Although dysthymia, a chronic, low-grade form of depression, has a morbidity rate as high as that of major depression, and increases the risk for major depressive disorder, limited information is available concerning the etiology of this illness. In the present report we review literature concerning the biological and characterological features of dysthymia, the effectiveness of antidepressant treatments, the influence of stressors in the precipitation and maintenance of the disorder, and both quality of life and psychosocial correlates of the illness. We also provisionally suggest that dysthymia may stem from disturbances of neuroendocrine and neurotransmitter functioning (eg, corticotropin releasing hormone and arginine vasopressin within the hypothalamus, or alternatively monoamine variations within several extrahypothalamic sites), and may also involve cytokine activation. The central disturbances may reflect phenotypic variations of neuroendocrine processes or sensitization of such mechanisms. It is suggested that chronic stressor experiences or stressors encountered early in life lead to the phenotypic neurochemical alterations, which then favor the development of the dysthymic state. Owing to the persistence of the neurochemical disturbances, vulnerability to double depression is increased, and in this instance treatment with antidepressants may attenuate the symptoms of major depression but not those of the basal dysthymic state. Moreover, the residual features of depression following treatment may be indicative of underlying neurochemical disturbances, and may also serve to increase the probability of illness recurrence or relapse.
Dysthymia, a chronic, low-grade form of depression, occurs in a substantial portion of the population, and increases the risk for major depressive disorder. Yet, relative to major depression, limited information is available concerning the behavioral concomitants, as well as the physiological correlates of dysthymia. Dysthymia and major depression share several features regarding stress/coping, and the response to pharmacotherapy. However, it appears that they can be distinguished from one another with respect to hypothalamic-pituitary-adrenal (HPA) functioning, and there is evidence that these depressive subtypes can be differentiated with regard to their cytokine correlates. Yet, there is reason to suppose that, owing to its chronic nature, dysthymia may be associated with persistent functional changes of HPA activity, as well as several adjunctive features, including altered stressor and uplift perceptions, coping styles, and quality of life.1, 2, 3, 4 Together, these factors may perpetuate the illness, promote relapse following treatment, and increase the risk for superimposed major depression (double depression).
The broad purpose of the present review is to elucidate several behavioral, neuroendocrine and immune/ cytokine characteristics of dysthymia. To this end, we provide an overview of the characteristics of dysthymia, including a description of the clinical and epidemiological aspects of the illness, comorbid features of dysthymia, as well as a review of the data suggesting a role for genetic factors. Given that dysthymia, or at least some subtypes of the illness, may involve biological underpinnings, a brief review is provided regarding the neuroendocrine, neurochemical and cytokine correlates of dysthymia, and an overview is provided concerning the efficacy of pharmacotherapy in the treatment of this disorder.
It is proposed that dysthymia may be related to subtle effects of stressors and inadequate coping styles, and may be exacerbated by the presence of ongoing psychosocial impairments. A highly provisional model is proposed concerning the etiological processes subserving dysthymia, including various facets of the HPA axis (eg, phenotypic variations of corticotropin releasing hormone (CRH) and arginine vasopressin (AVP), down-regulation of adrenal functioning) and forebrain serotonergic mechanisms. However, given the paucity of data concerning the effects of various challenges (eg, dexamethasone, ACTH, CRH, TSH, stressors, as well as challenges in the presence or absence of metyrapone) on HPA hormones among dysthymic individuals, the conclusions that can be derived are tentative and must necessarily await further data.
