Major Depressive Disorder (MDD) is typically considered a mental illness, yet pathology associated with MDD is evident in cells and organs throughout the body. For example, MDD is associated with an increased risk of developing atherosclerosis, heart disease, hypertension, stroke, cognitive decline, and dementia (including Alzheimer's disease), osteoporosis, immune impairments (eg, “immunosenescence”), obesity, metabolic syndrome, insulin resistance, and type 2 diabetes,1-4 and individuals who are afflicted both by MDD and one of these diseases have a poorer prognosis than individuals afflicted by either alone.3 This increased risk of serious medical diseases is not fully explained by lifestyle choices such as diet, exercise, and smoking, and the reasons for the heightened risk remain unknown.4 Moreover, many of the medical comorbidities seen in MDD are diseases more commonly seen with advanced age, and MDD has even been characterized as a disease of “accelerated aging.”1,5,6 In this review article, we explore certain biological mediators that are dysregulated in MDD and that may contribute to the depressed state itself, to the comorbid medical conditions, and to “accelerated aging.” Discovering novel pathological mediators in MDD could help identify new targets for treating depression and its comorbid medical conditions and could help reclassify MDD as a multisystem disorder rather than one confined to the brain.

Theoretical model

We propose a model of MDD comprised of certain pathogenic processes that are interlinked and often recursive, that occur in the brain and in the periphery, and that can culminate in cellular damage, cellular aging, and disease.6-10 This model is presented schematically in Figure 1 and is briefly described in this introduction; the individual moderators and mediators are described in greater detail in the remainder of this article. This model is not intended to be complete or all-encompassing but is meant to highlight and connect certain interesting findings in the study of depression. It does not propose that each component is necessary or sufficient, or that the specified mediators are the sole routes to MDD. It also does not speak to the directions of causality between depression and physical pathology. Further, many of the specified mediators may serve either protective or destructive functions depending on their context and chronicity.11-13 Nonetheless, the model presented here provides testable hypotheses for further investigation and provides rationales for considering novel treatment approaches. Earlier reviews of this model have been published elsewhere.5,6,10

Figure 1.
Figure 1.: Model of multiple pathways leading to psychiatric and physical llness and cell aging. In conjunction with genetic and epigenetic moderators, elevated cortisol levels, associated with downregulation of glucocorticoid receptor (GC) function (GC resistance) may result in altered immune function, leading to excessive synthesis of proinflammatory cytokines. Changes in glucocorticoidmediated activities also result in genomic changes (altered levels of certain neurotransmitters, neurotrophins, and other mediators), as well as dysregulation of the limbic-hypothalamicpituitary adrenal (LHPA) axis that might contribute to neuroendangerment or neurotoxicity, perhaps leading to depressive or cognitive symptoms. Dysregulation of the LPHA axis can also lead to intracellular glucose deficiency, glutamatergic hyperactivity, increased cellular calcium concentrations, mitochondrial damage, free radial generation, and increased oxidative stress. This cascade of events, coupled with a milieu of increased inflammatory cytokines, may lead to accelerated cellular aging via effects on the telomere/telomerase maintenance system. Dysregulation of normal compensatory mechanisms, such as increased neurosteroid or neurotrophin production, may further result in inability to reduce cellular damage, and thereby exacerbate destructive processes. This juxtaposition of enhanced destructive processes with diminished protective/restorative processes may culminate in cellular damage, apoptosis, and physical disease. Allopreg, allopregnanolone; BDNF, brain-derived neurotrophic factor; CRH, corticotropin releasing hormone; DHEA, dehydroepiandrosterone; GR, glucocorticoid receptor

