
Rating Scale ; PSQI: The Pittsburgh Sleep Quality Index; UKU: The Udvalg for Kilske Undergeselser Side
Effects Rating Scale

1. INTRODUCTION
Bipolar disorder (BPD) is a chronic mental disorder that includes recurring episodes of mania
and depression with a significant effect on functioning and quality of life (Angst, 2007). Patients
with BPD suffer mostly from depression. Bipolar I patients spend three times and bipolar II patients
spend four times more days in depression than in mania (Kupka et al., 2007). Current treatment
options for bipolar depression have low response rates, slow onset action beginning on average in
two to eight weeks, and many side effects (Tseng et al., 2016). Taking into consideration that 30%
of all bipolar patients report suicide attempts during depressive episodes and that treatment
responses in a significant ratio of patients are still low despite mood stabilizers, antidepressants, and
atypical antipsychotics, which have recently provided evidence-based results for treatment, there is
an obvious need for more effective, safe and fast-acting treatment modalities in this patient group
(Köhler et al., 2014; Musetti et al., 2013; Tondo et al., 2015).
Bipolar depression (BD) usually follows a seasonal rhythm, is recurrent, may present with
atypical depressive symptoms, is usually cyclical and has an etiology in which mainly biological
factors play a role. In this respect, seasonal depression and bipolar depression seem to share many
common features (Geoffroy et al., 2014; Geoffroy et al., 2015). Both of these disorders are thought
to share common factors in etiopathogenesis as well as in clinical presentation. As an example,
there are significant dysregulations in melatonin secretion, disruptions in sleep-wake cycles and
circadian rhythm in almost all BPD patients during illness episodes (Tseng et al., 2016). Also, a
number of studies have showed that 11-50% of patients with seasonal depression actually have
BPD (Sohn and Lam, 2004).
Seasonal affective disorder (SAD) was initially defined by Rosenthal et al. in 1984 as a
clinical morbidity with a cyclical course that presents with atypical symptoms such as hypersomnia,
increase in appetite, carbohydrate cravings and psychomotor retardation that present in winter
months and wane in summer. Studies have shown that this subtype of depression responds well to
light therapy (Nussbaumer et al., 2015; Rosenthal et al., 1984). The etiology of SAD seems to be
related to insufficient daylight exposure resulting in melatonin excretion irregularities and
disruptions in the circadian rhythm in genetically predisposed individuals (Pail et al., 2011).
Although the exact mechanism of action of bright light therapy (BLT) is still unclear, due to the
high rates of effectiveness in SAD it is presumed that it regulates the circadian rhythm by acting
through several mechanisms, and reinforces the delayed phase by inhibiting melatonin
(Nussbaumer et al., 2015).
BLT is seen as an effective, fast-acting and well-tolerated treatment for SAD (Rosenthal et
al., 1984), and although still questionable in non-seasonal types of depression, many studies have
reported it to be effective and as a treatment option for BD (Oldham and Ciraulo, 2014). The study
rationale of our current research rests on the hypothesis that given the similarities of the presenting
symptoms, BLT can be an effective treatment modality for BD as it is for SAD.
The number of randomized controlled studies on the effectiveness of BLT for bipolar
depression is very limited, contrary to the abundance of studies done with regard to SAD (Geoffroy
et al., 2015; Tseng et al., 2016). Since many patients with bipolar disorder have to be on a number
of medications for long-term treatment, the availability and efficiency of a fast-acting and a
comparably safer form of treatment must be extensively studied (Terman and Terman, 2005).
Based on these findings, we carried out a 2-week randomized single-blind placebo-controlled
study to evaluate the efficacy of bright light therapy exposure in the mornings as an add-on therapy.
The aims of the study were to assess the remission and the response rates, to evaluate the safety of
BLT and to determine its effect on sleep quality. The hypotheses of this study are: 1) BLT as add-
on therapy is expected to be significantly more effective than DL for bipolar depression; 2) BLT is
expected to be similar to placebo in terms of side effects, 3) because of its effects on circadian
rhythm BLT is also expected to improve the patient’s sleep quality.
2. MATERIALS AND METHOD

This study was conducted at Marmara University, from 2013 to 2014, and approved by
Medical Faculty Ethics Committee of Marmara University. Patients were recruited from the
Outpatient Clinic and the Bipolar Disorder Clinic in Marmara University Pendik Research and
Training Hospital and recruitment period was limited to one year. Since the number of randomized
controlled studies on the effectiveness of BLT on bipolar depression is very limited, contrary to the
abundance of studies done on SAD, we planned to recruit the patients during one year (to include
all seasons), in order to be able to determinate also the non-seasonal effectiveness of BLT. The
patients were fully informed about the study and those agreed to take part in the study signed
written informed consent.

