The Nature of Depression: An Updated Review

By Ahmed Moustafa

The Nature of Depression: An Updated Review provides clear connections between psychiatric and neurological disorders. Unlike prior books on depression, this book covers many neurological and psychiatric disorders, including Parkinson’s disease, major depressive disorder, Alzheimer’s disease, PTSD, addiction and anxiety disorder. In addition, this book covers different forms of depression, including transition-induced depression and the development of depression following major life events, including birth of a child, menopause and retirement.

• Covers depression comorbidity with psychological and neurological disorders
• Reviews comorbidity with addiction, anxiety, trauma and psychosis
• Compares the symptoms of subclinical depression to major depression
• Discusses how stress and sleep impact depression
• Theorizes the path of depression following negative life transitions

• Language: English
• Publisher: Academic Press
• Release date: Oct 9, 2020
• ISBN: 9780128176771

Ahmed Moustafa

Dr. Ahmed Moustafa is a Professor of Psychology and Computational Modeling at School of Psychology, Bond University, Gold Coast, Queensland, Australia. Prior to moving to Bond University, Ahmed was an associate professor in Psychology and Neuroscience at Marcs Institute for Brain, Behavior, and Development & School of Psychology, Western Sydney University. Ahmed is trained in computer science, psychology, neuroscience, and cognitive science. His early training took place at Cairo University in mathematics and computer science. Before joining Western Sydney University as a lab director, Ahmed spent 11 years in America working on several psychology and neuroscience projects. Ahmed conducts research on computational and neuropsychological studies of addiction, schizophrenia, Parkinson’s disease, PTSD, depression, Alzheimer’s disease. He has published over 240 papers in high-ranking journals including Science, PNAS, Journal of Neuroscience, Brain, Neuroscience and Biobehavioral Reviews, Nature (Parkinson’s disease), Neuron, among others. Ahmed has obtained grant funding from Australia, USA, Qatar, UAE, Turkey, and other countries. Ahmed has recently published ten books: (1) Computational models of brain and behavior; (2) Social Cognition in Psychosis, (3) computational Neuroscience Models of the Basal Ganglia, (4) Cognitive, Clinical, and Neural Aspects of Drug Addiction; (5) The Nature of Depression: An updated review; (6) Big data in psychiatry and neurology; (7) Alzheimer’s Disease: Understanding Biomarkers, Big Data, and Therapy. Elsevier; (8) Cognitive and Behavioral Dysfunction in Schizophrenia; (9) Female Pioneers from Ancient Egypt and the Middle East; and (10) Mental health effects of COVID-19. In the last 10 years, Ahmed has published collaboratively with 71 colleagues, has more than 510 co-authors, from 35 institutions in 14 countries. Ahmed is now Editor-in-Chief of Discover Psychology, a new journal by Springer Nature.

Book preview

book.

Part I

Recent topics on major depressive disorders

Chapter 1: Peripheral biomarkers in major depressive disorders

Judit Lazarya; Daniel Miezahb; Ahmed A. Moustafac,d; Szabolcs Kéria    a National Institute of Psychiatry and Addictions, Budapest, Hungary

b Psychology Department, Macquarie University, Sydney, NSW, Australia

c School of Psychology and Marcs Institute for Brain and Behaviour, Western Sydney University, Sydney, NSW, Australia

d Department of Human Anatomy and Physiology, Faculty of Health Sciences, University of Johannesburg, Johannesburg, South Africa

Abstract

Peripheral biomarkers would be useful tools for the daily clinical diagnosis, monitoring, and therapy of major depressive disorders (MDD). Although there are intensive investigations to identify specific biomarkers, applicable and validated tests are still not available. In this review, we provide an up-to-date summary of the most robust biomarker studies of MDD. There are accumulating results in this field primarily in relation to the hypothalamus-pituitary-adrenal (HPA) axis, the inflammatory system, and the neurogenesis/growth factors, but these data are heterogeneous and the investigated biomarkers are not specific for MDD. As a vicious circle, the discrepancy of biomarker studies’ results can be explained by the difficulties of phenotypic measurements, which should be improved by biomarker tests. Thus, more sophisticated phenotype definitions (e.g. subphenotypes, enodphenotypes, etc.) are needed for more accurate results. The development of standardized biomarker lab tests is possible in the near future, especially, as we now have more sensitive and high throughput detecting systems. The biomarkers would greatly help not only in clinical practice but also in drug development.