In brief, psychological and physical stressors trigger physiological responses that are acutely important for successful adaptation to the stress (“stress arousal”). However, when stress responses are disrupted or inappropriately prolonged, endangering effects may supersede the protective ones. The “cost” to the organism of maintaining these physiological responses over prolonged periods has been termed “allostatic load”13 or “arousal pathology,”14 and it has repeatedly been associated with poor medical outcomes.12 In addition to chronicity of the stress response, certain psychological, environmental, genetic, and epigenetic circumstances (discussed below) favor dysregulation of two main stress response effectors, the limbic-hypothalamic -pituitaryadrenal (LHPA) axis and the locus coeruleus noradrenergic (NE) system.15 A particular problem may arise when these two systems, which are generally counterregulatory, activate one another for prolonged periods of time (as may be seen in melancholic depression).15 The failure of glucocorticoids (GCs) to effectively counter-regulate stress-induced NE and LHPA activity may underlie critical aspects of MDD.15 Prolonged LHPA axis dysregulation can lead to neuroendangering or neurotoxic effects in vulnerable brain regions (eg, prefrontal cortex and hippocampus).16 It can also lead to energetic disturbances (decreased intracellular glucose availability and insulin resistance), glutamatergic hyperactivity/excitotoxicity, increased intracellular calcium concentrations, mitochondrial damage, free radical generation and oxidative stress, immune alterations (leading to a proinflammatory milieu), and accelerated cell aging (via effects on the telomere/telomerase maintenance system). The nature of cortisol abnormalities in MDD is complex, however, and will be discussed below. Prolonged activation of central NE systems, as often seen in melancholic depression, may be associated with worsened outcome in cardiovascular diseases and with accelerated cell aging at the level of the telomere.9,15

In addition to increases in these destructive processes, normal compensatory or reparative processes may be diminished, eg, diminution of counter-regulatory neurosteroids, eg, dehydroepiandrosterone (DHEA),17 and allopregnanolone,18 decreased antioxidant compounds, diminished anti-inflammatory/immunomodulatory cytokines, decreased neurotrophic factor concentrations, eg, brainderived neurotrophic factor (BDNF), and altered telomerase activity. This juxtaposition of enhanced destructive processes with diminished (or inadequate) protective or restorative ones can culminate in cellular damage and physical disease (Table I). This model will be explored in greater depth in the following sections.

Possibly damaging and protective mediators in major depression LHPA, limbic-hypothalamic-pituitary-adrenal; DHEA, dehydroepiandrosterone; BDNF, brain-derived neurotrophic factor. * Evidence is mixed as to whether DHEA concentrations are elevated or lowered in depression. ** Evidence is mixed as to whether the anti-inflammatory/immunomodulatory cytokine, IL10, is elevated or lowered in depression. *** Evidence is mixed as to whether telomerase activity is elevated or lowered in states of chronic stress and depression.

Potentially damaging mediatorsPotentially protective mediators
• Hyperactive LHPA axis and hypercortisolemia (with net hypercortisolism or hypocortisolism)• Neurosteroids (eg, DHEA* and allopregnanolone)
• Insulin sensitivity
• Intracellular glucose
• Synaptic glutamate and excitotoxicity• Antioxidants
• Anti-inflammatory/immuno-modulatory cytokines**
• Intracytoplasmic calcium
• Free radicals with oxidative stress• Neurotrophic factors (eg, BDNF)
• Inflammatory cytokines• Telomerase***


Psychological stress and individual differences

Psychological stress is frequently a precipitant of depressive episodes,19 and under certain circumstances it can initiate the biochemical cascade described here.7,8,10,13,16,20 It is apparent, though, that individuals respond very differently to stress, due, in part, to differences in coping strategies, disposition, temperament, and cognitive attributional styles.21-23 These can moderate stress-associated biological changes such as LHPA axis arousal,23 inflammation,22,24 neurogenesis,25 amygdala arousal,26 and cell aging. In the first study examining a personality trait and telomere length, O'Donovan et al found that pessimism was related to shorter telomere length, as well as higher IL-6 concentrations.22 In a study of the effects of early-life parental loss on later-life depression, the quality of the family and home's adaptation to the loss was the single most power-ful predictor of adult psychopathology, and was more important than the loss itself.27 Biochemical aspects of resilience vs stress vulnerability will not be covered here but have recently been reviewed.28

Adverse childhood events

Alexander Pope noted in 1734, that “as the twig is bent, the tree is inclined.” A rapidly expanding body of evidence suggests that early-life adversity (such as parental loss, neglect, and abuse) predisposes to adult depression27,29 as well as to LHPA axis hyper-reactivity to stress,27,30 increased allostatic load,13,31 diminished hippocampal volume (although this is controversial),32 lower brain serotonin transporter binding potential,33 and a myriad of adult physical diseases.34 Childhood adversity also predisposes to alterations in many of the mediators presented in our model of stress/depression/illness/cell aging, such as: inflammation,35,36 oxidative stress,37 neurotrophic factors,38 neurosteroids,39 glucose/insulin/ insulin-like growth factor (IGF-1) regulation,40 telomerase activity,41 and telomere length.36,42-44 Alterations in LHPA axis activity (increased or decreased) have been well described in victims of childhood adversity, even when the individuals are not currently depressed.27,30 In fact, several instances of neurobiological changes reported in MDD may be more attributable to histories of early-life adversity,30,32 which are over-represented among individuals with MDD, than to the MDD itself. Thus, early-life adversity seems capable of “reprogramming” the individual to a certain lifetime repertoire of altered physiological responses to stress and to vulnerability to psychiatric and physical illness. This reprogramming toward stress arousal and preparedness may be adaptive when the individual is likely to be confronted with a lifetime of continuous adversity, but is clearly disadvantageous otherwise. The causes of early adversity-induced behavioral and biochemical changes, and the explanation for the very long-lasting effects of such adversity, are the subject of intense investigation. One explanation that has attracted much attention is epigenetic changes,45,46 discussed in the next section.