Dysthymia: clinical and epidemiological features
Dysthymia, literally meaning ‘being of bad mood’ or ‘ill-humor’ is an illness characterized by a number of affective, neurovegetative and cognitive symptoms. The diagnosis of dysthymic disorder was introduced in the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III) to characterize chronic depression of 2 or more years and to encompass disorders which had previously been considered characterologically based, including neurotic depression, chronic minor depression, and characterological depression.5 While the severity of dysthymia is usually less profound than that of acute major depressive disorder, symptoms may fluctuate in intensity. Furthermore, several subtypes of dysthymia have been proposed based on specific symptoms, family history, and age of onset; subaffective dysthymia is thought to be of biological origin, while character spectrum dysthymia is more personality-based.6, 7 Currently, DSM-IV stipulates that a diagnosis of dysthymia includes depressed mood, coupled with two or more of the following: poor appetite or overeating, insomnia or hypersomnia, low energy/fatigue, low self-esteem, poor concentration or difficulty making decisions, and feelings of hopelessness.5 Dysthymia also frequently has an early and insidious onset, and is associated with pathology of character, albeit these may not play an etiological role. There has been some debate as to whether these symptoms are, in fact, most characteristic of dysthymia, and based on field trials, the DSM-IV (Appendix B) offers an alternative set of criteria. These include: (a) low self-esteem, self confidence, or feelings of inadequacy; (b) feelings of pessimism, despair, and hopelessness; (c) anhedonia (generalized loss of interest or pleasure); (d) social withdrawal; (e) chronic fatigue or tiredness; (f) feelings of guilt, or brooding about the past; (g) irritability or excessive anger; (h) reduced activity, effectiveness, or productivity; and (i) difficulty in thinking, as reflected by poor concentration, memory or decisiveness.
Like the DSM-IV, the ICD-10 defines dysthymia as a chronic depression which fails to meet the severity and duration criteria for recurrent depressive disorder. However, depressive illness, of at least mild duration, may have occurred previously.8 According to the ICD-10, dysthymia often begins early in adult life, and when late onset dysthymia occurs it is often secondary to a severe major depressive disorder or in response to environmental stressors (eg, bereavement). Moreover, dysthymia is thought to include persistent anxiety-depression, depressive neurosis, depressive personality disorder, as well as neurotic depression of more than 2 years duration.7
The characteristics of dysthymia in many ways overlap with those of major depression, although in dysthymia, symptoms outnumber signs (ie, objective characteristics such as vegetative symptoms and psychomotor changes are typically absent).9 Relative to major depressive disorder, symptoms occur at a lower frequency in dysthymic patients, but are qualitatively similar. Moreover, relative to neurovegetative and psychomotor features, social-motivational impairments tend to be more characteristic of dysthymia.3, 10 In many dysthymic patients an intermittent emergence of major depression may occur (double depression), with the dysthymic state usually recurring upon remission of the major depressive episode.9 Likewise, dysthymia may emerge as a residual syndrome of acute depression;6, 11 however, the diagnosis of dysthymia would not be applied if the initial episode of chronic depressive symptoms met the criteria for and was sufficiently severe to be diagnosed as a major depressive episode.
Dysthymia is more common among women than among men (2:1), has a 1-year prevalence rate (for the United States) estimated as high as 5.4%,12 and the highest life-time prevalence of the affective disorders.13 However, only a relatively small proportion of dysthymic individuals seek treatment for their illness, likely owing to the mild nature of the symptoms and their insidious onset, coupled with the individual's lack of insight. Indeed, the illness often appears in childhood and adolescence,14, 15 and as a result of the long-term, low-grade nature of the illness, dysthymia might not be perceived as differing from the individual's norm. It is of interest to note that in a study of a fairly large number of dysthymic patients, only 41% had received any form of pharmacotherapy, and just 56% had received psychotherapy, attesting to the under-treatment of this disorder.16