2.1 Participants

The inclusion criteria for this study were: 1) between ages of 18-65; 2) a diagnosis of Bipolar
I or II disorder, depressed phase according to DSM-IV criteria; 3) total score of 17 or more on the
Hamilton Depression Rating Scale (HAM-D) at baseline visit; 4) recipient of an appropriate and
stable dose of maintenance treatment for at least one month prior to enrollment.
Patients with psychotic symptoms, risk of suicide, a chronic and serious medical condition,
ophthalmological, and neurological problems (e.g. migraine, epilepsy, etc.), which could be
triggered by exposure to bright light, were excluded from the study. In order to eliminate the
possible increased placebo effect, which might be caused by the design of the study (visiting the
hospital premises every day for two weeks and the availability of clinical evaluation three times in
this period) patients with mild depression (HAM-D: 7-16) were not included in the study. In
addition, patients for whom a new mood stabilizer, antidepressant, and/or antipsychotic regimen has
been added or increased in dosage in the last month were excluded from the study. Participants
were also excluded who had prior experience with light therapy.
According to the power analysis, the number of patients we planned to recruit the study per
each group was 20. During the recruitment period, 51 patients were referred for the study and all of
them were interviewed. After the assessments, four patients due to risk of suicide, five patients with
mild depression, two patients with migraines, and three patients with psychotic symptoms were
excluded from the study. In addition, five patients did not accept to have admitted to the hospital for
two weeks. As a result, 32 participants with bipolar depression who met all inclusion/exclusion
criteria were enrolled and randomized according to their admission order, to either BLT (n=16) or
dim light (n=16) group.
 2.2. Procedures
All diagnostic interviews were carried out by the same researchers. To validate the diagnosis
of bipolar depression and to assess any comorbid conditions, Structured Clinical Interview for
DSM-IV Axis I Disorders (SCID-I) was administered. In the same interview, a semi-structured data
form, evaluating sociodemographic and clinical features prepared by the researchers; the Structured
Interview Guide for the Hamilton Depression Rating Scale; Seasonal Affective Disorders Version
(SIGH-SAD); the Montgomery-Asberg Depression Rating Scale (MADRS); the Hamilton
Depression Rating Scale (HAM-D); and the Pittsburgh Sleep Quality Index (PSQI) were applied.
Patients were randomized according to their admission order, to either BLT or dim light (DL)
group and they remained blind to treatment assignment during the study. 16 patients in the study
group received BLT with an intensity of 10000 lux, and 16 patients in the placebo group received
light intensity less than 500 lux. During the study period, any change in treatment regimens was not
allowed except lorazepam when needed as an add-on treatment. In both groups, the sessions were
scheduled between 08:00-10:00 A.M. for 30 minutes and for 14 consecutive days.
In both groups, light therapy was carried out in two identical rooms with room lighting by
default below 50 lux, and the intensity of light exposure from the devices were measured by lux
meters. The same rooms were used throughout the study and the placement of the light boxes
remained consistent throughout the study.
After completing the 7th and 14th sessions, the patients were assessed with HAM-D and
MADRS, regarding any change in depressive symptoms. To assess the changes of sleep quality,
patients were asked to fill out the PSQI forms for the previous week’s observations. The UKU Side
Effects Rating Scale was administered to evaluate any side effects with regards to light therapy.
Although not a part of the UKU rating scale, eye and sight problems due to light therapy (blurred
vision, tiredness of eyes, etc.) were also assessed in this interview. A 50% decrease in HAM-D and
MADRS scores sessions was interpreted as “response” and reduction to 7 points in HAM-D and 9
points in MADRS was interpreted as “remission” at the end of study.

2.3. Characteristics of Light Boxes

Bright light therapy was administered by means of the device MELISSA (New Territories,
Hong Kong), with dimensions 50 x 35cm that emits 10000 lux intensity light from two fluorescent
lamps 72 Watts each, when one is sitting 40-70cm’s away from the device.
The DL device used in the study was developed by the research team to exactly replicate the
BLT device. The light’s brightness varies directly with the luminosity and indirectly with the square
of the distance between the source and the object. The measurements with a calibrated lux-meter
revealed that approximately 100-150cm distance from the light source gives 500 lux brightness with
the DL device used in the study.

2.4. Instruments

The semi-structured data form, prepared by the researchers, had 57 questions. This form
included questions addressing clinical characteristics such as atypical, melancholic or catatonic
features; past psychiatric treatment history; suicide attempts; alcohol and drug use disorders;
presence of medical disorders; and family history for psychiatric disorders as well as
sociodemographic data.
The Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) is a clinical and
structured interview form to assess the presence of major DSM-IV Axis I disorders. By facilitating
to increase the reliability of the diagnosis and the screening of DSM-IV diagnostic criteria, it helps
to increase diagnostic validity and the systematic identification of symptoms (Lobbestael et al.,
2011).
The Montgomery-Asberg Depression Rating Scale (MADRS) is a scale developed to assess
the level and severity change in the nuclear depressive symptoms in patients with depression
(Montgomery and Asberg, 1979). It consists of 10 items. All items are scored between 0-6 (Hawley
et al., 2002). Scores between 0-9 points show no depression; 10-28 points indicate mild depression;
29-35 points indicate medium depression and points above 36 indicate severe depression (Müller et
al., 2000).
The Hamilton Depression Rating Scale (HAM-D) is a test used to evaluate the severity of
depression with the presence of symptoms in one-week’s time. Since the test was initially
developed for hospitalized patients, it has an emphasis on melancholic and physical symptoms of
depression. Scoring is based on the 17-item scale and scores of 0–7 indicate no depression, 8–16
suggest mild depression, 17–23 moderate depression and scores over 24 are indicative of severe
depression (Williams, 1988; Zimmerman et al., 2013).
The Structured Interview Guide for the Hamilton Depression Rating Scale, Seasonal
Affective Disorders Version (SIGH-SAD) evaluates atypical depressive symptoms and daily
fluctuations as well as the presence of standard depressive symptomatology. It is a likert type scale
evaluated by the rater with 29 questions: 17 of the questions are from classical HAM-D, four of
them assess other depressive symptoms and eight of them assess atypical depressive symptoms.
Items are scored from 0-4 to 0-2. The total number of points reflects an overall score (Williams et
al., 1988).
The Pittsburgh Sleep Quality Index (PSQI) evaluates the sleep quality and the type and
severity of sleep disorder in a one-month’s period. It consists of 24 questions, 19 of which have
been assessed by the patient himself/herself and five questions assessed by the patient’s sleep/bed
partner. The 19 self-questions inquire about seven dimensions such as subjective sleep quality,
latency of sleep, length of sleep, sleep habits, sleep disturbance, usage of sleep pills, and daytime
dysfunctionality. All items score between 0 (no problem) and 3 (severe problem). Total score to be
obtained from the scale is between 0 and 21. Scores equal to and below five indicate a good quality
of sleep (Buysse et al., 1991).
The UKU Side Effects Rating Scale (UKU-Udvalg for Kilske Undergeselser) is used to
evaluate side effects seen in the use of psychotropic medications with reference to cause-and-effect.
It consists of 52 items with 10 psychological, 10 neurological, 13 autonomic, and 29 other
symptoms. Rather than interpreting scores, the scale helps to compare which medications may
cause which symptoms and in what severity (Lingjaerde et al., 1987). The patients in our study
group were also asked three more questions about eye and sight (e.g. blurred vision, tiredness of
eyes) that are common to light therapy.