Genetic and epigenetic moderators

A number of variants in candidate genes have been implicated in contributing to maladaptive and resilient responses that underlie alterations in neuronal plasticity and subsequent behavioral depression.47 Evidence is strongest for genes involved in HPA regulation and stress (corticotrophin-releasing hormone [CRH]1; glucocorticoid receptor [GR]), regulatory neurotransmitters, transporters, and receptors (serotonin (5-HT)1A, 5HT2, 5-HTTLPR, NET), neurotrophic factors, (brain-derived neurotrophic factor [BDNF], nuclear factor-kappaB, mitogen-activated protein kinase-1) and transcription factors (cAMP response element binding, Re-1 silencing transcription factor, delta FosB), but variations in other secondary modulatory factors (g-aminobutyric acid [GABA], catechol-O-methyl transferase, monoamine oxidase, dynorphin, neuropeptide-Y) have also been hypothesized to be important in determining individual differences in stress response.48 Studies of the CRH-1 gene in humans, for example, have shown that specific variants are associated with differential hormonal responses to stress, and with differing rates of depression and suicidal behavior.49 Increasingly, such genetic effects have themselves been found to be modulated by individual variation in environmental context and history (gene x environment, GxE).45 Epigenetics, which focuses on nongenomic alterations of gene expression, provides a mechanism for understanding such findings, through alteration of DNA methylation and subsequent silencing of gene expression or through physical changes in DNA packaging into histones.50 A comprehensive review of this literature is beyond the scope of this article, but the findings of selective recent studies in these areas are illustrative of the regulatory complexity that influences the possible translation of stressful experiences into depression. Long-lasting epigenetic effects of early life experience on hypothalamicpituitary-adrenal (HPA) responses have been demonstrated most clearly in animal models of differential maternal behavior and social isolation.45,46 Augmented maternal care was associated with reduced hypothalamic response to stress in rat pups and altered expression of CRH into adulthood.51 Suggestive human data compatible with these mechanisms have been reported.28,52 Oberlander et al,53 for example, found that prenatal exposure to third trimester maternal depression was associated with increased methylation of the glucocorticoid receptor gene at 3 months of age in the newborn child, while McGowan45 reported decreased levels of GR expression in the hippocampus of suicide victims with a history of childhood abuse, in comparison with those without such history and to controls. Tyrka and colleagues54 have also shown that variants in the CRH1 receptor gene appear to interact with a history of childhood abuse in determining cortical response to CRH. A separate body of research has focused on genetic investigations in components of serotonergic function, most commonly on a variant in the serotonin promoter (5HTTLPR), and, to a lesser extent, on serotonin receptor genes.55 In a small-scale study that remains controversial, Caspi et al56 reported that the effect of a variant in 5HTTLPR on increasing risk of depression was dependent upon a history of previous life stresses; several large-scale attempts at replication failed to support these conclusions and subsequent meta-analyses have been both positive and negative.57,58 Ressler et al59 have suggested that gene x gene x environment interactions may be involved, and reported that 5-HTTLPR alleles interacted with CRH1 haplotypes and child abuse history in predicting depressive symptoms. Others, however, have found it hard to demonstrate such effects.55 Yet another example of a potential GxGxE interaction was found in a study by Kauffman et al60 of child abuse victims, in whom BDNF and 5-HTTLPR genotypes interacted with maltreatment history in predicting depression, with social support showing some moderating influence. Despite the persuasive empirical animal data, the clinical relevance of epigenetic effects of stress on human emotional behavior is yet to be convincingly established.