Although it had been suggested that pharmacotherapy may be less effective in the treatment of dysthymia relative to major depressive disorder, recent evidence has indicated that such intervention may, in fact, be a treatment of choice for dysthymia. However, the treatment response may vary with the specific subtype of the illness.4, 17 Indeed, it appeared that pharmacotherapy was less effective in ‘pure dysthymia’ than in dysthymia with a history of major depression or in patients with concurrent major depression.18 Because dysthymia was initially thought to be more of a characterological disturbance than a biologically-based illness, and because of the oft-noted superiority of pharmacotherapy in major depression relative to that observed in dysthymia,17 the illness was typically treated using different forms of psychotherapy.19, 20, 21 As such, the lack of controlled studies comparing the relative efficacy of psychotherapy and pharmacotherapy (either alone or in combination) is surprising. It was recently reported, however, that in the short-run, pharmacotherapy (using a selective serotonin reuptake inhibitor) was more efficacious than group cognitive behavior therapy (CBT) in alleviating symptoms of dysthymia,22 although CBT attenuated several functional aspects of the illness. Using a somewhat more protracted regimen (16 weeks), CBT was effective in a study involving a small number of dysthymic subjects. While this effect appeared to be somewhat reduced relative to that associated with fluoxetine, the effects of the treatments were not significantly different.23
Symptom and illness comorbidity
The presence of comorbid features with dysthymia can be related to any number of factors. On the one hand, the comorbidity may simply reflect the nosology of the syndromes, which have overlapping symptoms.1 On the other hand, comorbidity may reflect common underlying mechanisms, or in the case of debilitating medical conditions, the dysthymic state may represent a subsyndromal depression resulting from the primary illness. It is possible, as well, that the development of dysthymia may be secondary to features such as personality disorders or to excessive anxiety, or conversely, dysthymia may give rise to these features. Whatever the case, it is of obvious diagnostic and therapeutic value to discern the presence and progression of comorbid features.1
Major depression is often superimposed on a dysthymic state (double depression), and is associated with a high rate of illness recurrence.24, 25 In their epidemiological study, Weissman et al26 reported that 40% of dysthymic patients exhibited comorbid major depression, while more recent evidence suggested that as many as 62% of dysthymic patients met criteria for current major depression, and 80% for lifetime major depression.14 Treatment outcome was significantly better in major depressives than among double depressives, as was the recurrence rate over 2 years.25 It has been suggested that the chronology of dysthymia relative to major depression may influence subsequent treatment response. For example, it was reported that patients who had experienced dysthymia subsequent to their first major depressive episode, showed a reduced treatment response relative to those patients in whom onset of dysthymia had preceded their first major depression.27
As many as 75% of dysthymic patients suffer from some comorbid psychiatric disorder, of which depression, anxiety, and substance abuse are the most common.26 Indeed, approximately one-half (50.7%) of Shelton et al’s16 sample of 410 dysthymic patients reported a history of major depression, while 26.3% had a history of substance abuse, and 68.2% were diagnosed with a comorbid personality disorder. With respect to comorbid anxiety, it was demonstrated that while social phobia and panic attacks were significantly more common among unipolar and bipolar depressives (respectively) than dysthymic patients, the latter exhibited a higher prevalence of generalized anxiety disorder.28 As such, dysthymia ought to be more closely scrutinized for comorbid anxiety disorder, as it might be associated with a more severe and enduring symptom profile, as observed in major depression with co-existing anxiety disorder,29 and might thus require adjunctive forms of pharmacotherapy. In fact, it was recently reported that a coexisting anxiety disorder may indicate increased risk for persistent depression.30 In addition to comorbid anxiety, dysthymia often co-exists with personality disorders.31 As well, more double depressive patients than pure dysthymics met criteria for at least one personality disorder. Conversely, compared to episodic major depressives, significantly more dysthymics met criteria for an axis II disorder. Interestingly, dysthymic patients scored significantly higher than major depressive patients on all of the 13 dimensions of the Personality Disorder Examination, with the most common axis II disorders being borderline, avoidant and histrionic.31 Essentially, dysthymic subjects were found to display personality disorders in the DSM cluster B (antisocial, narcissistic, borderline and histrionic)31, 32 although high rates of avoidant and dependent personality disorders were also apparent.31, 32, 33 Indeed, it was reported that subaffective and character spectrum dysthymics could be distinguished from one another on the basis of DSM cluster C characteristics, with subaffective subjects exhibiting greater avoidant personality and dependent personality relative to character spectrum patients.34 It was further reported that the occurrence of such personality disorder comorbidity was particularly notable in early-onset dysthymic patients.35, 36 Finally, it appeared that depressive personality disorder (DPD) was more closely aligned with dysthymia than with major depression.37