Keywords

Major depressive disorder (MDD); Biomarkers; Cognition; Treatment-resistant patients; Subtypes

Introduction

The identification of specific peripheral biomarkers would be a great step in the clinical management of major depressive disorder (MDD). However, despite the number intensive investigations, peripheral biomarkers of MDD are still not available. The use of peripheral biomarkers would help in the differential diagnosis as well as in the drug selection with higher response rates for first choice and in state monitoring. Due to the current protocol of the treatment (trial-and-error), MDD patients have to endure side-effects and progression of depression in the case of ineffective antidepressant treatment for 3–4 weeks because their therapeutic effect develops considerably slowly. The lengthy process of trying different antidepressants to find the most effective medication may decrease the confidence of the patient, which further worsen the chance for recovery. The typical reasons of ineffectively attempting to find antidepressants are inadequate diagnosis based on insufficient information about the patient (e.g. lack of medical history) and this difficulty could be improved by biomarker measurements, as we argue in this article.

Untreated MDD can increase the risk of suicide and lead to a great financial burden (Ferrari et al., 2013). Accordingly, there is a large number of investigations on the potential peripheral biomarkers to achieve more optimal therapeutic intervention with the reduced depressed episode (Kraus, Kadriu, Lanzenberger, Zarate Jr., & Kasper, 2019). The most widely studied biomarkers are related to the hypothalamus-pituitary-adrenal (HPA) axis, the inflammatory system, and the neurotrophic factors (Menke, 2019; Pinto, Moulin, & Amaral, 2017). Recently more special molecules, such as micro-RNA, mitochondrial components or epigenetic patterns have also been studied. Although these are promising investigations, further studies are required to confirm them. In addition, there are other neural and cognitive biomarkers related to the development of depression. Below, we discuss each of these biomarkers in detail.

HPA axis related biomarkers

The HPA axis plays a key role in the pathomechanism of MDD and it forms complex networks with other molecular systems involved in depressive symptoms (Pariante & Lightman, 2008). Stressful stimuli induce corticotropin-releasing hormone (CRH), which is produced in the paraventricular nuclei of the hypothalamus, then adrenocorticotropic hormone (ACTH) is released from the pituitary gland to the blood stream. The circulating ACTH causes glucocorticoid release from the adrenal cortex, which thenpenetrateso all organs in the body including the brain. Glucocorticoids bind to the glucocorticoid receptors (GR) and the mineralocorticoid receptors (MR). Besides the effects on peripheral tissue as a part of the stress response, they mediate an inhibiting effect on the HPA axis as well via blocking the production of hypothalamic CRH and pituitary ACTH. This mechanism is the key point of the negative feedback on the HPA system, and it depends on the glucocorticoid level of blood. There is an enormous body of evidence supporting that dysregulation of the negative feedback is the core of depression (Holsboer, 2000a, 2000b). In agreement with this, CRH mRNA is reduced in patients with MDD and the serum level of CRH is correlated with the severity of depression (Catalan, Gallart, Castellanos, & Galard, 1998). Further, GR resistance was proved in treatment resistant depression, and a sustained high level of glucocorticoids predicted early relapse (Bauer et al., 2003; Ising et al., 2007; Juruena et al., 2009; Ribeiro, Tandon, Grunhaus, & Greden, 1993).

The most important condition of balanced HPA function is intact GR and MR signaling. Therefore, the MR/GR ratio was investigated in multiple studies as potential biomarkers (Klok et al., 2011; Qi et al., 2013). In chronic stress, the inhibition of neurogenesis through prolonged elevated serum glucocorticoid level and reduced MR/GR ratio were observed in patients with MDD (Anacker et al., 2013; Holsboer, 2000a). Moreover, a sustained high level of glucocorticoids leads to the desensitization and the resistance of GR. In the presence of GR resistance, the negative feedback turns into a positive loop resulting in a pathological state (Stapelberg et al., 2018). GR resistance is a complex process and it can lead to depression in numerous ways. One such potential way is via the neuro-inflammatory pathway (Stapelberg et al., 2018).