2.5. Statistics
The data from this study were analyzed statistically by the “Statistical Program for Social
Sciences” (SPSS) 20.0 version (SPSS Inc, Chicago, IL). Categorical variables between groups were
compared with chi-square test and the continuous variables analyzed by the Mann-Whitney U test.
The change from baseline over the course of treatment period, repeated measures such as
depression scores were assessed by ANOVA. Statistical significance level was taken as p<0.05.
3. RESULTS

Thirty-two patients who met the inclusion criteria were enrolled in the study. Patients were
randomly assigned to two groups: 16 patients in study group receiving BLT with an intensity of
10000 lux, and 16 patients in the placebo group receiving light intensity less than 500 lux.
Sociodemographic characteristics of the patients are presented in Table 1. There were no significant
differences between the groups regarding age, education level and employment except sex. Female
preponderance was significantly higher in placebo group (p<0.05).

3.1. Clinical Characteristics of the Groups and Psychiatric and Family History

17 of the patients (53%) were diagnosed as BPD type I, and 15 (47%) of them as type II.
Mean age for onset of illness was 24+5, and age interval of onset was 15 to 35. For majority of the
patients, the first illness episode was depression (n=24, 75%). The number of depressive episodes
in the whole group was between 2 and 30 and number of manic episodes was between 1 and 30.
In 7 (22%) patients there was a seasonal pattern for depressive episodes. There were no
significant differences between the groups regarding clinical features and psychiatric history as
presented in Table 2. Groups were compared by factors such as bipolar disorder in family history,
family history other than psychiatric morbidity, alcohol use disorder and suicide attempts and there
was no statistically significant difference between the groups. Family history of the groups is
presented in Table 2.

3.2. Pharmacological Agents Used in Treatment

Since all patients included in the study were followed in our Outpatient clinic and the Bipolar
Disorder Clinic, appropriate medication doses were monitored and all participants were adherent to
treatment. Both BLT and dim-light were applied as an add-on treatment to existing pharmacological
treatments. When all medications currently used for treatment were categorized under lithium,
lamotrigine, valproic acid, carbamazepin, antipsychotics and antidepressant categories, there was no
significant difference between the groups. Majority of the treatments consisted of lithium and the
antidepressant usage was similar between the groups. The pharmacological agents used in treatment
are presented in Table 3.

3.3. Depression Scores Before Treatment Sessions

Depression scores prior to light therapy sessions were assessed with HAM-D and MADRS
scales and there was no statistically significant difference between the groups. The mean scores for
HAM-D in BLT group was 24.0+5.0 and between 17 and 36; in DL group was 23.5+6.0; the mean
MADRS scores in BLT group was 32.0+8 and between 22 and 50; in DL group was 27.5+6.

3.4. Scores After Treatment Sessions

After two weeks, HAM-D scores decreased in both groups. In the BLT group, HAM-D scores
decreased by 12+4 points in the first week and 4+5 points in the second week of treatment; in the
DL group, scores decreased by 6+4 points in the first week and increased by 1+3 point in the
second week. Repeated measures ANOVA showed that the decrease was more significant in the
first week than in the second week in the BLT group (F=75.5, df =2, p<0.001). Weekly changes in
HAM-D scores are presented in Figure 1.
MADRS scores also decreased in both groups at the end of the two-week period. In the BLT
group, MADRS scores decreased by 16+8 points in the first week and 5+8 points in the second
week of treatment. In the DL group, MADRS scores decreased by 5+8 points in the first week and
0.5+7 points in the second week. Repeated measures ANOVA showed that the decrease was more
significant in the first week than in the second week in BLT group (F=62, df=2, p<0.001). Weekly
changes in MADRS scores are presented in Figure 2.

