GME Logo

User login

User menu

GME Research Review

GME Research Review is a monthly newsletter where internationally recognized experts select, summarize, and provide a clinical commentary on the latest published research in psychiatry.  Each summary has been derived from the relevant article’s abstract and the clinical commentary has been provided by our expert.

ISSUE 14, May 2013 — Guest Commentator: Dan V. Iosifescu, MD, MSc, Mount Sinai School of Medicine, New York, NY


Efficacy and Mood Conversion Rate During Long-term Fluoxetine v. Lithium Monotherapy in Rapid- and Non-rapid-cycling Bipolar II Disorder

Amsterdam JD, Luo L, Shults J.
Br J Psychiatry. 2013;202:301-316.

Objectives: Multiple studies suggest an association between rapid cycling and antidepressant use in bipolar disorder.  To test the hypothesis that antidepressant monotherapy is associated with poor outcomes, this study reports an exploratory analysis of the safety and efficacy of fluoxetine versus lithium monotherapy in individuals with rapid- and non-rapid-cycling bipolar II disorder.


  • Participants included 167 adult patients meeting DSM-IV criteria for bipolar II disorder (by SCID-I/P): 89 women (36.9±12.7 years old) and 78 men (37.9±12.9 years old).  All subjects had recovered from a major depressive episode and experienced minimal current depressive symptoms (Hamilton Depression Rating Scale [HRSD-17] score ≤8). Estimates of the number of prior depressive and hypomanic episodes (defined by DSM-IV criteria) and subsyndromal hypomanic episodes, were obtained.  The participants' condition was classified as rapid cycling if they had an average of ≥4 affective episodes per year during the course of their illness.
  • The study is a randomized, double-blind, placebo-controlled comparison of fluoxetine versus lithium monotherapy in patients initially stabilized on fluoxetine monotherapy. Initial fluoxetine monotherapy 80 mg daily was administered for up to 12 weeks to 37 participants with rapid- and 111 with non-rapid-cycling bipolar disorder. Twenty-five subjects (67.6%) with rapid- and 58 (52.3%) with non-rapid-cycling bipolar disorder recovered with fluoxetine (final HRSD score ≤8) were randomized to long-term monotherapy with either fluoxetine 10–40 mg daily, lithium 300–1200 mg daily (with a serum level of 0.5–1.5 mmol/l), or placebo for 50 weeks.
  • The HRSD, Young Mania Rating Scale (YMRS) and mood conversion measures were obtained at baseline (i.e. randomization) and during double-blind therapy.  The primary outcome measure was the proportion of participants with rapid- v. non-rapid-cycling bipolar disorder with relapse or recurrence of a major depressive episode. Secondary outcomes included the risk of depressive relapse, change over time in YMRS scores (in patients experiencing change in YMRS scores) and frequency of syndromal and subsyndromal mood conversion episodes.



Rapid Cycling

Non-rapid Cycling

P value

Depressive Relapse

9 (36%)

29 (51.8%)


Relapse with:  

   - Fluoxetine
   - Lithium
   - Placebo







Odds of depressive relapse

No significant difference; OR=0.6, 95%, CI 0.2-1.8





No significant difference




YMRS change

No significant difference

Conclusions: In this study of rapid- and non-rapid-cycling subjects with bipolar disorder type II, the rates of depressive relapse and syndromal and subsyndromal hypomania were similar during long-term lithium and fluoxetine monotherapy and placebo.