Biochemical mediators


Elevated circulating GC levels are often observed in depressed individuals (especially in those with severe, melancholic, psychotic, or inpatient depressions), although considerable variability exists between studies, between individuals, and even within individuals over time, and some individuals are even hypocortisolemic.61,62 The physiological significance of increased circulating GC levels remains unknown, and it is debatable whether hypercortisolemia results in hypercortisolism at the cellular level, or, rather, in hypocortisolism, perhaps due to downregulation of the GR (often referred to as “GC resistance”).20 Thus, determination of “net” GC activity in depressed individuals at the intracellular level has remained elusive.63,64 In fact, different subclasses of depressed individuals may show opposite patterns of limbic-hypothalamic-pituitary-adrenal (LHPA) axis activity,15 and levels of LHPA activation may be more related to individual depressive symptoms than to the depressive syndrome per se.65 Further, it is possible that both hypo- and hypercortisolism are related to depression, in an inverted-U shaped manner.62 Complicating our understanding of this issue, novel treatment strategies that decrease or increase GC activity may show antidepressant effects in certain patients.66-69 The “hypocortisolism” hypothesis is supported by findings that proinflammatory cytokine levels (eg, tumor necrosis factor [TNF]-a, interleukin [IL]-1ß and IL-6) tend to be increased in the serum of depressed patients, and that proinflammatory cytokines may contribute to depressive symptomatology. Since cortisol typically has anti-inflammatory actions and suppresses proinflammatory cytokines (although there are instances to the contrary [eg, ref 70]), the coexistence of elevated cortisol and elevated proinflammatory cytokine levels suggests an insensitivity to cortisol at the level of the lymphocyte GR.20 Further supporting this notion, inflammatory cytokines downregulate GRs.20 Also, antidepressants typically increase GR binding activity,20 although in so doing, negative feedback onto the HPA axis is increased.71 On the other hand, the “hypercortisolism” hypothesis is supported by certain phenotypic somatic features suggestive of cortisol excess and end-organ cortisol receptor overactivation in some individuals with depression, eg, osteoporosis, insulin resistance, type 2 diabetes, a relative hypokalemic alkalosis accompanied by neutrophilia and lymphocytosis, hypertension, metabolic syndrome, and visceral/intra-abdominal adiposity.72,73 Further support of net GC overactivation is provided by evidence of altered expression of target genes such as BDNF, which are believed to be under negative regulatory control by cortisol.74

Pathologically elevated or diminished GC activity might have adverse neurobehavioral and physical health sequellae.72,75 Chronic hypercortisolemia, in particular, has been proposed by Sapolsky and others16 to result in a biochemical “cascade,” which can culminate in cell endangerment or cell death in certain cells, including cells in the hippocampus. In the simplest description of this model, GC excess engenders a state of intracellular glucoprivation (insufficient intracellular glucose energy stores) in certain cells, impairing the ability of glia and other cells to clear synaptic glutamate. The resulting excitotoxicity results in excessive influx and release of calcium into the cytoplasm, which contributes to oxidative damage, proteolysis, and cytoskeletal damage. Unchecked, these processes can culminate in diminished cell viability or cell death.


Although cortisol concentrations are often reported as elevated in depression, CSF concentrations of the potent GABA-A receptor agonist neurosteroid, allopregnanolone, are decreased in unmedicated depressives, and CSF levels of allopregnanolone increase with treatment in direct proportion to the antidepressant effect.76 Selective serotonin reuptake inhibitor (SSRI) antidepressants rapidly increase allopregnanolone synthesis, and this may contribute to their anxiolytic effects.77,78 Another neurosteroid, DHEA, which may have “anticortisol” effects, has been reported to be both high and low in depression.17 Notably, both of these neurosteroids modulate HPA axis activity17,18 and immune system activity,17,79 antagonize oxidative stress17,80 and have certain neuroprotective effects.17,81 Depressed patients entering remission show decreases in plasma cortisol concentrations along with increases in plasma allopregnanolone concentrations.82 Endogenous decreases in this neurosteroid concentrations or exogenously produced increases in their concentrations might be expected to have damaging or beneficial effects, respectively, in the context of depression,17,78,83,84 and treatment trials have demonstrated significant antidepressant effects of exogenously administered DHEA.17 Animal models suggest that 3 a hydroxy-5 a reduced steroids (allopregnanolone and allotetrahydrodeoxycorticosterone) are responsive to stress85 and may function to restore normal g-aminobutyric acid (GABA)-ergic and hypothalamicpituitary-adrenal function following stress.18,85 In vitro, allopregnanolone suppresses release of gonadotropinreleasing hormone86 or CRH87 via a GABA-A mediated mechanism. Allopregnanolone or allotetrahydrodeoxycorticosterone can also attenuate stress-induced increases in plasma ACTH and corticosterone and can affect arginine vasopression transcription in the hypothalamus (paraventricular nucleus).18 Under chronic stress or in psychiatric disorders, dysregulation of the HPA axis could be exacerbated if there is insufficient activity of these “counter-regulatory” neurosteroids. In addition to protection against acute or chronic stress, neurosteroids such as allopregnanolone and allotetrahydrodeoxycorticosterone may be neuroprotective against early life stressors88 or against deleterious effects of social isolation.89 In this way, these neurosteroids may be neuroprotective during development and may affect future responsiveness to stress.