3.5. Treatment Response and Remission Rates

The response rate assessed by HAM-D scores in the BLT group was 69% (in 11 patients), and
12.5% (in 2 patients) in DL group. The remission rate in the BLT group was 44% (in 7 patients),
and 6% (in only 1 patient) in DL group. HAM-D scores reflect significant response and remission
rates in the BLT group compared with the placebo group.
The response rate assessed by MADRS scores in the BLT group was 81% (in 13 patients),
and 19% (in 3 patients) in DL group. The remission rate in BLT group was 44% (in 7 patients), and
12.5% (in 2 patients) in DL group. MADRS scores reflect significant response and remission rates
in the BLT group compared with the DL group. The response and remission rates of both groups
are presented in Table 3. Since there are sex differences between the groups, the results of treatment
with regards to sex are also evaluated; no significant difference was found between two groups for
response and remission rates.
 3.6. Sleep Quality Changes After Treatment Sessions

Sleep quality was assessed at three points during the study using the PSQI: once in the
pretreatment phase, once at the end of the first week, and once at the end of the second week. The
differences between groups in PSQI mean scores in the pretreatment phase and at the end of first
week were not statistically significant. At the end of second week, the increase in sleep quality
scores in BLT group was found to be significantly higher than the placebo group (p=0.028).

3.7. Seasonal Pattern

In this study, 25% (n=4) of patients receiving BLT were found to exhibit a seasonal pattern.
All patients with a seasonal pattern qualified for remission criteria whereas nine out of 12 patients
without a seasonal pattern qualified for treatment response and three for remission criteria.
Although there was no significant difference between treatment responders and non-responders in
terms of seasonality (p=0.52), remission rates in patients with a seasonal pattern were significantly
higher compared to those without (p=0.01). The results are presented in Table 4.

3.8. Side Effects

All patients were evaluated with the UKU Side Effects Rating Scale at the end of first and
second weeks. At the end of week one, one patient reported insomnia in the BLT group. After
excluding the possibility of a manic/hypomanic switch, lorazepam 1 mg/day was prescribed for that
patient. Two patients in the BLT group with headaches reported that the symptoms responded well
to nonsteroidal anti-inflammatories. One of these patients reported that the headache disappeared in
the first week, and the other patient reported a migraine diagnosis and her headache continued. One
patient reported a state of confusion and sedation that started right after the light therapy session
and that lasted for a couple of hours. She stated that she stared into the light during the session
although precisely informed not to do so. She was examined physically, biochemically, and
neurologically without any abnormal findings. Her EEG examination was in normal range. Her
lithium levels were found to be within therapeutic range; she wanted to continue sessions and she
completed the study without any adverse event.
In the DL group, at the end of week one, one patient reported nausea, one patient reported
headaches, one patient reported palpitations and headache during the sessions. At the end of second
week, the patient with headache reported disappearance of her complaint whereas others reported
continuation of their complaints. One patient reported irritability in the second week of treatment.
 While side effects in the BLT group showed a decline by week two, they remained until the
end of two weeks in the DL group. There were no statistically significant differences in side effects
observed between the groups (p>0.1). There were no patients who switched to hypomania/mania
and no patients that dropped off from the study due to side effects or any adverse event.