Cytokines as potential biomarkers in MDD

Cytokines are widely investigated as a MDD biomarker since there are strong evidences supporting the inflammatory theory of the MDD pathomechanism (Dowlati et al., 2010). Molecular and pharmacological studies unanimously suggest that the overactivation, sustained, or inadequate activation of the inflammatory system plays a pivotal role in the pathomechanism of MDD (Dowlati et al., 2010). Further, the anti-inflammatory effect of certain antidepressant drugs were approved (Hannestad, DellaGioia, & Bloch, 2011), and some studies confirmed that depressive symptoms were reduced by monoclonal cytokine antibody in patients with chronic inflammatory diseases (Haroon, Raison, & Miller, 2012; Jha & Trivedi, 2018). Moreover, there are some important data showing that patients with higher inflammatory parameters show weaker response to antidepressants (Benedetti, Lucca, Brambilla, Colombo, & Smeraldi, 2002; Lanquillon, Krieg, Bening-Abu-Shach, & Vedder, 2000; Sluzewska, Sobieska, & Rybakowski, 1997).

The ratio of pro- and anti-inflammatory cytokines may indicate the development of MDD (Maes & Carvalho, 2018). The immature immune cells produce proinflammatory cytokines, such as IL-6, IL-1β, and TNF-α, and they modulate the function of T-helper 1 (Th1) lymphocytes. Proinflammatory cytokines stimulate the production of acute phase proteins in peripheral cells contributing to the systemic low-grade inflammation. Anti-inflammatory cytokines (IL-4, IL-5, and IL-10) are expressed by T-helper 2 (Th2) lymphocytes, and they antagonize the effects of proinflammatory cytokines. The equilibrium of the two subgroups of cytokines is under the regulation of the HPA axis. The GR itself is able to bind to the glucocorticoid response element (GRE) of the DNA, which promotes the transcription of the proinflammatory cytokines. Pathological GR activity and glucocorticoid resistance lead to a dysfunctional T-cell function and abnormal neurogenesis and neuroplasticity, which is a further connecting point between the HPA and inflammation. An altered number and dysfunctions of Th2, Th17, and NK (Natural Killer) cells were demonstrated in patients with MDD, thus these dysfunctional immune cells are recommended as biomarkers in MDD (Mazza et al., 2018).

Some studies have also focused on the level of cytokines in the human serum in association with MDD (Himmerich, Patsalos, Lichtblau, Ibrahim, & Dalton, 2019; Lin, Ding, Wu, Dong, & Li, 2018). Despite the increasing number of positive data, the results are not equivocally replicated, and the findings are not specific for depression. It has been shown that a higher level of proinflammatory cytokines before the treatment predicted lack of future clinical response to antidepressants (Cattaneo et al., 2013; Zunszain et al., 2012).

It is noteworthy that studies on special subgroups of depression resulted in more promising data than studies comparing patients with MDD and healthy controls. Some studies suggest that a higher IL-6 level is associated with melancholic depression, while an elevated TNF-alpha level is correlated with atypical depression, which implicates a different biological mechanism underlying the two phenotypes (Yang et al., 2018). However, in another investigation, there were higher IL-6, TNF-alpha and C-Reactive Protein (CRP) concentrations in atypical depression compared to melancholic and healthy controls (Yang et al., 2018). Partly in agreement with this finding, significantly reduced CRP was reported in melancholic depression, while it was not observed in atypical depression (Yang et al., 2018). In a recently published study, it was found that elevated pro-inflammatory cytokine levels in post-stroke state predicted depression (Levada & Troyan, 2018). According to a currently published study, elevated CRP level was found in treatment-resistant patients with depression, moreover, it is suggested that second-line treatment with anti-inflammatory agents may be effective in these cases (Chamberlain et al., 2019). In a recentstudy, defective inflammatory pathways were demonstrated in never-treated patients with depression based on pro- and anti-inflammatory cytokine profile measurements (Syed et al., 2018). Interestingly greater reduction of proinflammatory markers during cognitive-behavioral psychotherapy was associated with more pronounced clinical improvement in unmedicated MDD, although IL-6 and CRP were not among the most sensitive clinical state markers (Keri, Szabo, & Kelemen, 2014). The heterogeneity of these prior results can be explained by differences in sample size, inclusion and exclusion criteria, mean age of study participants, measurement of depression, different phenotypes of patient groups, comorbid conditions, and theused antidepressants (Yang et al., 2018).