The detrimental effects of neurosteroid dysregulation on stress responses has been particularly documented in women with premenstrual dysphoric disorder (PMDD).90,91 PMDD is a depressive disorder that is characterized by cyclic recurrence, during the luteal phase of the menstrual cycle, of a variety of physical and emotional symptoms that are so severe as to interfere with daily activities. In these studies in women with PMDD, both high and low concentrations of allopregnanolone during the luteal phase of the menstrual have been reported. However, women with PMDD have reduced responsiveness to neurosteroids (on GABA-A receptors)92 as well as a blunted stress response (failing to demonstrate an increase in allopregnanolone concentrations after acute stress) in women with PMDD and a prior history of depression.90 Furthermore, women with PMDD who also had prior histories of depression showed significant decreases in allopregnanolone after acute stress.90 These data highlight that long-term histories of depression may be associated with persistent, long-term effects on the responsivity of the neurosteroid system, as well as long-term effects on modulation of the HPA axis following stress.

Glucose and insulin regulation

Abnormalities of glucose homeostasis (eg, insulin resistance and impaired glucose tolerance) are seen in MDD, even in individuals who are nonobese and not diabetic.93 These glucose and insulin abnormalities are most pronounced in hypercortisolemic depressed individuals,94 as would be predicted based on cortisol's well-known antiinsulin effects. Hypercortisolemic depressives, compared with normocortisolemic ones, are also at increased risk of having increased abdominal (visceral) fat deposition95 and the metabolic syndrome,96 which are also risk factors for cardiovascular disease. Insulin resistance and diminished cellular glucose uptake can also lead to a dangerous “energetic crisis.”7,16 When this occurs in the hippocampus,16 for example, hippocampal excitotoxicity may develop, since there is insufficient energy available to clear glutamate from the synapse. Thereafter, cytosolic calcium is mobilized, triggering oxygen free radical formation and cytoskeletal proteolysis. The relevance of this in humans was demonstrated in a PET scan study, in which cortisol administration to normal individuals resulted in significant reductions in hippocampal glucose utilization.97 The importance of hippocampal insulin resistance for depression and cognitive disorders (eg, Alzheimer's disease) is the subject of active investigation.98,99

Over and above these direct effects on energy balance, prolonged exposure to glucose intolerance and insulin resistance is associated with accelerated biological aging7,100 including shortened telomere length,101 and visceral adiposity is associated with increased inflammation and oxidation,102,103 both of which, themselves, promote accelerated biological aging.7 These will be further discussed below in the sections on inflammation, oxidation, and cell aging.

Immune function

Dysregulation of the LHPA axis contributes to immune dysregulation in depression, and immune dysregulation, in turn, can activate the HPA axis and precipitate depressive symptoms.20 Immune dysregulation may be an important pathway by which depression heightens the risk of serious medical comorbidity.7,104,105 Several major proinflammatory cytokines, such as IL-1ß, IL-2, IL-6 and TNF-a, are elevated in depression, either basally or in response to mitogen stimulation or acute stress.20,106,107 Conversely, certain anti-inflammatory or immunomodulatory cytokines, such as IL-1 receptor antagonist and IL10 may be decreased or dysregulated.106 Indeed, the ratio of proinflammatory to anti-inflammatory/ immunomodulatory cytokines may be disturbed in depression and could result in net increased inflammatory activity106 as well as in oxidative stress.108 Converging findings suggest that high peripheral levels of inflammatory cytokines, such as IL-6, are associated with the activation of central inflammatory mechanisms that can adversely affect the hippocampus, where IL-6 receptors are abundantly expressed.109 High proinflammatory cytokine levels, for example, may directly contribute to depression, decreased neurotrophic support, and altered glutamate release/reuptake and hippocampal neurodegeneration,110 and, plasma IL-6 levels are inversely correlated with hippocampal gray matter in healthy humans.111 Further, inappropriately and chronically elevated proinflammatory cytokines can contribute to accelerated biological aging (eg, premature shortening of immune cell telomeres112). Interestingly, the development of immunosenescence (eg, the loss of the CD28 marker from CD8+ T cells), can further aggravate the proinflammatory milieu, since CD8+CD28- cells hypersecrete IL-6.113 It should be noted, however, that due to the complexity of cytokine actions in neurons and glia, the end effect of individual cytokines may be either detrimental or protective, depending on the circumstances.106