4. DISCUSSION

This randomized, single blind study aimed to address the efficacy of BLT for bipolar
depression when it is administered in the mornings for a two-week period and it has been found that
both treatment response and remission rates in this treatment modality are significantly higher than
dim-light with side effects seen similar to that of the placebo group.
Unipolar depression is a heterogeneous disorder with various symptom presentations and
courses, clinical characteristics and etiopathogenesis including many different factors like genetics,
psychosocial factors, bereavement, loss, unemployment, separation, etc. (Goodwin and Jamison,
2007). It is not easily comprehensible how a disorder’s subsets with such a variation in clinical
characteristics and prognosis can respond well to a single form of treatment. To view and evaluate
seasonal depression in a separate category has enabled the use of BLT as a specific and effective
method of treatment. BLT has proven to be an effective and well-tolerated form of first line
treatment for seasonal depression for the last 30 years (Oldham and Ciraulo, 2014).
Research has shown that bipolar depression seems to present with clinical characteristics,
biological features, course and prognosis that are considerably different than unipolar depression.
This becomes more evident from the observed inefficiency of the treatment modalities for unipolar
depression when used for bipolar depression (Hirschfeld, 2014). Bipolar depression frequently
seems to follow a seasonal rhythm, with repetitive atypical depressive features, is cyclical and has
an etiology that is dominated mostly biological factors (Goodwin and Jamison, 2007). In this
respect, bipolar depression shares many common characteristics with seasonal depression. These
two disorders are thought to share common factors in etiopathogenesis along with clinical
characteristics. As an example, almost all patients diagnosed with bipolar disorder have coexisting
sleep-wake cycle and circadian rhythm patterns during illness episodes. Various studies have shown
that circadian rhythm disruptions play a major role in the pathophysiology of seasonal depression
and that 11-50% of these patients are diagnosed as bipolar disorder (Sohn and Lam, 2004).
Previous studies on BLT responses have been carried out in patients with regards to
depression being seasonal or non-seasonal. Although a multitude of studies carried out in seasonal
depression indicate the superior efficacy of light therapy compared to the placebo in this group,
there are contradicting data from a limited number of studies on its efficacy in the non-seasonal
depression group (Golden et al., 2005). A comprehensive review that evaluated 20 placebo-
controlled studies reported that BLT is superior to placebo in non-seasonal depression. This review
emphasizes that there are more significant differences in the effectiveness of BLT in high quality
studies (Tuunainen et al., 2004). A meta-analysis, based on randomized, controlled trials with
stringent inclusion criteria suggests that BLT is efficacious for SAD and non-SAD with effect sizes
equal to the most antidepressant pharmacotherapy trials (Golden et al., 2005). A recent meta-
analysis, which included 458 patients, revealed that BLT is an effective treatment as an
augmentation therapy compared to antidepressant usage alone. In addition, the effect size of BLT
was found similar to that of other common augmentation strategies (Penders et al., 2016).
In this study, the decision to include patients with bipolar depression was based on the
hypothesis that since the clinical characteristics of bipolar depression resemble that of seasonal
depression, this similarity may as well be reflected in treatment responses to BLT, which is a
proven method of treatment of the latter. This form of treatment may be especially important in the
management of resistant cases of bipolar depression with limited options of treatment and may
prove as an effective alternative for clinicians for long-lasting treatment responses and an
improvement of functionality for patients.
The results from this study indicate that BLT as a treatment for bipolar depression is superior
to placebo in rates of treatment response and remission, 81% of patients in the BLT group had a
good treatment response and 44% justified for remission criteria. The rates for treatment response
and for remission in the control group were found to be 19% and 6% respectively. There are a
limited number of studies evaluating the effectiveness of BLT in bipolar depression (Geoffroy et
al., 2015). One of these studies compares, irrespective to seasonality, the response rates to BLT in
bipolar depression and unipolar depression. At the end of the first week of treatment, patients with
bipolar depression were found to respond three times more to BLT compared to patients with
unipolar depression (90% and 30%, respectively) (Deltito et al., 1991). Another study done with
patients with treatment resistant bipolar depression included sleep deprivation and light therapy in
combination, and found that 62% of patients had a decrease of more than 50% of the total HAM-D
scores in the first six days of treatment (Wu et al., 2009). In another 6-week RCT study with BD,
light therapy administered as an add-on therapy, the BLT group had significantly higher remission
rates (56%) versus the DL group (14.3%) (Sit et al., 2017). In addition, a recent meta-analysis,
which investigated the efficacy of BLT for BD, revealed that BLT is an effective and safe treatment
option as an adjunctive therapy, as in our study (Tseng et al., 2016). According to the American
Psychiatric Association Practice Guideline for the treatment of patients with major depressive
disorder, BLT might be used to treat non-seasonal affective disorder as well as seasonal depression
(APA, 2010).
 In this study, the first week of treatment MADRS scores in the BLT group decreased by