Neurotrophic factors as biomarkers in MDD

The neurotrophic theory of depression was described by Duman, Heninger and Nestler in 1997 focusing on the potential role of Brain Derived Neurotrophic Factor (BDNF) (Duman, Heninger, & Nestler, 1997). Neurotrophic factors play a fundamental role in neurogenesis, neuroproliferation and plasticity, which, in case of pathological modulation, are key points in the development of depression (Lee & Kim, 2011). In addition, experimental data have shown that these neurotrophic factors mediate the effect of antidepressants and electroconvulsive therapy (ECT) with increased levels during the treatment (Bocchio-Chiavetto et al., 2010; Cattaneo et al., 2010; Pandey et al., 2010). Several neurotrophic factors related to the development of MDD are reported in the literature including GDNF (glial cell-derived neurotrophic factor), IGF-1 (insulin-like growth factor 1), VEGF (vascular endothelial growth factor), NGF (nerve growth factor), NGR-1 (nogo-66 receptor 1), FGF (fibroblast growth factor), and TGF-β (transforming growth factor) (Sharma, da Costa e Silva, Soares, Carvalho, & Quevedo, 2016).

The pathological role of BDNF in MDD is also related to the HPA axis (Stapelberg et al., 2018). The phosphorylation of the GR determines receptor function, and dysfunctional phosphorylation is related to the development of MDD (Stapelberg et al., 2018). In chronic stress, an increased phosphorylation of GR was observed, and it inhibited the proliferating effect of BDNF in the pituitary progenitor cells. The concentrations of BDNF were reported to be significantly lower in patients with MDD and during the depressive episode in patients with bipolar depression as well (Piccinni et al., 2015). Moreover, the concentrations of BDNF were correlated with the severity and the duration of depressive episodes, and with the presence of psychotic symptoms (Galvez-Contreras et al., 2016). Polyakova, Schroeter, et al. (2015) and Polyakova, Stuke, et al. (2015) concluded in their meta-analysis that only the serum level and not the plasma concentration showsa pronounced association with a depressive episode (Polyakova, Schroeter, et al., 2015; Polyakova, Stuke, et al., 2015). Nevertheless, the unchanged BDNF level in both the plasma and serum in the first weeks of treatment predicted treatment resistance (Lieb et al., 2018). In victims of suicide, the BDNF level was significantly lower than in individuals died by natural ways (Galvez-Contreras et al., 2016).

Besides BDNF, VEGF was also investigated as a potential biomarker in MDD. The level of VEGF in the serum was elevated in patients suffering from MDD in several studies; however, some authors reported decreased levels or a lack of association between serum VEGF level and MDD severity (Takebayashi, Hashimoto, Hisaoka, Tsuchioka, & Kunugi, 2010). In our studies, we found a significantly higher VEGF concentration in treatment resistant depression (Elemery, Kiss, Gonda, Dome, & Lazary, 2017; Halmai et al., 2013). Further investigations are required to clarify these associations between VEGF level and MDD. The complex relationship among HPA axis, inflammatory system, and neurotrophic factors are shown in Fig. 1.

Fig. 1 The interacting matrix of HPA axis, inflammation, and neurotrophic factors in healthy (A) and pathological (B) cases. In healthy conditions (A) the level of glucocorticoids (Glcs) in the serum increased temporarily and they bound mainly to the mineralocorticoid receptors (MR). In a pathological state (e.g. chronic stress) (B) the sustained glucocorticoid level lead to overactivation and then the resistance of the glucocorticoid receptors (GR), high CRH levels (negative feedback turns into a positive loop), together with increased inflammation and blocked neurotrophic effects.