Stress and increased LHPA axis activity can also increase oxidative stress and decrease antioxidant defenses.5 ,7,114 Oxidative stress often increases with aging and various disease states, while antioxidant and antiinflammatory activities decrease, resulting in a heightened likelihood of cellular damage and of a senescent phenotype.7,115 The co-occurrence of oxidative stress and inflammation (the so-called “evil twins” of brain aging115), as may be seen in depression, post-traumatic stress disorder (PTSD), stroke, Alzheimer's disease, and others, can be especially detrimental. Oxidative stress occurs when the production of oxygen free radicals (and other oxidized molecules) exceeds the capacity of the body's antioxidants to neutralize them. Oxidative stress damages DNA, protein, lipids, and other macromolecules in many tissues, with telomeres (discussed below) and the brain being particularly sensitive. Elevated plasma and/or urine oxidative stress markers (eg, increased F2-isoprostanes and 8-hydroxydeoxyguanosine [8-OHdG], along with decreased antioxidant compounds, such as Vitamin C, Vitamin E, and Coenzyme Q) have been reported in individuals with depression and in those with chronic psychological stress, and the concentration of peripheral oxidative stress markers is positively correlated with the severity and chronicity of depression.114,116 Further, the ratio of serum oxidized lipids (F2-isoprostanes) to antioxidants (Vitamin E) is directly related to psychological stress.8 Importantly, this ratio (and the ratio of F2-isoprostanes to another antioxidant, Vitamin C) is inversely related to telomere length in chronically stressed caregivers8 and in individuals with major depression.117 Oxidative stress markers are also correlated with decreased telomerase activity.118 Further, diminished levels of antioxidants reportedly lower BDNF activity.119 Interestingly, antidepressants decrease oxidative stress.120 Since cellular oxidative damage may be an important component of the aging process, prolonged or repeated exposure to oxidative stress might accelerate aspects of biological aging and promote the development of aging-related diseases in depressed individuals.114 It is unknown whether antioxidant treatment would retard stress- or depressionrelated aging; this is discussed below under “novel treatment implications.”

Brain-derived neurotrophic factor

The “neurotrophic model” of depression74 emphasizes the centrality of neurogenesis and neuronal plasticity in the pathophysiology of depression. It posits that diminished hippocampal BDNF activity, caused by stress or excessive GCs, impairs the ability of stem cells in the subgranular zone of the dentate gyrus (as well as cells in the subventricular zone, projecting to the prefrontal cortex) to remain viable and to proliferate into mature cells. It is not known whether such effects can cause depression, but they may be relevant to the mechanism of action of antidepressant treatments.121 Unmedicated patients with depression have decreased hippocampal (at autopsy) and serum concentrations of BDNF121,122 Over 20 studies have documented decreased serum concentrations of BDNF in unmedicated depressed individuals; this is now one of the most consistently replicated biochemical findings in major depression.121,123 Further, serum BDNF concentrations increase with antidepressant treatment.121,123 The relationship of peripheral BDNF concentrations to central ones is not known, but even peripherally administered BDNF abrogates depressive and anxiety-like behaviors and increases hippocampal neurogenesis in mice, suggesting that serum BDNF concentrations are functionally significant for brain function and are more than merely a biomarker.124 A role of BDNF in antidepressant mechanisms of action is supported by findings that hippocampal neurogenesis (in animals) and serum BDNF concentrations (in depressed humans) increase with antidepressant treatment,121,123 and that hippocampal neurogenesis and intact BDNF expression are required for behavioral effects of antidepressants in animals.125,126