Clinical Commentary
Most current treatment guidelines advise against the use of antidepressant monotherapy (without concomitant therapy with a mood stabilizer) in bipolar disorder, especially for patients with rapid-cycling bipolar disorder. The current report appears to challenge this expert consensus: the authors report that patients with a history of rapid-cycling bipolar II disorder show no increased risk of affective switches during 50 weeks of fluoxetine monotherapy (compared to lithium monotherapy). Before deciding to discard the current treatment guidelines in bipolar disorder, it is useful to note several details. First, the classification of “rapid cycling” in this study involved an average of 4 or more syndromal or subsyndromal episodes per year during the course of their illness (not in the past year, as per DSM-IV-TR criteria). Some of the subjects designated as rapid cycling in this study may have had fewer episodes in the last year compared to the non-rapid cycling subjects. In the rapid-cycling group the most recent depressive episode was 13.3 months, which is not consistent with current rapid cycling. Second, all subjects randomized to lithium or fluoxetine had previously tolerated a 12-week course of high dose (80 mg) fluoxetine monotherapy.  Since only subjects stable at the end of the initial treatment entered the randomized phase of the study, we can conclude these subjects represent a subgroup of patients pre-selected for their ability to not experience affective shifts during fluoxetine monotherapy. On balance, the current study importantly highlights the existence of a subgroup of subjects with bipolar disorder type II which may tolerate and experience good outcomes on long-term fluoxetine monotherapy. This usefully challenges the dogma of always avoiding antidepressants in bipolar subjects. However, the specific design of the study prevents us from generalizing these results to the entire population of subjects with bipolar II disorder (and especially those bipolar subjects with recent rapid-cycling).



The Sertraline vs Electrical Current Therapy for Treating Depression Clinical Study: Results From a Factorial, Randomized, Controlled Trial

Brunoni AR, Valiengo L, Baccaro A, et al
JAMA Psychiatry. 2013;70(4):383-391. 

The aim of this study was to assess the efficacy and safety of transcranial direct current stimulation (tDCS) compared to a standard SSRI antidepressant (sertraline) in subjects with major depressive disorder (MDD). 


  • Participants were 120 subjects meeting DSM-IV criteria for nonpsychotic, unipolar MDD (diagnosed with the MINI structured interview).  All subjects were antidepressant-free and had moderate to severe depressive symptoms (HRSD-17 score ≥ 18). Most subjects had low level of treatment resistance.
  • Participants were randomized using a 2×2 factorial design to sertraline/placebo and active/sham tDCS, constituting 4 groups: 1) sham tDCS and placebo (sham+placebo); 2) sham tDCS and sertraline (sertraline only); 3) active tDCS and placebo (tDCS only); and 4) active tDCS and sertraline (combined treatment).  
  • Active and sham tDCS were administered as twelve 30-minute tDCS sessions (10 consecutive tDCS sessions, Monday to Friday, in the first 2 weeks, then 2 additional sessions at 2 and 4 weeks).
  • The tDCS montage involves 2 electrodes places over the forehead: the anode over the F3 and the cathode over the F4 areas (according to the International 10-20 EEG electrode positioning system). These electrode positions correspond to the left and right dorso-lateral prefrontal cortex (DLPFC). 
  • For active tDCS a direct current of 2 mA was applied for 30 minutes at each session. 
  • Sham tDCS used the same montage but the device was turned off after 1 minute of stimulation (mimicking a mild local scratching sensation to preserve blinding).
  • The pharmacological intervention was a fixed dose of sertraline 50 mg/d or placebo.
  • The primary outcome measure was the change in Montgomery-Asberg Depression Rating Scale (MADRS) score from baseline to 6 weeks (end point).


Treatment Group

Mean Difference in MADRS Scores Compared to Sertraline + tDCS1

CI; P Value

Sertraline 2

8.5 points

95% CI, 2.96 to 14.03; P=.002


5.9 points

95% CI, 0.36 to 11.43; P=.03

Placebo+sham tDCS

11.5 points

95% CI, 6.03 to 17.10; P>.001

            1At 6 weeks, the combination of tDCS and sertraline had superior efficacy compared to either treatment alone.
            2 tDCS and sertraline used alone were associated with comparable efficacy (mean difference=2.6 points; 95% CI, -2.90 to 8.13; P=.35.
            3 Active tDCS alone (but not sertraline alone) was superior to placebo+sham tDCS.

      Adverse effects include skin redness on the scalp in active tDCS (P = .03). There were 7 episodes of treatment-emergent mania or hypomania, 5 occurring in the combined treatment group. The integrity of the blinding was problematic: participants correctly guessed both sertraline and tDCS use (although the authors stress that most participants were “only moderately confident in their choices”).

Conclusions: In this group of MDD subjects with low levels of treatment resistance, treatment with tDCS or low-dose sertraline was associated with similar rates of improvement and adverse events. The combination of tDCS and sertraline had superior efficacy compared to either treatment alone or compared to sham. 

