Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jul 10.
Published in final edited form as: Compr Psychiatry. 2012 Aug 15;54(1):1–6. doi: 10.1016/j.comppsych.2012.06.006

The diagnosis of depression: current and emerging methods

Katie M Smith a,*, Perry F Renshaw b, John Bilello a
PMCID: PMC5502713  NIHMSID: NIHMS873499  PMID: 22901834

Abstract

Depression is one of the leading causes of disability in adolescents and adults, particularly starting from age 15 years and older. Diagnosis of depression has traditionally been made based on clinical criteria, including patient current symptoms and history. This process is widely used but relies on subjective interpretation. To standardize both the data obtained and data interpretation, various interview-based instruments and noninterview methods exist for screening and testing for depression in various clinical settings.

This article evaluates the technical basis for and clinical performance of these various instruments and methods to diagnosis depression in clinical settings. Traditional tools include physician-administered or patient self-administered interview tools that have reasonable clinical accuracy depending on the threshold score and may lead to a full diagnostic evaluation for high-risk patients. In addition, older laboratory methods such as the dexamethasone test have contributed to the diagnosis of depression over a long period. Newer diagnostic methods such as genomics, proteomics, and metabolomics are technically sophisticated and objective and are beginning to emerge in psychiatry. Although promising, further evaluation of these methods is needed to fully demonstrate their clinical value and accuracy.

1. Introduction

Over 20 million people in the United States have a mood disorder, and only about 50% have been diagnosed and are being treated for this malady [1,2]. Major depressive disorder (MDD) affects approximately 14.8 million American adults or about 6.7% of the US population aged 18 years and older in a given year [1]. Rates of depression among children and adolescents approach that in adults with reports indicating prevalences of 3% and 6% [3], respectively, and the lifetime prevalence of MDD among adolescents may be as high as 20% [4–6]. MDD is the leading cause of disability in the United States for individuals between 15 and 44 years old [7]. It is a significant disorder in all ages requiring significant health care dollars to treat effectively.

Diagnosis in individuals young and old is difficult given that depressive symptoms can mimic other disorders and often, coexisting conditions can confound an accurate diagnosis. Many individuals are diagnosed based on treatment response to one or several antidepressant medications. Various clinical interview-based instruments as well as laboratory methods exist for depression screening, and several new approaches are in development, each demonstrating varying degrees of clinical development and robustness.

To understand the role of these very different diagnostic approaches to depression in the clinic, it is important to evaluate their methodology and clinical accuracy. Both nonlaboratory and laboratory methods have found use for depression diagnosis, yet these methods are not useful for assessing depression severity in previously diagnosed depression.

2. Nonlaboratory methods to diagnose MDD

The long-standing approach to depression diagnosis is often subject to great variation in the methods for information gathering and processing. For clinical study conduct, a structured or semistructured interview method (SDI) is standard; however, in routine practice, clinical diagnosis is used. A recent Psychiatric Times article described a meta-analysis of 38 studies, and nearly 16 000 patients showing these 2 approaches often result in different rather than comparable diagnoses [8]. A K statistic for each study ranged from 0.6 to 0.8 (interpreted as “acceptable” to “good”), and the statistic from the meta-analysis across all diagnoses was 0.27 [9]. This demonstrates a “gap” in the real world vs study-based methods to achieve accurate diagnoses.

Another study reported clinical diagnosis for MDD relative to SDI to be poor [10]. The implication of both reports is that, for many patients, a positive diagnosis of a particular psychiatric disorder would not be confirmed for most (>50%) of the subjects by an SDI method. Regardless of which approach is truly “more accurate,” many patients are misdiagnosed, thereby confounding a decision to treat or not to treat and in treatment selection for a given patient. Additional impact is a lower probability of successful outcomes for many patients.

3. Laboratory methods to diagnose MDD

The dexamethasone suppression test (DST) was developed to differentiate various types of Cushing syndrome and other conditions mediated by hypercorticism and also has been used for depression diagnosis [11,12]. The DST assesses the negative feedback of dexamethasone, a cortisol-like synthetic hormone on pituitary corticotropin release. Dexamethasone administration, typically at a low dose (1–2 mg), should reduce corticotropin levels and lead to decreased cortisol levels in healthy individuals, but in many depressed patients, cortisol levels remain elevated [11].

