Bright Light Therapy for Bipolar Depression: a Review of 6 Studies

Psychiatry Journal Club

Bright light therapy for bipolar depression: A review of 6 studies

Sy Atezaz Saeed, MD, MS, and Radhika J. Kothadia, MD

epressive episodes are part of Research to investigate novel therapeu- DSM-5 criteria for bipolar II dis- tics for bipolar depression is a high priority. order, and are also often experi- Patients with bipolar disorder are suscep- D 1 enced by patients with bipolar I disorder. tible to environmental cues that alter circa- Depressive episodes predominate the clini- dian rhythms and trigger relapse. Recent cal course of bipolar disorder.2,3 Compared studies have suggested that bright light with manic and hypomanic episodes, bipo- therapy (BLT), an accepted treatment for Sy Atezaz Saeed, MD, MS lar depressive episodes have a stronger seasonal depression, also may be useful for Department Editor association with long-term morbidity, sui- treating nonseasonal depression.10 Patients cidal behavior, and impaired functioning.4,5 with bipolar depression frequently have Approximately 20% to 60% of patients with delayed sleep phase and atypical depres- Recent evidence bipolar disorder attempt suicide at least sive features (hypersomnia, hyperpha- once in their lifetime, and 4% to 19% gia, and lethargy), which predict response suggests BLT die by suicide. Compared with the general to light therapy.11 In this article, we review may help reduce population, the risk of death by suicide is 6 recent studies that evaluated the effi- depressive symptoms 10 to 30 times higher in patients with bipo- cacy and safety of BLT for treating bipolar lar disorder.6 depression (Table,12-17 page 41). Treatment of bipolar depression is less investigated than treatment of unipolar 1. Wang S, Zhang Z, Yao L, et al. Bright light depression or bipolar mania. The mainstays therapy in treatment of patients with bipolar of treatment for bipolar depression include disorder: a systematic review and meta-anal- mood stabilizers (eg, lithium, valproic ysis. PLoS ONE. 2020;15(5):e0232798. doi: acid, or lamotrigine), second-generation 10.1371/journal.pone.0232798 antipsychotics (eg, risperidone, quetiap- ine, lurasidone, or olanzapine), adjunctive In this meta-analysis, Wang et al12 examined antidepressants (eg, selective serotonin the role of BLT in treating bipolar depres- reuptake inhibitors or bupropion), and sion. They also explored variables of BLT, combinations of the above. While signifi- including duration, timing, color, and color cant progress has been made in the treat-

ment of mania, achieving remission for Dr. Saeed is Professor and Chair, Department of Psychiatry and Discuss this article at patients with bipolar depression remains a Behavioral Medicine, East Carolina University Brody School www.facebook.com/ challenge. Anti-manic medications reduce of Medicine, Greenville, North Carolina. Dr. Saeed is Current MDedgePsychiatry Psychiatry’s Department Editor, Psychiatry Journal Club. Dr. Kothadia depressive symptoms in only one-third of is a PGY-1 Psychiatry Resident, Prisma Health–Midlands, Columbia, patients.7 Antidepressant monotherapy South Carolina. can induce hypomania and rapid cycling.8 Disclosures The authors report no financial relationships with any companies Electroconvulsive therapy has also been whose products are mentioned in this article, or with manufacturers used for treatment-resistant bipolar depres- of competing products. Current Psychiatry 40 May 2021 sion, but is usually reserved as a last resort.9 doi: 10.12788/cp.0114 Psychiatry Journal Club