Clinical Commentary 
Over the last decade a series of novel devices using different forms of energy for brain stimulation have been proposed as treatment strategies for mood disorders. Transcranial direct current stimulation is an old technology; its efficacy as an antidepressant appears questionable based on poorly designed older studies. This is the first well-designed study testing this treatment, but in a population with very low levels of treatment resistance and in comparison to a low dose of a standard antidepressant. The blinding appears imperfect, which raises additional questions on the full validity of the results. The tDCS treatment is easy to deliver and may be associated with standard antidepressants for additional efficacy. On balance, tDCS appears a valid treatment for milder, non treatment-resistant forms of depression.


Practice-Based Versus Telemedicine-based Collaborative Care for Depression in Rural Federally Qualified Health Centers: A Pragmatic Randomized Comparative Effectiveness Trial

Fortney JC, Pyne JM, Mouden SB, et al
Am J Psychiatry. 2013;170(4):414-425.

Practice-based collaborative care is a complex, evidence-based practice; it is however difficult to implement in smaller primary care practices that lack on-site mental health staff. Telemedicine-based collaborative care may represent a solution, as it allows integration of mental health providers into remote primary care settings. The objective of this multisite randomized comparative effectiveness trial was to evaluate the outcomes of patients assigned to practice-based and telemedicine-based collaborative care. 


  • From 2007 to 2009, 19,285 patients at five federally qualified health centers serving medically underserved populations were screened for depression with the Patient Health Questionnaire (PHQ-9).
  • None of the clinics had mental health staff.
  • 15% (N=2,863) of the patients screened positive (PHQ-9 ≥10). 
  • 364 patients who screened positive were enrolled in the study and followed for 18 months, randomly assigned to one of the two interventions. Most of the subjects in the study were in rural locations, unemployed, and uninsured. 
  • Those assigned to practice-based collaborative care received evidence-based care from an on-site primary care provider and a nurse care manager (nurses with no prior mental health experience who received a 1-day training in depression care management). No mental health provider supervised nurse care managers.
  • Patients assigned to telemedicine-based collaborative care received evidence-based care from an on-site primary care provider and an off-site team: a nurse care manager and a pharmacist by telephone, and a psychologist and a psychiatrist via videoconferencing. In the collaborative care group patients received medication management and CBT (if needed) via videoconference. 
  • The primary clinical outcome measures were treatment response, remission, and change in depression severity.  Follow-up was performed with blinded telephone interviews at 6, 12, and 18 months.



Telemedicine-based Group


Better outcomes for both response and remission

Greater reductions in depression severity over time (Per Hopkins Symptom Checklist)

Higher patient satisfaction with treatment

Note: Improvements in outcomes appears to be attributable to higher fidelity to the collaborative care evidence base
in the telemedicine-based group compared to the practice-based group. 

Conclusions: In the context of small remote primary care practices with no mental health staff, better trained/supervised nurse care managers obtain superior outcomes via telemedicine compared to minimally trained local part-time care managers with no additional supervision from mental health professionals and pharmacists.

Clinical Commentary
The effectiveness of collaborative care programs in the treatment of depression in primary care has been supported by several previous studies. However, in most previous studies collaborative care involved local care managers and mental health specialists. The current study has important implication for the organization of collaborative care in remote settings lacking local mental health providers. For such settings, the current study suggests that centralized care management using full-time managers supervised by specialists yields superior outcomes, even if such care is delivered via teleconference. The benefit of local relationships between patients and providers appears to be surpassed by the higher quality of the care and monitoring delivered as part of the centralized program. 


Contribution of Common Genetic Variants to Antidepressant Response

Tansey KE, Guipponi M, Hu X, et al
Biol Psychiatry. 2013;73(7):679-682.

The field of pharmacogenetics assumes that drug responses are heritable traits.  However, whether an individual’s response to a given drug is heritable has not, in general, been established.  The aim of this study was to estimate the heritability of response to antidepressants, by combining genetic data and clinical outcome data after treatment with SSRI or SNRI antidepressants from two large studies in MDD.