means of 16 and 5 points in the placebo group whereas at the end of the second week, the scores
decreased by 5 and 0.5 respectively. These results convey the possibility that the antidepressant
effects of BLT emerge in the first week and the results are prominent in the first days of treatment.
Although not within the scope of the study, the patients with positive responses to treatment
continued to exhibit both good response and remission rates one month after the study as they were
followed in the same outpatient clinic in which the study was conducted.
In accordance with our findings, all other studies indicate that the antidepressant effects of
BLT begin in the first days and that this effect is prominent in the first week of treatment (Golden et
al., 2005; Terman and Terman, 2005; Tuunainen et al., 2004). This fast acting antidepressant effect
may help clinicians in the time period it takes for current antidepressants to take effect.
In our study no clinical features other than seasonality had an effect on treatment outcomes
for BLT. All patients with a seasonal pattern in BLT group justified for remission criteria.
Literature shows that remission rates may be as high as 75% with BLT in patients with a seasonal
pattern compared with non-seasonal pattern (Terman and Terman, 2005). Our study indicates that
seasonality has no effect on treatment response rates but a significant one on remission.
Although there was no statistically significant difference, it is notable that the placebo group
included a higher number of patients with bipolar disorder type II. The studies investigating the
treatment response in bipolar depression, obtained contradictory results between bipolar I and II in
terms of treatment response rates and treatment resistance (McElroy et al., 2010; Thase et al., 2006;
Young et al., 2013). Low response rates in the placebo group of this study may be a result of the
higher number of patients with bipolar II in the placebo group compared to the study group.
There was no significant difference in sexes for response and remission rates in this study.
Although there is no data with regards to sex differences in other studies, we need to emphasize this
finding since there was a sex difference in our treatment groups.
Although there was no difference in sleep quality findings between the groups in the first and
second evaluations, there was a significant improvement in the sleep quality of patients in BLT
group at the end of the second week of treatment. The prominent effect of light on depressive
symptoms in the first week was not obvious on sleep quality. This finding may suggest that light
has a slower action on sleep compared to other depressive symptoms. Other studies indicate that
light therapy improves sleep quality by acting on sleep phases (Terman and Terman, 2005).
Our study indicates that BLT is not different from placebo with regards to side effects. In the
first week of treatment, 25% of patients in BLT group and 19% in the placebo group reported side
effects like headache, insomnia, and nausea. Headache was the most common side effect reported in
both groups. These side effects decreased in the second week of treatment in BLT group and no
other side effects were observed. Other studies reported side effects in 6-16% of patients and that
they disappear by the end of week one (Terman and Terman, 2005). One study reported that 34% of
patients experienced side effects in the first day of treatment and that a majority of side effects
decreased by the fifth day and that 10% of patients had lasting side effects (Kogan and Guilford,
1998). In that study the side effects in the control group decreased by a lesser degree and there were
new side effects in the second week of treatment. These findings were interpreted as being related
to continuing depressive symptoms in the control group. Some authors who point out to the
limitation that side effects were assessed through patient reports have carried out two placebo
controlled studies with healthy volunteers and have found out that there was no difference between
the groups as in accordance with our findings (Botanov and Ilardi, 2013; Volz et al., 1991).
An important limitation in the treatment of bipolar depression is the presence of the risk to
switch to mania. A review that examines 20 BLT studies with non-seasonal depression found that
the most common side effect was switch to hypomania, which was not a finding in our study
(Tuunainen et al., 2004). Likewise, in a recent randomized controlled trial involving bipolar
depression, light therapy was administered as adjunctive treatment and no hypomanic shift was
observed (Sit et al., 2017). The reason behind the absence of a hypomanic/manic switch in patients
could be explained by the applying of BLT as an add-on treatment to mood stabilizer or atypical
antipsychotics.
Other studies have shown that BLT is well tolerated compared to other pharmacological
agents and that drop-out rates due to side effects are much less (Terman and Terman, 2005). The
rate of side effects in our study was moderate and there were no drop-outs due to the side effects.
According to the power analysis, the number of patients we planned to recruit the study per each
group was 20. However, due to the study protocol, the recruitment phase was limited to one year,
the required number could not be reached. The single-blind design of the study, small sample size,
and the presence of sex differences between the groups are the limitations of this study. In addition,
while no manic shift occurred during this study it would have been better to use a scale to check
manic symptoms.
Consequently, a treatment option that is fast-acting like BLT may facilitate recovery in the
acute phase of bipolar depression and may help to form better adherence and higher remission rates.
There is a need for large sample sized, double-blind controlled studies for establishing the efficacy
and safety of BLT for the treatment of bipolar depression in future.