Despite being a pioneering laboratory method, the DST has proven inconvenient for patients and lacked good clinical performance for depression assessment [13]. The test may differentiate severe melancholic depression, mania, or acute psychosis from chronic psychosis (87% specificity) or dysthymia (77% specificity) [14]. Dexamethasone suppression test accuracy can be compromised by various factors. The Table summarizes these confounders and the clinical performance characteristics of DST along with the other methods discussed herein.

Table.

Comparison of methods outside primary care to diagnose MDD

Diagnostic method Clinical performance in MDD
Structured Diagnostic Interview
 SDI vs clinical evaluation κ = 0.45 [9]
 SDI vs PHQ Sensitivity, 40%; specificity, 87% [10]
DST
 DST vs clinical evaluation Sensitivity, 60%–70%; specificity, 70%–90% [11]
Confounders: patient use of barbiturates, anticonvulsants, or corticosteroids; use of alcohol; bioavailability of dexamethasone; acute “stress”; severe weight loss
Genomics
 Genomic signature vs CIDI (v. 2.1) Sensitivity, 87.5% (n = 42); 76.9% (n = 26)
Specificity, 61.5% (n = 42); 71.4% (n = 26) [16]
No reports on clinical performance or effect of medication or comorbidities
Proteomics No reports on clinical performance or effects of medication or comorbidities
Metabolomics P > .05 for depressed vs remitted and never depressed groups [23]
No reports on clinical performance or effect of medication or comorbidities
Biological pathway biomarker panel
 Biomarkers panel vs SDI Sensitivity, 95%, specificity, 87.5% [36,37]
No reports on effect of medication or comorbidities
Open in a new tab

4. Future directions

More advanced technology is emerging to assess depression, and although these newer techniques have not yet found full adoption in psychiatry, the practicing psychiatrist should be made aware given that similar methods have found use in other therapeutic areas. These are discussed below.

5. Genomic methods

With the sequencing of the human genome in the 1990s and development of new techniques for rapid sequencing, genomic applications in many medical areas, including psychiatry, are emerging.

Measuring gene expression profiles in blood cells holds promise for identifying disease classifiers and risk markers in psychiatric disorders, and one recent study describes such a profile for MDD [15,16]. The approach involved lipopolysacchride stimulation of blood cells in patients with MDD and normal samples to generate a high signal–to–basal state genomic response that could be associated with MDD. In a primary cohort of MDD and healthy control subjects, a 7-gene molecular profile was identified that distinguished patients with MDD from healthy controls. These results were confirmed using an independent gene detection method (polymerase chain reaction) (P = .007) and validated using polymerase chain reaction in a replication cohort where separation of the 2 groups was also found to be statistically significant (P = .019). An estimate of the gene panel’s clinical performance showed good sensitivity but poor specificity [16].

6. Proteomic and metabolomic methods

Proteomic techniques facilitate the study of multiple protein and modified protein products of disease-relevant genes. This can lead to the study of differentially expressed protein profiles in diseased vs nondiseased populations. Numerous studies of brain tissue, blood, and cerebral spinal fluid in schizophrenia and bipolar patients have shown some common protein alterations in energy and phospholipid metabolism as well as Ca2+ homeostasis [17–19]. Few systematic studies have been done on either brain tissue or with biological fluids from depressed patients. Most work have been done in animal models with a focus on biomarker changes due to antidepressant treatment rather than to identify biomarkers to diagnose depression. Some studies show that antidepressant treatment may modulate neurogenesis-related markers such as brain-derived neurotrophic factor (BDNF) [20]. Many studies have been reported using high anxiety behavior (HAB) and low anxiety behavior (LAB) murine models. The HAB mice represent a comorbid depression-like behavior and are hyperanxious. Proteomic analysis identified glyoxalase-I as a protein marker down-regulated in various areas of the brain in 1 study and a modified enolase phosphatase isoform in HAB mice in another [21,22]. Elucidation of drug mechanisms and potential biomarkers using proteomic techniques has great promise, but further exploration is needed.