  • The authors have combined two large samples of adults with MDD, with prospectively recorded outcome of antidepressant treatment and genome-wide genotyping were used in this analysis, the Novel Methods Leading to New Medications in Depression and Schizophrenia (NEWMEDS) and the Sequenced Treatment Alternatives to Relieve Depression (STAR*D). 
  • NEWMEDS includes 2146 individuals with available blood DNA, treated with SSRI (escitalopram, citalopram, paroxetine, sertraline, fluoxetine) or NRI (nortriptyline, reboxetine) antidepressants for 6-12 weeks.  STAR*D included 1,948 MDD subjects with available blood DNA treated for up to 14 weeks with the SSRI citalopram.
  • Unrelated Caucasian individuals (n=1,790 from NEWMEDS and n=1,107 from STAR*D) who passed stringent genotype quality control were included in the analyses. 
  • Antidepressant response was defined as a continuous variable [ie, change from baseline in depression severity on the primary depression rating scale (MADRS, HAMD)], adjusted for age, sex, and recruiting center/study.
  • Heritability was estimated with a novel method (Genome-wide Complex Trait Analysis, GCTA), which uses genome-wide single-nucleotide polymorphisms (SNPs) to estimate directly the total additive genetic variance. GTCA approach compares how similar pairs of individuals are both genetically (lining up many SNPs across the genome) and phenotypically (in terms of response to antidepressants) to estimate heritability. If a trait is heritable, then individuals who are genetically more similar should also be phenotypically more similar. The analyses were restricted to 71,297 autosomal SNPs that were common across the two genotyping platforms used in the two studies. These SNPs were evenly spread out across the genome and covered all autosomes.


  • The additive effects of common genetic polymorphisms across the human genome explain approximately 42.0% (SE=.180, P=.009) of individual variation in response to antidepressants.
  • In 2,273 subjects treated with SSRI antidepressants, common SNPs explained .428 (SE=.230) of variance in response to serotonergic antidepressants.
  • The similar estimates for all antidepressants and for SSRI antidepressants suggest that most of the genetic contribution is common between SSRI and NRI antidepressants.
  • This estimate of variance of antidepressant response explained by common genetic variation, together with the lack of any positive genome-wide association studies suggests that antidepressant response is a polygenic trait with a contribution from multiple common genetic variants across the genome, similar to other complex traits, such as schizophrenia.

  These results suggest that response to antidepressants is a complex trait with substantial contribution from a large number of common genetic variants of small effect. 

Clinical Commentary
Multiple previous pharmacogenetics studies have tried to detect genetic markers of antidepressant response, with very limited results so far. The current report suggests that a significant (42%) proportion of the variance in antidepressant response may be related to genetic factors.  However, the data implicate a high number of common genetic variants scattered across the genome, none of which have very large effects, but cumulatively contribute to a substantial proportion of variation.  This makes more unlikely the possibility of a simple genetic test to predict antidepressant response in clinical populations. Moreover, these data do not imply that one could explain 42% of the variability in antidepressant response, or have 42% accuracy, in a very complex genetic test taking into account all the SNPs highlighted by on these data; the accuracy of such genetic tests may be significantly lower.  This report is definitely not the final word on this subject and future studies, with even larger samples, may focus our understanding of genetic predisposition for antidepressant response on a narrower group of genes.  However this is a sobering result, which highlights the difficulty of finding a genetic test to accurately predict antidepressant response in clinical populations.


Lithium, Gray Matter, and Magnetic Resonance Imaging Signal
Cousins DA, Aribisala B, Nicol Ferrier I, et al
Biol Psychiatry. 2013;73(7):652-657.

Objectives: Multiple previous magnetic resonance imaging (MRI) studies have reported that lithium can increase the volume of gray matter or the thickness of cortical layers in the human brain.  This finding has been interpreted as the result of neurotrophic or neuroprotective effects of the drug (which have also been detected in animal models).  The current study has tested an alternate hypothesis, that lithium may influence directly the intensity of the MRI signal and that reported changes in gray matter volumes may be artifacts resulting from the altered image contrasts. This hypothesis was tested by measuring gray matter volume change after a short course of lithium with two neuroimaging methods: voxel-based morphometry and Structural Image Evaluation using Normalization of Atrophy (SIENA). 