Neuroimaging biomarkers of MDD

Neuroimaging studies have reported some potential biomarkers for MDD. EEG studies have reported antidepressant treatment response (ATR) index as a potential biomarker in MDD (Leuchter, Cook, Gilmer, et al., 2009; Leuchter, Cook, Marangell, et al., 2009). In one study, it was found that ATR index or value (measured by EEG) predicted responsiveness to different antidepressants such as escitalopram and bupropion (Leuchter, Cook, Gilmer, et al., 2009). However, when the two drugs were combined in treating MDD patients, the ATR index or value could not predict responsiveness to them. Future studies are needed to examine the possibility of ATR index in predicting responsiveness to combined antidepressant medications for the treatment of MDD.

Changes in EEG frequency bands have also been explored as a potential biomarker in MDD. It has been found that changes in EEG alpha band activity before treatment with antidepressants (such as clomipramine, imipramine, paroxetine, and fluoxetine) can show differences between MDD patients who are responding or not responding to treatment (Bruder et al., 2001; Knott, Mahoney, Kennedy, & Evans, 2000). Changes in EEG theta band activity has been studied as a biomarker in MDD. A decline in EEG theta band activity before treatment has been linked to response to treatment with antidepressants such as TCA and imipramine (Iosifescu et al., 2009). On the other hand, some researchers have found that an increase in the activity of EEG theta band prior to treatment can distinguish between those who responded to or those who did not respond to paroxetine during six weeks after treatment (Knott et al., 2000). Such inconsistencies between the previous studies need to be further examined thoroughly by future research.

The alpha hemispheric asymmetry is another possible biomarker of MDD which has been investigated in EEG research. It has been found that alpha asymmetry predicted the outcome of treatment with antidepressants (TCAs and the SSRI such as fluoxetine) between 4 weeks and 12 weeks after the treatment (Bruder et al., 2001). Theta cordance has been suggested a potential biomarker of MDD in EEG research. Researhcers have reported that a decline in frontal theta cordance in MDD patients which was examined after one week of treatment with antidepressants (such as fluoxetine, and venlafaxine) predicted response to treatment (Bares et al., 2008; Cook et al., 2002). In EEG studies, event-related potential (ERP) has been reported as a potential biomarker in MDD. While the latency of the P300 component of the ERP may be delayed in MDD patients, it has been found to return to normal levels in four weeks posttreatment with antidepressant medications (Hetzel et al., 2005). Moreover, strong loudness dependent auditory evoked potential (LDAEP) component of the ERP at threshold levels was linked with response to antidepressants such as fluoxetine and citalopram (Juckel et al., 2007).

Structural and functional neuroimaging studies conducted on MDD have revealed some biomarkers for the disorder. The hippocampus is one of the biomarkers identified by such studies. Changes in the volume of hippocampus have been reported to play a role in the occurrence of MDD (Bremner et al., 2000). There are decreases in hippocampus volumes in depression (Cole, Costafreda, McGuffin, & Fu, 2011). Studies have reported a reversal in reduced hippocampal volume in MDD patients when antidepressants are used to treat them (Arnone et al., 2013; Tendolkar et al., 2013). Findings from structural and functional neuroimaging studies have suggested that the changes in anterior cingulate cortex is a potential biomarker in MDD. An increasein gray matter levels in the anterior cingulate cortex (ACC) before treatment predicted remission to the antidepressant fluoxetine (Costafreda, Chu, Ashburner, & Fu, 2009). Metabolic decreases in the ACC have also been linked with a response to the antidepressant, fluoxetine (Mayberg et al., 2000). Changes in the prefrontal cortex is another potential biomarker in MDD. A study showed that higher baseline activities in the ventrolateral prefrontal cortex resulted in a poorer response to antidepressant medications in MDD patients (Keedwell et al., 2010). It has been reported that low baseline activities in the dorsolateral prefrontal cortex lead to improvement in depressive symptoms (Miller et al., 2013). Functional neuroimaging studies have indicated that the amygdala is a possible biomarker in MDD. More activities in the amygdala before treatment was shown to predict improvement in depressive symptoms after MDD patients were treated with escitalopram (Langenecker et al., 2007). Imaging data have reported on the insula as possible biomarker in MDD research. A recent functional imaging study reported that greater activities in the bilateral insula resulted in improvement in depressive symptoms when antidepressants were used in treating MDD patients (Cullen et al., 2016). In addition, metabolic decreases in the insula have been linked with response to antidepressant Fluoxetine in MDD studies (Mayberg et al., 2000).