Proteomic analyses of brain tissue have led to the study of metabolites or metabolomics, which enables the separation and measurement of many small molecules comprising metabolic pathways. The goal, as with genomics and proteomics, is to identify “signatures” of metabolites differentially expressed in diseased vs nondiseased states to understand disease-driven biochemical changes. A recent metabolomic study in depressed elderly adults analyzed 800 metabolites and identified several fatty acids (stearate, palmitate, myristate, and oleate) along with glycerol and γ-aminobutyric acid as having statistically lower blood levels in currently depressed patients compared with never depressed controls. Mass spectrometry analysis of plasma metabolites yielded an expression pattern that was most similar between the remitted group and the never depressed control group, and both differed from the currently depressed group [23]. The study suggests that changes in lipid and neurotransmitter metabolism associated with depression may alter these metabolic pathways. These metabolites then might serve as a “signature” to identify depressed patients, pending additional study to validate these findings.

7. Systematic biological pathway analysis

Components of several biological pathways undergo changes in various psychiatric conditions or in response to antipsychiatric therapy [24,25]. Pathways such as the inflammatory, neurotrophic, metabolic, and hypothalamic-pituitary-adrenal (HPA) axis express and respond to many regulatory biochemicals that include steroids, neuropeptides, cytokines, and neurotransmitters [26–28]. These systematic interactions provide the molecular basis for integrated neuroendocrine-immune responses to homeostatic perturbations induced by stress, inflammation, or infection.

Many components of these pathways have been implicated in depression. Inflammatory markers have been studied in MDD as potential biomarkers [29]. Severe depressive illness has been associated with elevations of cytokines or their soluble receptors, including interleukin (IL)-2, soluble IL-2 receptors (sIL-2R), IL-1b, IL-1 receptor antagonist (IL-1Ra), IL-6, soluble IL-6 receptor (sIL-6R), and γ-interferon [26,30].

A case control study found that serum resistin (a cytokine and metabolic marker) levels correlate with symptoms of atypical depression and that levels of adiponectin were lower in depressed subjects than controls [31].

Reports of BDNF, a neurotrophic factor, suggest that serum levels are generally lower in depressed subjects compared with controls and that serum BDNF levels increase in response to antidepressant treatment. Brain-derived neurotrophic factor alone may, however, lack some specificity for depression diagnosis, and the mechanism of BDNF release is not clearly understood [32]. Serum BDNF may serve as a biomarker to discriminate between unipolar and bipolar depression [33].

Dysfunction in the HPA axis has been observed in the etiology of depression. Hypercortisolism in depression is described by elevated mean 24-hour serum cortisol concentrations and increased 24-hour urinary excretion of cortisol [34]. In endocrine testing, using the DST, serum cortisol, and adrenocorticotrophin concentrations are not suppressed in some 20% to 50% of patients [35].

Recently, a new blood test has been described for MDD diagnosis that is composed of representative components of the neurotrophic, metabolic, inflammatory, and HPA axis pathways [36,37]. The test consists of 9 biomarkers, including α-1 antitrypsin, apolipoprotein CIII, BDNF, cortisol, epidermal growth factor, myeloperoxidase, prolactin, resistin, and soluble tumor necrosis factor α type II, all measured by standard immunoassays. Thyroid stimulating hormone (TSH) is included and separately reported to assess hypo- or hyperthyroidism in the overall evaluation of depression. Patients with endogenous depression often have lower levels of basal serum TSH and lower TSH changes from baseline to peak [38]. Depression may be associated with subclinical hypothyroidism or mild thyroid failure [39].