  • Participants were 32 healthy young men. Subjects with psychiatric or medical conditions were excluded. In study 1, eight subjects received MRI before and after 11 days of open-label lithium to detect the effects of lithium on proton relaxation times and on the brain MRI volumetric analysis. In study 2, MRIs were performed at baseline and after a course of high-dose lithium (n=9), low-dose lithium (n=9), or placebo (n=6). In all instances, anatomical and quantitative scan data were acquired before medication administration and after 11 days of lithium.
  • All scans were performed on 3 Tesla Achieva whole-body scanner (Philips Medical Systems) with an 8-channel SENSE head coil. The protocol comprised: 1) high-resolution three-dimensional (3D) T1-weighted anatomical (repetition time [TR] = 9.6 msec; echo time [TE] = 4.6 msec; flip angle = 8°; field of view = 240×240 mm; contiguous); 2) fast quantitative T1 measurement with a custom inversion recovery prepared echo-planar imaging sequence (TR = 15 sec; TE = 24 msec; inversion time = .25–2.5 sec in 12 uniform steps; matrix 128×128, 72 slices, isotropic 2-mm resolution); and 3) low-resolution Bo field-map with a dual echo 3D gradient-echo (TR = 27 msec, TE = 2.6, 6.1 msec).
  • Image analysis was performed on a Linux platform using several software packages (MATLAB; SPM8; MRICRO; FSL).  To perform voxel-based morphometry (VBM), T1-weighted anatomical images were normalized and segmented with a three-compartment model in SPM8 (gray matter/white matter/CSF).  Global volumes for gray matter, white matter, and cerebrospinal fluid (CSF) were calculated by summing the voxel intensity values over each segmented image. 
  • The fully automated Structural Image Evaluation using Normalization of Atrophy (SIENA) software was also used to measure brain volume change between baseline and endpoint scans. SIENA estimates volume change by finding the brain/non-brain edge points on images and then measuring the perpendicular edge displacement between baseline and endpoint scans for each subject. The mean edge displacement is converted into a global estimate of percentage brain volume change.
  • Quantitative T1 relaxometry analysis was also performed. 




Lithium (11 days)


Placebo (11 days)


P Value

Grey Matter Volume:

Voxel–based Morphometry


No change


Grey Matter Volume:


No change

No change


Grey Matter
T1 Relaxation


No change


Conclusion: Magnetic resonance images of the brain differ before and after a short course of lithium, but this difference might derive from a change in the characteristics of the MRI signal rather than an actual increase in gray matter volume.

Clinical Commentary
Lithium remains the gold standard in the treatment of bipolar disorder. However, its clinical effects occur over longer periods of treatment.  Multiple studies have highlighted neurotrophic or neuroprotective effects of lithium in animals, while high-definition MRI studies have reported significant increases in gray matter volume even after relatively brief periods of lithium treatment (several weeks). The current report raises the likelihood that such significant increases in gray matter volume after short periods of lithium treatment probably represent an artifact. This should not change our enthusiasm for using lithium for bipolar subjects, but it puts in perspective simplistic biological “explanations” of lithium efficacy.


The Association Between Low Vitamin D and Depressive Disorders

Milaneschi Y, Hoogendijk W, Lips P, et al.
Molecular Psychiatry advance online publication 9 April 2013; doi: 10.1038/mp.2013.36

Objective:  It has been hypothesized that hypovitaminosis D is associated with depression but epidemiological evidence supporting this assertion is limited and contradictory. This study investigated the association between depressive disorders and related clinical characteristics with blood concentrations of 25-hydroxyvitamin D [25(OH)D] and parathyroid hormone (PTH) in a large cohort of depressed patients, subjects with remitted depression and healthy controls. Associations between serum 25(OH)D and specific clinical features and the course of illness in  currently depressed subjects were also tested.