Table Bright light therapy for bipolar depression: 6 studies Study Design Outcomes Wang et al12 Review and meta-analysis of 5 RCTs and Compared with placebo, BLT 7 cohort studies evaluating BLT for 847 significantly decreased depression patients with bipolar depression severity (P < .05) Lam et al13 Review and meta-analysis of 7 RCTs Compared with control conditions, BLT evaluating BLT for 259 patients with was associated with a small to moderate bipolar depression but statistically significant improvement in clinician-rated depressive symptoms (P = .03) Hirakawa et al14 Review and meta-analysis of 4 RCTs BLT had a significant effect on response evaluating BLT for 190 patients with rates (P = .002) but not on remission bipolar depression rates (P = .34) Clinical Point D’Agostino Review of 6 studies evaluating triple All studies reported improvement in et al15 chronotherapy (total sleep deprivation, HAM-D scores at the end of treatment, One meta-analysis sleep phase advance, and BLT) for 190 with response rates ranging from 50% patients with MDD or bipolar depression to 84% found that BLT at Dallaspezia Meta-analysis of 11 studies (including Analyses of all 11 studies and of just various durations, and Benedetti16 5 RCTs) evaluating BLT for depressive the 5 RCTs found BLT had a positive symptoms in 195 patients significant effect on depressive timings, colors, and symptoms (both P < .001) color temperatures Takeshima Review and meta-analysis of 6 studies No significant differences were found et al17 evaluating BLT for manic and depressive between BLT and placebo for rate of reduced depression symptoms in 280 patients with bipolar remission from depressive episodes, severity disorder rate of manic switching, or depressive symptom scores BLT: bright light therapy; HAM-D: Hamilton Depression Rating Scale; MDD: major depressive disorder; RCTs: randomized controlled trials

temperature, and how these factors may for depressive symptoms, and efficacy of affect the severity of depressive symptoms. different light color/color temperatures for depressive symptoms. Study design • Two researchers conducted a system- Outcomes atic literature search on PubMed, Web of • As assessed by the Hamilton Depression Science, Embase, Cochrane Library, and Rating Scale (HAM-D); Inventory of Cumulative Index of Nursing and Allied Depressive Symptomatology, Clinician Health Literature (CINAHL), as well as 4 Rating; or the Structured Interview Guide for Chinese databases from inception to March the HAM-D, depression severity significantly 2020. Search terms included “phototherapy,” decreased (P < .05) with BLT intensity ≥5,000 “bright light therapy,” “bipolar disorder,” lux when compared with placebo. and “bipolar affective disorder.” • Subgroup analyses suggested that • Inclusion criteria called for random- BLT can improve depression severity with ized controlled trials (RCTs) or cohort or without adjuvant therapy. Duration of studies that used a clearly defined diagno- <10 hours and >10 hours with morning sis of bipolar depression. Five RCTs and 7 light vs morning plus evening light ther- cohort studies with a total of 847 partici- apy all produced a significant decrease in pants were included. depressive symptoms (P < .05). • The primary outcomes were depression • White light therapy also significantly Current Psychiatry severity, efficacy of duration/timing of BLT decreased depression severity (P < .05). Vol. 20, No. 5 41 continued Psychiatry Journal Club