These biomarkers were identified within a compiled list of 90 to 100 blood biomarkers (proteins, cytokines, hormones, etc) cited in the scientific literature or emerging research with a role in MDD or other psychiatric disorders (diagnosis, prognosis, etc). A hypothesis-driven program was initiated to demonstrate that some of these serum biomarkers could distinguish patients with MDD from healthy individuals provided that testing methods or reagents were commercially available for this purpose. The goal was to identify a subset of biomarkers that could quantitatively separate the patients with MDD from healthy individuals as a first step to a diagnostic MDD panel. Target biomarkers with comparable performance were further refined for the final panel based on their association with 1 of the 4 pathways described above. Finally, good clinical sensitivity and specificity (>85%) were key outcome measures to select the final 9 biomarkers capable of separating MDD from normal subjects.

Serum measurements of each of the 9 biomarkers were analyzed using a bilogistic algorithm to provide a final predictive diagnostic score. Scores ranged from 1 to 9 signifying an increasing likelihood of MDD.

8. Retrospective study

A retrospective case control study was conducted with a heterogeneous population of MDD patient samples to rigorously test the discriminatory capability of the candidate biomarkers to differentiate MDD from healthy subjects. A training set of 50 serum samples from patients with MDD and 20 healthy volunteers was obtained from a biobank (PrecisionMed, San Diego, CA) and were maintained frozen at −80°C until testing. In the MDD, 48 of 50 and 20 of 20 in the healthy subject groups were evaluable. In the MDD group, 2 of 50 had insufficient serum volume to test all target biomarkers, so they were not evaluable.

The MDD samples were from men; women were excluded to avoid confounding factors, resulting from pre-and postmenopausal hormonal changes. These patients with MDD met Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) diagnoses of 296.31, 296.32, 296.33, or 296.35 and had family histories of depression in a first-degree relative. Exclusion criteria eliminated anyone with a DSM-IV diagnosis of 296.2, assuming that a single depressive episode was not definitive for a diagnosis of depression, or anyone who had a history of head trauma, delusions, hallucinations, seizure disorder, or alcohol abuse. Blood draws were performed while the patients were in the depressed state. Evaluable men with MDD were of Northern European descent, with a mean age of 39 years (age interval, 22–50 years). Twenty were free of psychiatric drugs, and 30 were on various antidepressants either as monotherapy or combination therapy.

The healthy subjects were men and were age matched to the patients with MDD with a mean age of 36.3 years (age interval, 22–50 years). A preliminary screen by an experienced psychiatrist excluded those who had history of major neurologic, medical, or psychiatric disorders or a prior head injury.

The clinical performance of the MDD-specific biomarker panel showed a clinical sensitivity of 95% (confidence interval [CI], 73.1%–99.7%) and a clinical specificity of 87.5% (CI, 74.1%–94.8%).

9. Validation study

A prospective case-controlled validation study was conducted in 2 community-based psychiatric practices under an institutional review board–approved protocol. Subjects were MDD psychiatric outpatients with a clinical diagnosis of MDD or clinical symptoms consistent with depression and were age- and sex-matched to healthy controls who were healthy volunteers. After completing informed consent, 28 MDD (10 men and 18 women) and 28 healthy subjects were prospectively enrolled. Subjects with MDD on antidepressants remained on those treatments. Medical history and a single blood sample for biomarker analysis were collected.

The MDD group had a mean age of 37.1 years (age interval, 21–50 years) and presented with a range of psychiatric disorders based on their DSM-IV diagnosis, including major depression, single-episode, or recurrent MDD, and all had a 21-item Hamilton Depression Rating (HAM-D) score more than 14. A few patients had a diagnosis of more than 1 psychiatric disorder, for example, MDD or generalized anxiety together with panic disorder. This group also had comorbidities, including arthritis, diabetes, hypertension, and obesity. The investigator did not feel that any existing comorbidity would affect the individual’s mental status. All patients with MDD were on 1 or combination treatment regimens for their psychiatric disorder (eg, Escitalopram, Lorazepam, Buproprion, and Duloxetine). The 28 healthy adults (10 men and 18 women) had a mean age of 37.2 years (age interval, 22.1–50 years). They were screened by an experienced psychiatrist and enrolled if they did not meet DSM-IV criteria for major Axis I disorder (except social anxiety disorder or a specific phobia) and had a HAM-D score less than 7. Exclusions included subjects with a disease associated with the HPA axis (eg, Addison disease), autoimmune or inflammatory disease (eg, rheumatoid arthritis), or a medical condition or regularly used nonsteroidal anti-inflammatory medications, steroids, or any other medication known to affect the inflammatory or HPA axis processes. Pregnant women were excluded as well as anyone who met DSM-IV for substance abuse or dependence 1 month before the study or had a history of adverse events or fainting during blood drawing.