  • Participants were part of the Netherlands Study of Depression and Anxiety (NESDA). Between 2004–2007, 2981 participants aged 18–65 years (including subjects with a current or past depressive and/or anxiety disorder and healthy controls) were recruited from the community (19%), general practice (54%), and secondary mental health care (27%).  After 2 years, a face-to-face follow-up assessment was conducted (response rate= 87.1%). At baseline, participants provided blood samples and underwent medical examinations and psychiatric interviews. 
  • DSM-IV diagnoses of depressive and anxiety disorders were ascertained with a structured interview (Composite Interview Diagnostic Instrument, CIDI).  
  • Vitamin D analyses were performed in 2,386 participants with a current (ie, within the past 6 months; N=1,158) or remitted (ie, lifetime, but not current; N=815) depressive disorder and healthy controls (N=511). 
  • Analyses on the course of depressive disorders were based on 902 participants depressed at baseline who participated in the 2-year follow-up.
  • Vitamin D status was measured by assessing circulating levels of 25(OH)D, which is the combined product of cutaneous synthesis from solar exposure and dietary sources. Fasting blood samples were obtained in the morning around 8 am and kept frozen at −80 °C and never thawed before analysis. Serum 25(OH)D was measured using isotope dilution—online solid phase extraction liquid chromatography–tandem mass spectrometry (ID-XLC-MS/MS).  
  • Serum intact PTH was determined using an immunometric assay (Abbott Laboratories, Abbott Park, Illinois).
  • Clinical characteristics measured for subjects with current depressive disorders:
    • Severity of depression (28-item self-report IDS)
    • Depressive symptoms duration during the past 4 years (Life Chart Interview)
    • Age of onset and presence of comorbid anxiety disorders (from the CIDI) and use of medications including antidepressants.
    • Course of depressive disorders was determined using two different outcome measures: (1) the presence (yes/no) of a DSM-IV depressive diagnosis (past 6-months) at 2-year follow-up; (2) duration of depressive symptoms over 2 years (derived from the Life Chart Interview).
    • Duration of symptoms was calculated as the percentage of time during follow-up with symptoms of at least mild severity.   
  • Putative confounders, selected a priori on the basis of previously reported associations with vitamin D and depression, included:
    • Sociodemographic variables (age, sex, and years of education)
    • Smoking status and alcohol use
    • Body mass index (BMI) and physical activity
    • The number of chronic diseases for which subjects received treatment (eg, cardiovascular disorders, diabetes, lung disease, arthritis, cancer, ulcer, intestinal problem, liver disease, epilepsy, and thyroid gland disease)
    • Use of vitamin D supplements
    • The number of sunlight hours in the 10 weeks preceding blood drawing
    • The degree of urbanization (number of inhabitants per km2).


 Depressed Patients

P value, Statistics

Lower 25 (OH)D levels compared with controls

P=0.001, Cohen’s d=0.21

Lowest 25 (OH)D levels in those with most severe depression

P=0.001, Cohen’s d=0.44

Inverse relationship between vitamin D status and:
    -     Symptoms severity

-          Risk of depressive disorder at 2-year follow-up



P=0.003, β=−0.19, s.e.=0.07,

P=0.03, Relative risk=0.90, 95% CI=0.82–0.99

Conclusions: This large cohort study indicates that low levels of 25(OH)D were associated to the presence and severity of depressive disorder, suggesting that hypovitaminosis D may represent an underlying biological vulnerability for depression.

Clinical Commentary
Low levels of vitamin D are relatively common in the general population, especially at northern latitudes and during winter months. This large study has the merit of carefully assessing the association between vitamin D levels and depression while taking into account a host of possible confounders, including sociodemographic factors, sunlight, urbanization, lifestyle and health. As such, the association reported appears credible: low serum vitamin D levels were found in subjects with both current and remitted depression, and among currently depressed patients lower vitamin D levels were associated with a less favorable course. This study lends credibility to previously reported results on the efficacy of vitamin D supplementation in the treatment of depression. Of note, this cross sectional study does not inform on the directionality of this association. It is possible that low vitamin D levels may trigger depression, but it is also possible that changes in lifestyle associated with depression (obesity, staying indoors and avoiding sunlight, etc.) may result in lower vitamin D levels. However, given the inexpensive and generally benign nature of vitamin D supplementation, these data justify future efforts to study the role of vitamin D supplementation as part of preventive or treatment interventions in major depression. 



To contact GME, email us at [email protected]

GME does not provide medical advice. The website and articles are intended for informational purposes only. They are not a substitute for professional medical advice, diagnosis or treatment. Never ignore professional medical advice in seeking treatment because of something you have read on the GME Website. If you think you may have a medical emergency, immediately call your doctor or dial 911.

Log In

Join GME For Free

Become a GME subscriber and gain full access to our extensive library of 700+ psychiatric medical education videos, free CME webcasts, latest research updates, and more. To sample our content, watch the featured videos below.
Join Today!