Color temperatures >4,500K and <4,500K Data for remission obtained from 5 trials both significantly decreased depression showed no significant difference between severity (P < .05). BLT and control (P = .09). • Compared with control, BLT was not Conclusion associated with an increased risk of affec- • BLT (at various durations, timings, tive switches (P= .67). colors, and color temperatures) can reduce depression severity. Conclusion • This analysis only included studies • This study suggests a small to moder- that showed short-term improvements in ate but significant effect of BLT in reducing depressive symptoms, which brings into depressive symptoms. question the long-term utility of BLT. • Study limitations included inconsistent light parameters, short follow-up time, Clinical Point 2. Lam RW, Teng MY, Jung YE, et al. Light small sample sizes, and the possibility that A review of 7 RCTs therapy for patients with bipolar depression: control conditions had treatment effects (eg, systematic review and meta-analysis of ran- dim light as control vs no light). found a small to domized controlled trials. Can J Psychiatry. moderate but 2020;65(5):290-300. 3. Hirakawa H, Terao T, Muronaga M, et al. significant effect Adjunctive bright light therapy for treating of BLT in reducing Lam et al13 examined the role of BLT for bipolar depression: a systematic review and patients with bipolar depression in a sys- meta-analysis of randomized controlled tri- depressive tematic review and meta-analysis. als. Brain Behav. 2020;10(12):ee01876. doi. symptoms org/10.1002/brb3.1876 Study design • Investigators conducted a systematic Hirakawa et al14 assessed the role of adjunc- review of RCTs of BLT for patients with tive BLT for treating bipolar depression. bipolar depression. Articles were obtained Previous meta-analyses focused on case- from Web of Science, Embase, MEDLINE, control studies that assessed the effects PsycInfo, and Clinicaltrials.gov using the of BLT and sleep deprivation therapy on search terms “light therapy,” “photother- depressive symptoms, but this meta-analy apy,” “light treatment,” and “bipolar.” sis reviewed RCTs that did not include • Inclusion criteria required patients sleep deprivation therapy. diagnosed with bipolar disorder cur- rently experiencing a depressive episode, Study design a clinician-rated measure of depressive • Two authors searched Embase, symptomatology, a specific light interven- MEDLINE, Scopus, Cochrane Central tion, and a randomized trial design with a Register of Controlled Trials (CENTRAL), control. CINAHL, and Clinicaltrials.gov from • A total of 7 RCTs with 259 participants inception to September 2019 using the were reviewed. The primary outcome was terms “light therapy,” “phototherapy,” improvement in depressive symptoms and “bipolar disorder.” based on the 17-item HAM-D. • Inclusion criteria called for RCTs, participants age ≥18, a diagnosis of bipolar Outcomes disorder according to standard diagnostic • BLT was associated with a significant criteria, evaluation by a standardized scale improvement in clinician-rated depressive (HAM-D, Montgomery-Åsberg Depression symptoms (P = .03). Rating Scale [MADRS], Structured • Data for clinical response obtained Interview Guide for the Hamilton from 6 trials showed a significant differ- Depression Rating Scale with Atypical Current Psychiatry 42 May 2021 ence favoring BLT vs control (P = .024). Depression Supplement [SIGH-ADS]), Psychiatry Journal Club

and light therapy as the experimental depressive disorder (MDD) and bipolar group intervention. depression. • The main outcomes were response rate (defined as ≥50% reduction in depression Study design severity based on a standardized scale) and • Researchers conducted a systematic remission rate (defined as a reduction to 7 search on PubMed, Scopus, and Embase points on HAM-D, reduction to 9 points on from inception to December 2019 using the MADRS, and score <8 on SIGH-ADS). terms “depression,” “sleep deprivation,” • Four RCTs with a total of 190 par- “chronotherapy,” and related words. ticipants with bipolar depression were • The review included studies of all evaluated. execution modalities, sequences of inter- ventions, and types of control groups (eg, Outcomes active control vs placebo). The popula- • BLT had a significant effect on response tion included participants of any age with Clinical Point rate (P = .002). MDD or bipolar depression. In an analysis • There was no significant effect of BLT • Two authors independently screened on remission rates (P = .34). studies. Six articles published between of 4 RCTs with • No studies reported serious adverse 2009 and 2019 with a total of 190 patients 190 participants, effects. Minor effects included headache were included. BLT significantly (14.9% for BLT vs 12.5% for control), irrita- improved response bility (4.26% for BLT vs 2.08% for control), Outcomes rates, but not and sleep disturbance (2.13% for BLT vs • All studies reported improvement 2.08% for control). The manic switch rate in HAM-D scores at the end of treatment remission rates was 1.1% in BLT vs 1.2% in control. with triple chronotherapy, with response rates ranging from 50% to 84%. Conclusion • Most studies had a short follow-up • BLT is effective in reducing depressive period (up to 3 weeks). In these studies, symptoms in bipolar disorder, but does not response rates ranged from 58.3% to 61.5%. affect remission rates. One study that had a 7-week follow-up also • This meta-analysis was based on a reported a statistically significant response small number of RCTs, and light therapy rate in favor of triple chronotherapy. parameters were inconsistent across the • Remission rates, defined by different studies. Furthermore, most patients were cut-offs depending on which version of the also being treated with mood-stabilizing or HAM-D was used, were evaluated in 5 stud- antidepressant medications. ies. These rates ranged from 33.3% to 77%. • It is unclear if BLT is effective as mono- • Two studies that used the Columbia therapy, rather than as adjunctive therapy. Suicide Severity Rating Scale to assess the effect of triple chronotherapy on suicide risk 4. D’Agostino A, Ferrara P, Terzoni S, et al. reported a significant improvement in scores. Efficacy of triple chronotherapy in unipolar and bipolar depression: a systematic Conclusion review of available evidence. J Affect Disord. • Triple chronotherapy may be an effec- 2020;276:297-304. tive and safe adjunctive treatment for depression. Some studies suggest that it Triple chronotherapy is the combination also may play a role in remission from of total sleep deprivation, sleep phase depression and reducing suicide risk. advance, and BLT. D’Agostino et al15 reviewed all available evidence on the 5. Dallaspezia S, Benedetti F. Antidepressant efficacy of triple chronotherapy inter- light therapy for bipolar patients: a meta- Current Psychiatry ventions in treating symptoms of major analyses. J Affect Disord. 2020;274:943-948. Vol. 20, No. 5 43 continued Psychiatry Journal Club