The clinical sensitivity was estimated to be 85.7% (CI, 66.4%–95.3%), and the clinical specificity was 89.3% (CI, 70.6%–97.2%).

A comparison of the clinical performance of all methods discussed herein is summarized in the Table.

10. Conclusion

Psychiatric patients and their health care providers need better methods to diagnose depression given that no laboratory test exists today that is widely used to assess depression. An accurate diagnosis is essential, given that 50% or more of psychiatric patients are prescribed a series of single or combination antidepressants over months or years. Methods providing more accurate diagnoses would enhance trust between patients and their caregivers, thereby strengthening this relationship.

Currently, the administration of clinical evaluations and standard interviews to assess depression can be compromised by poor patient reporting and patient symptoms that can overlap with other disorders. In primary care, structured clinical evaluations are not often part of routine practice, resulting in highly variable and often inaccurate diagnosis of depression. Any effort to standardize or otherwise improve upon these approaches will bring only incremental improvement, given that these tools are subjective in nature.

Among the laboratory methods, DST may still have value for some patients if used with discretion and supported by an abundance of other clinical information. Its clinical performance remains insufficient to support broad use in assessing MDD. Greater promise can be found in biologically based laboratory methods that are objective in both methodology and interpretation. An additional advantage is that patient specimen procurement, measurement, and analysis are done outside the busy psychiatrist’s office, thereby minimizing office resources currently required to conduct a typical psychiatric interview.

Genomic, proteomic, and metabolic profiling to identify biomarkers with application to diagnosis MDD are still in their infancy. Better progress is seen in the multiparameter biomarker panel and algorithm, for which there is a biological rationale, and results can be generated using standard laboratory methodology and a routine blood sample. Its clinical performance and reliability are quite promising and better than DST or other emerging multi-parameter methods. Although the early studies were done with intended heterogeneous MDD populations, this could result in unpredictability in more well-defined populations. Therefore, additional studies are warranted to further demonstrate the panel’s clinical accuracy in patients with MDD on a defined treatment regimen. It would be important also to examine the test correlation with severity of depression as reflected by HAM-D or Montgomery-Asberg Depression Rating Scale (MADRS) scores. Similarly, more needs to be understood about the effect of potential immunological and rheumatologic comorbidities on test specificity given that some biomarkers in the panel reflect inflammatory processes. In spite of these limitations, the initial clinical accuracy of this MDD diagnostic biomarker panel holds promise to enhance the psychiatric professional’s diagnostic armamentarium in MDD.

One recently reported independent study evaluated this biomarker panel initially in 36 patients with MDD and 43 nondepressed subjects [40]. The test demonstrated a sensitivity and specificity of 91.7% and 81.3%, respectively, in differentiating between the 2 groups. A follow-on replication study of 34 subjects with MDD yielded similar performance with a sensitivity and specificity of 91.1% and 81%, respectively. These results compare favorably with the biomarker panel performance reported herein.

Acknowledgments

Financial support: Support for the development of this manuscript was provided by Ridge Diagnostics, Inc, San Diego, CA.

Footnotes

Conflict of interest/financial disclosure: KMS and JB are employees of Ridge Diagnostics, Inc, San Diego, CA. JB also owns stock in the company. PFR is a professor of psychiatry, University of Utah School of Medicine, Salt Lake City, UT, and a consultant to Novartis, Kyowa Hakko Kirin, Ltd, and Ridge Diagnostics, Inc, and has stock in Ridge Diagnostics, Inc.