Apart from its direct neurotrophic actions, BDNF also has anti-inflammatory and antioxidant effects that may contribute to its neuroprotective efficacy,127 and BDNF, in concert with telomerase (discussed below) promotes the growth of developing neurons.128 In addition, BDNF (despite its name) has significant peripheral actions that are important for physical health, and the low levels of BDNF seen in MDD may be involved in certain comorbid illnesses such as cardiovascular disease, diabetes, obesity, and metabolic syndrome.129 For example, BDNF improves glucose and lipid profiles, enhances glucose utilization, suppresses food intake, has an insulinotropic effect and protects cells in the islets of Langerhans (reviewed in ref 129). Plasma levels of BDNF are low in type 2 diabetes and are inversely correlated with fasting glucose levels.129 Indeed, BDNF is increasingly considered not only a neurotrophin but a metabotrophin,129 and its dysregulation has been proposed as a unifying feature of several clustered conditions, such as MDD, Alzheimer's disease, and diabetes.130

Cell aging: telomeres and telomerase

Telomeres are DNA-protein complexes that cap the ends of linear DNA strands, protecting DNA from damage.131 When telomeres reach a critically short length, as may happen when cells undergo repeated mitotic divisions in the absence of adequate telomerase (eg, immune cells and stem cells, including neurogenic stem cells in the hippocampus), cells become susceptible to apoptosis and death. Even in nondividing cells, such as mature neurons, telomeres can become shortened by oxidative stress, which preferentially damages telomeres to a greater extent than nontelomeric DNA. This nonmitotic type of telomere shortening also increases susceptibly to apoptosis and cell death. Telomere length is a robust indicator of “biological age” (as opposed to just chronological age) and may represent a cumulative log of the number of cell divisions and a cumulative record of exposure to genotoxic and cytotoxic processes such as oxidation.7-9,113,131,132 Telomere length may also represent a biomarker for assessing an individual's cumulative exposure to, or ability to cope with, depression or stressful conditions. For example, chronically stressed8,9 or depressed133-135 individuals show premature leukocyte telomere shortening, a sign of cellular aging. In the former study, telomere length was inversely correlated with perceived stress and with cumulative duration of caregiving stress.8 The estimated magnitude of the acceleration of biological aging in these studies was not trivial; it was estimated as approximately 9 to 17 additional years of chronological aging in the stressed caregivers and approximately 6 to 10 years in the depressed individuals. Preliminary data from our group suggest that telomere loss in MDD is most apparent in those individuals with more chronic courses of depression,117 but another study did not observe that.135 Interestingly, individuals with histories of early-life adversity or abuse also have shortened leukocyte telomeres.36,42,43 Since individuals with MDD are more likely to have experienced earlylife adversity, it remains to be determined how much of the telomere shortening seen in studies of MDD relate to the MDD per se vs the histories of early-life adversity. In individuals with post-traumatic stress disorder (PTSD), telomere shortening was more closely linked to adverse childhood events than to the PTSD per se.44

The importance of accelerated telomere shortening for understanding comorbid medical illnesses and premature mortality in depressed individuals is highlighted by multiple studies in nondepressed populations showing significantly increased medical morbidity and earlier mortality in those with shortened telomeres.7,136 For example, shortened leukocyte telomeres are associated with a greater than 3-fold increase in the risk of myocardial infarction and stroke and with a greater than 8-fold increase in the risk of death from infectious disease.137 Thus, cell aging (as manifest by shortened telomeres), may provide a conceptual link between depression and its associated medical comorbidities and shortened life span.7,104,132 The causes of accelerated telomere loss in MDD are not known, but they may include chronic exposure to inflammation and oxidation, both of which are commonly seen in MDD and both of which are associated with telomere shortening. In our own studies, telomere length in MDD was inversely correlated with inflammation (IL-6 concentrations) and oxidative stress (the F2-isoprostane/ Vitamin C ratio).117

Telomere length is determined by the balance between telomere shortening stimuli (eg, mitotic divisions and exposure to inflammation and oxidation) and telomere lengthening or reparative stimuli. A major enzyme responsible for protecting, repairing, and lengthening telomeres is telomerase, a ribonucleoproptein enzyme that elongates telomeres, thereby counteracting telomere shortening and maintaining cellular viability.131 Telomerase may also have antiaging or cell survival-promoting effects independent of its effects on telomere length by regulating transcription of growth factors, synergizing with the neurotrophic effects of BDNF, having antioxidant effects and intrinsic antiapoptotic effects, protecting cells from necrosis, and stimulating cell growth in adverse conditions (eg, ref 128). In one study in which telomere shortening was observed, telomerase activity

was significantly diminished in stressed (generally nondepressed) caregivers8 but, in another caregiver study (in which caregivers were more depressed than controls), telomerase activity was significantly increased.138 We recently found that telomerase activity was significantly increased in unmedicated depressed individuals.139 It is possible that increased telomerase activity, in the face of shortened telomeres, is an attempted compensatory response to telomere shortening.138,139