In a meta-analysis, Dallaspezia and 6. Takeshima M, Utsumi T, Aoki Y, et al. Benedetti16 evaluated 11 studies to assess Efficacy and safety of bright light therapy for the role of BLT for treating depres- manic and depressive symptoms in patients sive symptoms in patients with bipolar with bipolar disorder: a systematic review disorder. and meta-analysis. Psychiatry Clin Neurosci. 2020;74(4):247-256. Study design • Researchers searched literature pub- Takeshima et al17 conducted a systematic lished on PubMed with the terms “mood review and meta-analysis to evaluate the disorder,” “depression,” and “light therapy.” efficacy and safety of BLT for manic and • Eleven studies with a total of 195 par- depressive symptoms in patients with ticipants were included. Five studies were bipolar disorder. They also evaluated if BLT RCTs. could prevent recurrent mood episodes in Clinical Point • The primary outcome was severity of patients with bipolar disorder. Triple chronotherapy, depression based on scores on the HAM-D, Beck Depression Inventory, or SIGH-ADS. Study design which includes BLT, Secondary outcomes were light inten- • Researchers searched for studies of may be effective sity (measured in lux) and duration of BLT for bipolar disorder in MEDLINE, for the treatment of treatment. CENTRAL, Embase, PsychInfo, and patients with MDD Clincialtrials.gov using the terms “bipolar or bipolar depression Outcomes disorder,” “phototherapy,” and “random- • Analysis of all 11 studies revealed a pos- ized controlled trial.” itive effect of BLT on depressive symptoms • Two groups of 2 authors indepen- (P < .001). dently screened titles and abstracts for the • Analysis of just the 5 RCTs found a sig- following inclusion criteria: RCTs, 80% of nificant effect of BLT on depressive symp- patients diagnosed clinically with bipolar toms (P < .001). disorder, any type of light therapy, and • The switch rate due to BLT was lower control groups that included sham treat- than rates for patients being treated with ment or no light. Three groups of 2 authors antidepressant monotherapy (15% to 40%) then evaluated the quality of the studies or placebo (4.2%). and risk of bias. • Duration of treatment influenced treat- • Six studies with a total of 280 partici- ment outcomes (P = .05); a longer dura- pants were included. tion resulted in the highest clinical effect. • Primary outcome measures included However, regardless of duration, BLT rates of remission from depressive or manic showed higher antidepressant effects than episodes, rates of relapse from euthymic placebo. states, and changes in score on depression or • Higher light intensity was also found to mania rating scales. show greater efficacy. Outcomes Conclusion • No significant differences were • BLT is an effective adjunctive treat- found between BLT and placebo for rates ment for bipolar depression. of remission from depressive episodes • Higher light intensity and longer (P = .42), rates of manic switching duration of BLT may result in greater anti- (P = .26), or depressive symptom scores depressant effects, although the optimum (P = .30). duration and intensity are unknown. • Sensitivity analysis for 3 studies with • A significant limitation of this study low overall indirectness revealed that BLT was that the studies reviewed had high did have a significant antidepressant effect Current Psychiatry 44 May 2021 heterogeneity, and only a few were RCTs. (P = .006). continued on page 46 Psychiatry Journal Club