References

  • 1.Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity, and comorbidity of twelve-month DSM-IV disorders in the National Comorbidity Survey Replication (NCS-R) Arch Gen Psychiatry. 2005;62(6):617–27. doi: 10.1001/archpsyc.62.6.617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.U.S. Census Bureau. Table 2 Annual Estimates of the Population by Selected Age Groups and Sex for the United States: April 1, 2000 to July 1, 2004 (NC-EST2004-02) Source: Population Division. U.S. Census Bureau; Population Estimates by Demographic Characteristics. Release Date: June 9, 2005. [Google Scholar]
  • 3.Costello J, Erkanli A, Angold A. Is there an epidemic of child or adolescent depression? J Child Psychol Psychiatry. 2006;47(12):1263–71. doi: 10.1111/j.1469-7610.2006.01682.x. [DOI] [PubMed] [Google Scholar]
  • 4.Lewinsohn PM, Rohde P, Seeley JR. Major depressive disorder in older adolescents: prevalence, risk factors, and clinical implications. Clin Psychol Rev. 1998;18(7):765–94. doi: 10.1016/s0272-7358(98)00010-5. [DOI] [PubMed] [Google Scholar]
  • 5.Cheung A. Canadian community health survey: major depressive disorder and suicidality in adolescents. Health Policy. 2006;2(2):76–89. [PMC free article] [PubMed] [Google Scholar]
  • 6.Whitaker A, Johnson J, Shaffer D, et al. Uncommon troubles in young people: prevalence estimates of selected psychiatric disorders in a nonreferred adolescent population. Arch Gen Psychiatry. 1990;47(5):487–96. doi: 10.1001/archpsyc.1990.01810170087013. [DOI] [PubMed] [Google Scholar]
  • 7.The World Health Organization. The global burden of disease: 2004 update, Table A2: Burden of disease in DALYs by cause, sex and income group in WHO regions, estimates for 2004. Geneva, Switzerland: WHO; 2008. [Google Scholar]
  • 8.Rettew DC. Commentary, diagnoses from clinical evaluations and standardized diagnostic interviews don’t agree. Psychiatry Times. 2010;27(6):1–3. [Google Scholar]
  • 9.Rettew DC, Lynch AD, Achenbach TM, Dumenci L, Ivanova MY. Meta-analysis of agreement between diagnoses made from clinical evaluations and standardized diagnostic interviews. Int J Methods Psychiatr Res. 2009;18(3):169–84. doi: 10.1002/mpr.289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Lowe B, Spitzer RL, Grafe K, Kroenke K, Quenter A, Zipfel S, et al. Comparative validity of three screening questionnaires for DSM-IV depressive disorders and physicians’ diagnoses. J Affect Disord. 2004;78(2):131–40. doi: 10.1016/s0165-0327(02)00237-9. [DOI] [PubMed] [Google Scholar]
  • 11.Stewart PM. The adrenal cortex. In: Kronenberg HM, Melmed S, Polonsky KS, Larsen PR, editors. Williams Textbook of Endocrinology. ch. 14 Philadelphia, PA: Saunders Elsevier; 2008. [Google Scholar]
  • 12.Fountoulakis KN, Gonda X, Rihmer Z, Fokas C, Iacovides A. Revisiting the dexamethasone suppression test in unipolar major depression: an exploratory study. Ann Gen Psychiatry. 2008;7(22):1–9. doi: 10.1186/1744-859X-7-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.The dexamethasone suppression test: an overview of its current status in psychiatry. The APA Task Force on Laboratory Tests in Psychiatry. Am J Psychiatry. 1987;144:1253–62. doi: 10.1176/ajp.144.10.1253. [DOI] [PubMed] [Google Scholar]