ACKNOWLEDGEMENT
 We appreciate the valuable assistance of Meltem Kora, M.D. and Banu Hummel Ph.D. in
translating the article. This research did not receive any specific grant from funding agencies in the
public, commercial, or not-for-profit sectors.

References:

American Psychiatric Association. Practice guideline for the treatment of patients with major
depressive disorder, 2010. 3rd ed. Arlington (VA), American Psychiatric Publishing.
Angst, J., 2007. The bipolar spectrum. Br J Psychiatry 190(3), 189-191.
Botanov, Y., Ilardi, S.S., 2013. The acute side effects of bright light therapy: a placebo-controlled
investigation. PloS one. 8(9), e75893.
Buysse, D.J., Reynolds, C.F., Monk, T.H., Hoch, C.C., 1991. Quantification of subjective sleep
quality in healthy elderly men and women using the Pittsburgh Sleep Quality Index (PSQI).
J Sleep Res. Vol 14(4), 331-338.
Deltito, J.A., Moline, M., Pollak, C., Martin, L.Y., Maremmani, I., 1991. Effects of phototherapy on
non-seasonal unipolar and bipolar depressive spectrum disorders. J Affect Disord. 23(4),
231-237.
Geoffroy, P.A., Bellivier, F., Scott, J., Etain, B., 2014. Seasonality and bipolar disorder: a
systematic review, from admission rates to seasonality of symptoms. J Affect Disord. 168,
210-223.
Geoffroy P.A., Fovet, T., Micoulaud-Franchi, J.A., Boudebesse, C., Thomas, P., Etain, B., et
al.,2015. Bright light therapy in seasonal bipolar depressions. Encephale. 41(6), 527-533.
Golden, R.N., Gaynes, B.N., Ekstrom, R.D., Hamer, R.M., Jacobsen, F.M., Suppes, T., el al., 2005.
The efficacy of light therapy in the treatment of mood disorders: a review and meta-analysis
of the evidence. Am J Psychiatry. 162(4), 656-662.
Goodwin F.K., Jamison, K.R., 2007. Manic-depressive illness: bipolar disorders and recurrent
depression, Oxford University Press (Vol. 1).
Hawley, C.J., Gale, T.M., Sivakumaran, T., 2002. Defining remission by cut off score on the
MADRS: selecting the optimal value. J Affect Disord. 72(2): 177-184.
Hirschfeld, R.M., 2014. Differential diagnosis of bipolar disorder and major depressive disorder. J
Affect Disord. 169, S12-S16.
Kogan, A.O., Guilford, P.M., 1998. Side Effects of Short-Term 10,000-Lux Light Therapy. Am J
Psychiatry. 155(2): 293-294.
Köhler, S., Gaus, S., Bschor, T., 2014. The challenge of treatment in bipolar depression: evidence
from clinical guidelines, treatment recommendations and complex treatment situations.
Pharmacopsychiatry. 47(2), 53-59.
Kupka, R.W., Altshuler, L.L., Nolen, W.A., Suppes, T., Luckenbaugh, D.A., Leverich, G.S., et al.,
2007. Three times more days depressed than manic or hypomanic in both bipolar I and
bipolar II disorder. Bipolar Disord. 9(5), 531-535.
Lingjaerde, O.U., Ahlfors, P., Bech, S., Dencker, S.J., Elgen, K., 1987. The UKU side effect rating
scale: A new comprehensive rating scale for psychotropic drugs and a cross‐ sectional study
of side effects in neuroleptic‐ treated patients. Acta Psychiatr Scand. 76(s334), 1-100.
Lobbestael, J., Leurgans, M, Arntz, A., 2011. Inter‐ rater reliability of the Structured Clinical
Interview for DSM‐ IV Axis I disorders (SCID I) and Axis II disorders (SCID II). Clin
psychol psychot. 18(1), 75-79.
McElroy, S.L., Weisler, R.H., Chang, W., Olausson, B., Paulsson, B., Brecher, M., et al., 2010. A
double-blind, placebo-controlled study of quetiapine and paroxetine as monotherapy in
adults with bipolar depression (EMBOLDEN II). J Clin Psychiatry 71(2), 163
Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to be sensitive to change.
Br J Psychiatry 134(4), 382-389.
Musetti, L., Del Grande, C., Marazziti, D., Dell'Osso, L., 2013. Treatment of bipolar depression.
CNS Spectr. 18(04), 177-187.
Müller, M.J., Szegedi, A., Wetzel, H., Benkert, O., 2000. Moderate and severe depression:
Gradations for the Montgomery–Åsberg Depression Rating Scale. J Affect Disord. 60(2),
137-140.
Nussbaumer, B., Kaminski-Hartenthaler, A., Forneris, C.A., Morgan, L.C., Sonis, J.H., Gaynes,
B.N., et al., 2015. Light therapy for preventing seasonal affective disorder. Cochrane
Database Syst Rev. 10.1002/14651858.CD011269.pub2.
Oldham, M.A., Ciraulo, D.A., 2014. Bright light therapy for depression: a review of its effects on
chronobiology and the autonomic nervous system. Chronobiol Int. 31(3), 305-319.
Pail, G., Huf, W., Pjrek, E., Winkler, D., Willeit, M., Praschak-Rieder, N., et al., 2011. Bright-light
therapy in the treatment of mood disorders. Neuropsychobiology. 64(3), 152-162.
Penders, T.M., Stanciu, C.N., Schoemann, A.M., Ninan, P.T., Bloch, R., Saeed, S.A., 2016. Bright
Light Therapy as Augmentation of Pharmacotherapy for Treatment of Depression: A
Systematic Review and Meta-Analysis. Prim Care Companion CNS Disord. 18(5).
Rosenthal, N.E., Sack, D.A., Gillin, J.C., Lewy, A.J., Goodwin, F.K., Davenport, Y., et al., 1984.
Seasonal affective disorder. A description of the syndrome and preliminary findings with
light therapy. Arch Gen Psychiatry. 41(1), 72-80.
Sit, D., McGowan, J., Wiltrout, C., Diler, R.S., Dills, J.J., Luther, J., et al., 2017. Adjunctive Bright
Light Therapy for Bipolar Depression: A Randomized Double-Blind Placebo-Controlled
Trial. Am J Psychiatry. appi-ajp
Sohn, C.H., Lam, R.W., 2004. Treatment of seasonal affective disorder: unipolar versus bipolar
differences. Curr Psychiatry Rep. 6(6), 478-485.
Terman, M., Terman, J.S., 2005. Light therapy for seasonal and nonseasonal depression: efficacy,
protocol, safety, and side effects. CNS Spectr. 10(08), 647-663.
Thase, M.E., Macfadden, W., Weisler, R.H., Chang, W., Paulsson, B., Khan, A., et al., 2006.
Efficacy of quetiapine monotherapy in bipolar I and II depression: a double-blind, placebo-
controlled study (the BOLDER II study). J Clin Psychopharmacol 26(6), 600-609.
Tondo, L., Pompili, M., Forte, A., Baldessarini, R.J., 2016. Suicide attempts in bipolar disorders:
comprehensive review of 101 reports. Acta Psychiatr Scand. 133(3), 174-186.
Tseng, P.T., Chen, Y.W., Tu, K.Y., Chung, W., Wang, H.Y., Wu, C.K., et al., 2016. Light therapy
in the treatment of patients with bipolar depression: A meta-analytic study. Eur
Neuropsychopharm. 26(6), 1037-1047.
Tuunainen, A., Kripke, D.F., Endo, T., 2004. Light therapy for non‐ seasonal depression. The
Cochrane Library. 10.1002/14651858.CD004050.pub2.
Volz, H.P., Mackert, A., Stieglitz, R.D., 1991. Side-effects of phototherapy in nonseasonal
depressive disorder. Pharmacopsychiatry. 24(4): 141-143.
Williams, J.B.W., Link, M.J., Rosenthal, N.E., Terman, M., 1988. Structured Interview Guide for
the Hamilton Depression Rating Scale, Seasonal Affective Disorders Version (SIGH-SAD)
New York Psychiatric Institute, New York.
Williams, J.B.W., 1988. A structured interview guide for the Hamilton Depression Rating Scale.
Arch Gen Psychiat. 45(8): 742-747.
Wu, J.C., Kelsoe, J.R., Schachat, C., Bunney, B.G., DeModena, A., Golshan, S., et al., 2009. Rapid
and sustained antidepressant response with sleep deprivation and chronotherapy in bipolar
disorder. Biol Psychiatry. 66(3): 298-301.
Young, A.H., Calabrese, J.R., Gustafsson, U., Berk, M., McElroy, S.L., Thase, M.E., et al., 2013.
Quetiapine monotherapy in bipolar II depression: combined data from four large,
randomized studies. Int J Bipolar Disord. 1(1), 10.
Zimmerman, M., Martinez, J.H., Young, D., Chelminski, I., Dalrymple, K., 2013. Severity
classification on the Hamilton depression rating scale. J Affective Disord. 150(2), 384-388.
Table 1. Sociodemographic Characteristics