Pointing to the inter-relatedness of several of the mediators considered in this review, telomerase activity can be down-regulated by cortisol,140 tumor necrosis factor (TNF)-a and certain growth factors, and upregulated by IL-6 and certain other inflammatory cytokines, insulinlike growth factor-1, fibroblast growth factor-2, vascular endothelial growth factor, estrogen, and others.141

Novel treatment implications

To the extent the biochemical mediators we have described are pathophysiologically involved in MDD and its medical comorbidities, new classes of treatments should be considered, and certain noncanonical mechanisms of action of traditional antidepressants should be emphasized in new drug development. Some of these novel approaches are already under investigation, while others remain to be tested. In Table II, we list certain traditional and nontraditional, but mechanism-based, interventions that may ameliorate the biochemical mediators we have discussed. These interventions range from purely behavioral (eg, exercise and improved fitness, environmental enrichment, yoga and meditation, dietary macronutrient modifications and calorie restriction) (see refs 7,142-144 for description of these behavioral approaches) to more purely medication-based (see ref 145 for additional descriptions of novel biological mechanism-based therapeutics). For example, early work suggests the promise, at least in certain patients, of antiglucocorticoids,67-69 DHEA supplementation,17 insulin receptor sensitizers,99,146 glutamate antagonists,147 calcium blockers,148 anti-inflammatories,149 antioxidants,150 increased BDNF delivery to the brain, 124,151 and, most speculatively, telomerase enhancers.152,153

Potential mechanism-based therapeutic interventions. LHPA, limbic-hypothalamic-pituitary-adrenal; GC, glucocorticoid; GR, glucocorticoid receptor; CRH, corticotrophin-releasing hormone; DHEA, dehydroepiandrosterone; BDNF brain-derived neurotrophic factor; TNF, tumor necrosis factor; SSRI, selective serotonin reuptake inhibitor

Biochemical mediatorPotential treatment interventions
Stress vulnerabilityStress reduction; meditation; lifestyle changes7,142,154
Epigenetic changesEpigenetic reprogramming155,156
LHPA axis dysregulationAntidepressants upregulate GR function71
(Hypercortisolemia + GC resistance)CRH antagonists157
Cortisol antagonists and GR antagonists or agonists66,69
Glucose/insulin dysregulationInsulin receptor sensitizers99,146
Glutamate/excitotoxicityGlutamate antagonists147
Oxidative stressAntidepressants have antioxidant effects120
Intracellular calciumCalcium blockers148
InflammationAntidepressants have anti-inflammatory effects20
Anti-inflammatory drugs, TNF-a antagonists, etc149
Decreased counter-regulatory neurosteroids SSRIs increase allopregnanolone synthesis77,78
DHEA administration17
Decreased BDNFAntidepressants (esp SSRIs) increase BDNF concentrations122,123
Environmental enrichment143,144
Dietary restriction143
BDNF administration via novel routes or vectors124,151
Cell aging (telomeres; telomerase)Telomerase activation152,153

Summary: is depression accompanied by accelerated aging?

We began this review article by noting that depressed individuals are at increased risk of developing physical illnesses more commonly seen with aging. It remains unknown whether MDD and these medical conditions are causally related. This determination will be important in considering whether primary treatment of the depression (eg, with antidepressant medications or psychotherapy) should additionally treat some of the medical comorbidities (and vice versa) or whether the biochemical mediators that are common to both conditions (eg, inflammation and oxidation) should be a primary treatment focus. We also discussed the potent influence that early-life adversity can have on the subsequent development of depression and medical comorbidities. We noted that many of the biochemical mediators are linked to others, and that there are many examples of bidirectional influence. Finally, we postulated that certain of these mediators have the potential to accelerate cellular aging at the level of DNA. In any event, is important to recognize that MDD may be biologically heterogeneous, and this model may apply only to certain subsets of patients with MDD. This reconceptualization of MDD as a constellation of biochemical features conducive to physical as well as mental distress places MDD firmly in the taxonomy of physical disease and points to new types of treatment.