  • 14.Arana GW, Baldessarini RJ, Ornsteen M. The dexamethasone suppression test for diagnosis and prognosis in depression. Arch Gen Psychiatry. 1985;42(12):1193–204. doi: 10.1001/archpsyc.1985.01790350067012. [DOI] [PubMed] [Google Scholar]
  • 15.Le-Niculescu H, Kurian SM, Yehyawi N, Dike C, Patel SD, Edenberg HJ, et al. Identifying blood biomarkers for mood disorders using convergent functional genomics. Mol Psychiatry. 2009;14:156–74. doi: 10.1038/mp.2008.11. [DOI] [PubMed] [Google Scholar]
  • 16.Spijker S, Van Zanten JS, De Jong S, Penninx BW, van Dyck R, Zitman FG, et al. Stimulated gene expression profiles as a blood marker of major depressive disorder. Biol Psychiatry. 2010;68:179–86. doi: 10.1016/j.biopsych.2010.03.017. [DOI] [PubMed] [Google Scholar]
  • 17.Behan AT, Byrne C, Dunn MJ, Cagney G, Cotter DR. Proteomic analysis of membrane microdomain-associated proteins in the dorsolateral prefrontal cortex in schizophrenia and bipolar disorder reveals alterations in LAMP, STXBP1, and BASP1 protein expression. Mol Psychiatry. 2009;14:601–13. doi: 10.1038/mp.2008.7. [DOI] [PubMed] [Google Scholar]
  • 18.Prabakaran S, Wengenroth M, Lockstone HE, Lilley K, Leweke FM, Bahn S. 2-D DIGE analysis of liver and red blood cells provides further evidence for oxidative stress in schizophrenia. J Proteome Res. 2007;6:141–9. doi: 10.1021/pr060308a. [DOI] [PubMed] [Google Scholar]
  • 19.Martins-de-Souza D, Dias-Neto E, Schmitt A, Falkai P, Gormanns P, Maccarrone G, et al. Proteome analysis of schizophrenia brain tissue. World J Biol Psychiatry. 2010;11(2):110–20. doi: 10.3109/15622970903490626. [DOI] [PubMed] [Google Scholar]
  • 20.Nibuya M, Morinobu S, Duman RS. Regulation of BDNF and trkB mRNA in rat brain by chronic electroconvulsive seizure and antidepressant drug treatments. J Neurosci. 1995;15:7539–47. doi: 10.1523/JNEUROSCI.15-11-07539.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kromer SA, Kessler MS, Milfay D, Birg IN, Bunck M, Czibere L, et al. Identification of glyoxalase-I as a protein marker in a mouse model of extremes in trait anxiety. J Neurosci. 2005;25(17):4375–84. doi: 10.1523/JNEUROSCI.0115-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Ditzen C, Varadarajulu J, Czibere L, Gonik M, Targosz BS, Hambsch B, et al. Proteomic-based genotyping in a mouse model of trait anxiety exposes disease-relevant pathways. Mol Psychistry. 2010;15:702–11. doi: 10.1038/mp.2008.146. [DOI] [PubMed] [Google Scholar]
  • 23.Paige LA, Mitchell MW, Krishnan KR, Kaddurah-Daouk R, Steffens DC. A preliminary metabolomic analysis of older adults with and without depression. Int J Geriatr Psychiatry. 2007;22:418–23. doi: 10.1002/gps.1690. [DOI] [PubMed] [Google Scholar]
  • 24.Shelton RC. The molecular neurobiology of depression. Psychiatr Clin North Am. 2007;1:1–11. doi: 10.1016/j.psc.2006.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Kenis G, Maes M. Effects of antidepressants on the production of cytokines. Int J Neuropsychopharmacol. 2002;5(4):401–12. doi: 10.1017/S1461145702003164. [DOI] [PubMed] [Google Scholar]
  • 26.Maes M. Major depression and activation of the inflammatory response system. In: Dantzer R, Wollman EE, Yirmiya R, editors. Cytokines, Stress and Depression. New York: Kluwer Academic/Plenum Publishers; 1999. pp. 25–46. [DOI] [PubMed] [Google Scholar]
  • 27.Zhu CB, Blakely RD, Hewlette WA. The proinflammatory cytokines interleukin-1 beta and tumor necrosis factor-alpha activate serotonin transporters. Neuropsychopharmacology. 2006;31:2121–31. doi: 10.1038/sj.npp.1301029. [DOI] [PubMed] [Google Scholar]