Cognitive biomarkers of MDD

Some cognitive biomarkers have been identified in MDD research. One of such cognitive biomarkers in MDD is executive dysfunctions. Between 20% and 30% of MDD patients have severe deficits in executive functions (McIntyre et al., 2013). Some studies have reported that executive functions deficits in MDD may be linked with frontal lobe impairments (Koolschijn, van Haren, Lensvelt‐Mulders, Hulshoff Pol, & Kahn, 2009; Lorenzetti, Allen, Fornito, & Yücel, 2009). Regarding the HPA-inflammatory hypothesis of MDD, we showed that enhanced microglial activation in the prefrontal cortex of patients with MDD was associated with less efficient attentional processing (Li, Sagar, & Keri, 2018). Individuals with MDD display poor executive function than healthy controls especially during the early period of depression (Lee, Hermens, Porter, & Redoblado-Hodge, 2012). Psychomotor abnormalities have been suggested as one of the possible cognitive biomarkers in MDD (Bennabi, Vandel, Papaxanthis, Pozzo, & Haffen, 2013). Researchers have associated psychomotor deficits in MDD to impairments in frontostriatal dopaminergic neurotransmission (Hoeppner, Prudente-Morrissey, Herpertz, Benecke, & Walter, 2009). Interestingly, dopaminergic loss and decreased volumes in the hippocampal formation (CA2-CA3 subfields) may result in the emergence of depressive symptoms and memory imapirments (Gyorfi et al., 2017). It has also been reported that activations in the frontal, limbic, and temporal areas of the brain during psychomotor activities may cause better treatment response in people with MDD (Langenecker et al., 2007). Verbal memory deficits have also been identified as possible cognitive biomarkers of MDD. Verbal memory impairments seem to increase the possibility of poor treatment prognosis with antidepressants (Alexopoulos et al., 2005).

Concluding remarks

The studies on peripheral biomarkers have provideed promising results, but the findings are heterogenous and not specific for MDD. Therefore, currently, there are no recommended blood tests for the routine diagnosis of MDD. The relationship between peripheral biomarkers and changes in the central nervous system is also dubious, although. numerous studies addressed this issue, and some of these reports confirmed the validity and rationale of peripheral biomarkers (Felger et al., 2018; Kanegawa et al., 2016). Furthermore, some investigations showedthat dysfunctions of the blood brian barrier can be a potential pathomechism of certain psychiatric conditions including the affective disorders (Kealy, Greene, & Campbell, 2018).

It can be arguedfrom the biomarker research that the HPA axis plays a key role in the pathomechanism of depression, and it forms an overlapping molecular network with other biochemical pathways implicated in inflammation and neurotrophic factor regulation (Stapelberg et al., 2018). These neurobiological mechanisms may contribute to the cognitive and neuroimaging alterations in MDD with a special reference to the prefrontal cortex and hippocampal formation. However, despite the fact that there are some common molecular pathways in the pathomechanism of MDD, it is not likely that any of them may solely explain the pathophysiology of MDD. The above evolved results suggest that, instead of distinct diagnostic categories with rigid boundaries, there are different subgroups or phenotypical dimensions of depression characterized by different biological mechanisms (Pinto et al., 2017). We conclude is that a pattern of multiple markers, including cognitive and neural ones, would be more informative than a single biomarker. It is also noteworthy that potential peripheral biomarker tests cannot lack thorough clinical examination, and the interpretation of test results require a prudent approach because the level of commonly used biomarkers depends on numerous confounding conditions. Based on the results of biomarker investigations, which we discussed above, a clinical tool might be developed in the near future, which could help to provide a more accurate diagnosis, effective treatment design, and proper clinical state monitoring in patients with MDD. However, it is important to highlight that the use of such tools will not be able to replace, but rather aid the clinical examination of patients and the clinical knowledge of health care