BLT Controls p
(n=16) (n=16)
Age (Mean (SD)) 42.1 (9.1) 37.1 (8.2) NS*

Sex
Female 10 (62.5%) 16 (100%) 0.018**
Marital Status
Single 4 (25%) 6 (37.5%)
Married 10 (62.5%) 8 (50%)
NS
Widowed 1 (6%) 1 (6%)
Divorced - 1 (6%)
Separate 1 (6%) -
Education Level
Primary School 2 (12.5%) 4 (25%)
Secondary School 1 (6%) 1 (6%)
High School 5 (31%) 5 (31%) NS
Precollege 1 (6%) -
University 5 (31%) 5 (31%)

Ph.D. 2 (12.5%) 1 (6%)

Unemployment 8 (50%) 5 (31%) NS
BLT: Bright Light Therapy. SD: Standard deviation
*: NOT statistically significant
**Fisher’s exact test
Table 2. Psychiatric History, Clinical Features and Family History
BLT Controls
p
(n=16) (n=16)
Polarity
I 10 (62.5%) 7 (44%) NS*
II 6 (37.5%) 9 (56%)
Mean age of Onset (SD) 25.81 (5.84) 22.62 (5.36) NS
First episode
Mania 3 (19%) 3 (19%)
NS
Hypomania 2 (12.5%) -
Depression 11 (69%) 13 (81%)
No. of depressive episodes (Mean (SD)) 10.81 (10.22) 8.31 (6.32) NS
No. of manic episodes (Mean (SD)) 6.56 (7.85) 8.13 (8.11) NS
Seasonality 4 (25%) 3 (19%) NS
Manic switch rates with antidepressants 9 (56%) 7 (44%) NS
Rapid cycling - 1 (6%) NS
Mixed symptoms 4 (25%) 4 (25%) NS
Atypical depressive symptoms 6 (37.5%) 5 (31%) NS
Melancholic symptoms 7 (44%) 6 (37.5%) NS
Suicide attempts
None 9 (56%) 5 (31%)
NS
Once 3 (19%) 6 (37.5%)
Multiple 4 (25%) 5 (31%)
Hospitalization 9 (56%) 10 (62.5%) NS
Postpartum depression 1 (6%) 3 (19%) NS
Previous history of ECT 3 (19%) 4 (25%) NS
Alcohol use disorder 1 (6%) 4 (25%) NS
Psychiatric comorbidity 8 (50%) 10 (62.5%) NS
Other comorbidity 7 (44%) 6 (37.5%) NS
Family history of BPD 8 (50%) 10 (62.5%) NS
Family history of psychiatric morbidity 11 (69%) 12 (75%) NS
Family history of suicide attempts 3 (19%) 5 (31%) NS
Family history of alcohol use disorder 9 (56%) 6 (37.5%) NS
NS*: NOT statistically significant, BLT: Bright Light Therapy, SD: Standard deviation, ECT: Electroconvulsive therapy, BPD: Bipolar disorder

Table 3. Pharmacological Treatments

BLT Controls p
(n=16) (n=16)
Lithium 9 (56%) 8 (50%) NS*
Lamotrigine 2 (12.5%) 5 (31%) NS
Valproic acid 3 (19%) 3 (19%) NS
 Carbamazepine - 2 (12.5%) NS
Antipsychotics 6 (37.5%) 9 (56%) NS
Antidepressants 6 (37.5%) 6 (37.5%) NS
Monotherapy 7 (43.5%) 5 (31%) NS
NS*: NOT statistically significant, NS*: NOT statistically significant, BLT: Bright Light Therapy

Table 4. Response and Remission Rates

BLT Controls p
(n=16) (n=16)

HAM-D decrease by 50% 11 (69%) 2 (12.5%) 0.001*

HAM-D <7 7 (44%) 1 (6%) 0.037**

MADRS decrease by 50% 13 (81%) 3 (19%) <0.0001***

MADRS <9 points 7 (44%) 2 (12.5%) 0.05 ****

*: Chi-square Pearson value: 10.4, df: 1
**: Chi-square Pearson value: 6.0, df: 1
***: Chi-square Pearson value: 12.5, df: 1
****: Fisher’s exact test
BLT: Bright Light Therapy, MADRS: Montgomery-Asberg Depression Rating Scale, HAM-D: Hamilton Depression Rating Scale

Table 5. Seasonal Pattern and Treatment Response and Remission Rates

Seasonality (-) Seasonality (+) p

n=12 n=4
50% decrease in HAM-D 9 (75%) 4 (100%) NS*

HAM < 7 points 3 (25%) 4 (100%) 0.02

50% decrease in MADRS 9 (75%) 4 (100%) NS

MADRS < 9 points 3 (25%) 4 (100%) 0.02

NS*: NOT statistically significant, MADRS: Montgomery-Asberg Depression Rating Scale, HAM-D: Hamilton Depression Rating Scale

Highlights
• BLT administered as an add-on therapy for BD is significantly superior to placebo
• BLT is not different from placebo with regards to side effects
• The antidepressant effects of BLT emerge in the first week
• Since BLT is fast-acting option, it may facilitate recovery in the acute phase of BD
• It may increase remission rates by improving treatment compliance through fast action