  • 28.Tichomirowa MA, Keck ME, Schneider HJ, Paez-Pereda M, Renner U, Holsboer F, et al. Endocrine disturbances in depression. J Endocrinol Invest. 2005;28(1):89–99. doi: 10.1007/BF03345535. [DOI] [PubMed] [Google Scholar]
  • 29.Maes M, Yirmyia R, Noraberg J, Brene S, Hibbeln J, Perini G, et al. The inflammatory & neurodegenerative (I&ND) hypothesis of depression: leads for future research and new drug developments in depression. Metab Brain Dis. 2009;24(1):27–53. doi: 10.1007/s11011-008-9118-1. [DOI] [PubMed] [Google Scholar]
  • 30.Simon NM, McNamara K, Chow CW, Maser RS, Papakostas GI, Pollack MH, et al. A detailed examination of cytokine abnormalities in major depressive disorder. Eur Neuropsychopharmacol. 2008;18(3):230–3. doi: 10.1016/j.euroneuro.2007.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lehto SM, Huotari A, Niskanen L, Tolmunen T, Koivumaa-Honkanen H, Honkalampi K, et al. Serum adiponectin and resistin levels in major depressive disorder. Acta Psychiatr Scand. 2010;121(3):209–15. doi: 10.1111/j.1600-0447.2009.01463.x. [DOI] [PubMed] [Google Scholar]
  • 32.Sen S, Duman R, Sanacora G. Serum brain-derived neurotrophic factor, depression, and antidepressant medications: meta-analyses and implications. Biol Psychiatry. 2008;64(6):527–32. doi: 10.1016/j.biopsych.2008.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Fernandes BS, Gama CS, Kauer-Sant’Anna M, Lobato MI, Belmonte-de-Abreu P, Kapczinski F, et al. Serum brain-derived neurotrophic factor in bipolar and unipolar depression: a potential adjuntive tool for differential diagnosis. J Psychiatr Res. 2009;43(15):1200–4. doi: 10.1016/j.jpsychires.2009.04.010. [DOI] [PubMed] [Google Scholar]
  • 34.Brouwer JP, Appelhof BC, Hoogendijk WJG, Huyser J, Endert E, Zuketto C, et al. Thyroid and adrenal axis in major depression: a controlled study in outpatients. Eur J Endocrinol. 2005;152:185–91. doi: 10.1530/eje.1.01828. [DOI] [PubMed] [Google Scholar]
  • 35.Holsboer F, Barden N. Antidepressants and hypothalamic-pituitary-adenocortical regulation. Endocr Rev. 1996;17:187–205. doi: 10.1210/edrv-17-2-187. [DOI] [PubMed] [Google Scholar]
  • 36.Renshaw P, Bilello J, Pi B. Multianalyte biomarker blood test to aid in diagnosis, treatment and management of major depressive disorder. American Psychiatric Association Meeting; May 2009; poster NR7-014. [Google Scholar]
  • 37.Bilello J, Pi B, Lyoo IK. Biomarker hypermapping as an aid to the diagnosis, treatment and management of major depressive disorder. U.S. Psychiatric and Mental Health Congress; November, 2009; poster 106. [Google Scholar]
  • 38.Kirkegaard C, Faber J. The role of thyroid hormones in depression. Eur J Endocrinol. 1998;138:1–9. doi: 10.1530/eje.0.1380001. [DOI] [PubMed] [Google Scholar]
  • 39.McDermott MT, Ridgway EC. Subclinical hypothyroidism is mild thyroid failure and should be treated. J Clin Endocrinol Metab. 2001;86:4585–90. doi: 10.1210/jcem.86.10.7959. [DOI] [PubMed] [Google Scholar]
  • 40.Papakostas G, Shelton RC, Kinrys G, Henry ME, Bakow BR, Lipkin SH, et al. Assessment of a multi-assay, serum-based biological diagnostic test for major depressive disorder: a pilot and replication study. Mol Psychiatry. 2011:1–8. doi: 10.1038/mp.2011.166. [DOI] [PubMed] [Google Scholar]

RESOURCES