continued from page 44

• The most commonly reported adverse 6. Dome P, Rihmer Z, Gonda X. Suicide risk in bipolar disorder: effects of BLT were headache (4.7%) and a brief review. Medicina (Kaunas). 2019;55(8):403. 7. Sachs GS, Nierenberg AA, Calabrese JR, et al. Effectiveness of sleep disturbance (1.4%). adjunctive antidepressant treatment for bipolar depression. N Engl J Med. 2007;356(17):1711-1722. 8. Post RM, Altshuler LL, Leverich GS, et al. Mood switch in Conclusion bipolar depression: comparison of adjunctive venlafaxine, • This meta-analysis suggests that BLT bupropion, and sertraline. Br J Psychiatry. 2006;189:124-131. 9. Shah N, Grover S, Rao GP. Clinical practice guidelines does not have a significant antidepres- for management of bipolar disorder. Indian J Psychiatry. sant effect. However, a sensitivity analysis 2017;59(Suppl 1):S51-S66. of studies with low overall indirectness 10. Penders TM, Stanciu CN, Schoemann AM, et al. Bright light therapy as augmentation of pharmacotherapy for treatment showed that BLT does have a significant of depression: a systematic review and meta-analysis. Prim Care Companion CNS Disord. 2016;18(5). doi: 10.4088/ antidepressant effect. PCC.15r01906. • This review was based on a small num- 11. Terman M, Amira L, Terman JS, et al. Predictors of response and nonresponse to light treatment for winter depression. ber of RCTs that had inconsistent placebos Am J Psychiatry. 1996;153(11):1423-1429. Clinical Point (dim light, negative ion, no light, etc.) and 12. Wang S, Zhang Z, Yao L, et al. Bright light therapy in varying parameters of BLT (light intensity, treatment of patients with bipolar disorder: a systematic A review of 6 studies review and meta-analysis. PLoS ONE. 2020;15(5):e0232798. exposure duration, color of light), which doi: 10.1371/journal.pone.0232798 found no significant may have contributed to the inconsistent 13. Lam RW, Teng MY, Jung YE, et al. Light therapy for patients with bipolar depression: systematic review and meta- differences between results. analysis of randomized controlled trials. Can J Psychiatry. BLT and placebo on 2020;65(5):290-300. References 14. Hirakawa H, Terao T, Muronaga M, et al. Adjunctive bright light therapy for treating bipolar depression: a systematic several measures of 1. Diagnostic and statistical manual of mental disorders, 5th review and meta-analysis of randomized controlled ed. American Psychiatric Association; 2013. bipolar depression trials. Brain Behav. 2020;10(12):ee01876. doi.org/10.1002/ 2. Judd LL, Akiskal HS, Schettler PJ, et al. The long-term brb3.1876 natural history of the weekly symptomatic status of bipolar 15. D’Agostino A, Ferrara P, Terzoni S, et al. Efficacy of triple I disorder. Arch Gen Psychiatry. 2002;59(6):530-537. chronotherapy in unipolar and bipolar depression: a 3. Judd LL, Akiskal HS, Schettler PJ, et al. A prospective systematic review of available evidence. J Affect Disord. investigation of the natural history of the long-term 2020;276:297-304. weekly symptomatic status of bipolar II disorder. Arch Gen 16. Dallaspezia S, Benedetti F. Antidepressant light therapy Psychiatry. 2003;60(3):261-269. for bipolar patients: a meta-analyses. J Affect Disord. 2020; 4. Rihmer Z. S34.02 - Prediction and prevention of suicide in 274:943-948. bipolar disorders. European Psychiatry. 2008;23(S2):S45-S45. 17. Takeshima M, Utsumi T, Aoki Y, et al. Efficacy and safety of 5. Simon GE, Bauer MS, Ludman EJ, et al. Mood symptoms, bright light therapy for manic and depressive symptoms functional impairment, and disability in people with bipolar in patients with bipolar disorder: a systematic review disorder: specific effects of mania and depression. J Clin and meta-analysis. Psychiatry Clin Neurosci. 2020;74(4): Psychiatry. 2007;68(8):1237-1245. 247-256.

Current Psychiatry 46 May 2021