As in the case of other depressive syndromes,38 substance abuse was found to occur at a fairly high rate among dysthymic patients. For instance, it has been reported that subjects in alcohol treatment programs frequently met the criteria for dysthymic disorder.39 In a sample of subjects with substance-related disorder, a substantial portion (˜10%) exhibited comorbid dysthymia.40, 41 The dysthymic group also had a surprisingly early age of first use of caffeine (7.3 ± 7.0 years), and the investigators speculated that this may have constituted an early attempt to self-medicate.40
In considering comorbid features, it may be important to distinguish between subtypes of dysthymia based on age of illness onset. It had been reported that early- and late-onset dysthymic patients with superimposed major depression did not differ with respect to several clinical, psychosocial and cognitive indices, nor in terms of family history of depression and alcoholism.42 Subsequent studies, however, indicated that early-onset dysthymics exhibited greater use of emotion-focused coping strategies, a higher frequency of personality disorders, and increased life-time substance abuse disorder than their late-onset counterparts.36, 43, 44, 45, 46, 47, 48, 49, 50 Further, a greater number of early-onset dysthymic patients had a family history of a mood disorder, and displayed earlier onset and longer duration of the index major depressive episode.36, 50 To be sure, as indicated by Klein et al,36 the earlier onset of the index major depressive episode might simply reflect greater opportunity for such an occurrence given that dysthymia increases the risk of major depression. Moreover, early onset dysthymia being associated with poor interpersonal and psychosocial skills, may have favored the development of personality disorders and substance abuse. Yet, there is reason to suppose that biological factors may be more closely aligned with early onset of the illness, while late onset may be associated with character spectrum disorder.4 In fact, as will be discussed shortly, there have been several reports indicating that these subgroups may differ with respect to several neuroendocrine substrates, as well as in response to some antidepressant medications.
In addition to the high psychiatric comorbidity, dysthymia has also been associated with various other medical conditions. Given the similarity between dysthymia and chronic fatigue syndrome (CFS), particularly with respect to lethargy, lassitude, impaired concentration, and diminished drive, the possibility exists that a subgroup of chronic fatigue patients represents a variant of dysthymia.51 Indeed, it was reported that patients identified with fatigue were between six and 10 times more likely to suffer from dysthymia than the population at large.52 Similarly, fibromyalgia syndrome, which overlaps with CFS, has been associated with major depression and dysthymia.53, 54 Interestingly, fibromyalgia syndrome has been associated with adverse childhood experiences. Moreover, specific personality traits which may be related to such adverse experiences (eg, unstable self-esteem), may also be associated with fibromyalgia, just as they are associated with dysthymia.54 It is unclear, however, whether dysthymia is associated simply with the symptom of chronic fatigue, as opposed to the syndrome.51 It is significant, as will be discussed shortly, that dysthymia is associated with elevated production of interleukin-1 (IL-1β),55 a cytokine released by activated macrophages, just as CFS has been associated with an elevation of this cytokine.56, 57
In addition to CFS, like other mood and anxiety-related disorders, dysthymia has also been associated with increased co-occurrence of migraine (with aura), tension, and non-organic headaches.58, 59, 60 In this respect, it was reported that of the mood disorders, the prevalence of coexisting dysthymia was particularly notable.60 These data raise the possibility that headache and dysthymia share common underlying processes (eg, serotonergic mechanisms, given that both are positively influenced by serotonin reuptake inhibitors) or that environmental triggers, such as stressors, may be associated with both pathologies.
Comorbidity involving CFS and headaches in dysthymia is not unexpected given the potential for common mechanisms or environmental precipitants. More surprising, however, was the finding that a large number of patients suffering from Parkinson's disease exhibited concurrent major depressive disorder and/or dysthymia. Of course, such comorbidity might be predicted given that the stress of the neurodegenerative disturbance might lead to depressive affect. Interestingly, however, in a recent review of this literature, depressive features, including introversion and inflexibility, were reported to serve as premorbid predictors of the onset of Parkinson's disease or an accelerated cognitive decline.61 These investigators have gone so far as to suggest that psychological and pharmacological interventions may be appropriate in the treatment of Parkinsonian depression.
As in the case of neurodegenerative disorders, patients suffering from traumatic brain injury (TBI) reportedly suffer from a higher prevalence of mood disorders than the general population (ie, 25–50% major depression, 15–30% dysthymia, 9% mania).62 Given the large number of brain regions that may be primarily or secondarily influenced by head trauma, it is difficult to identify the specific nuclei or pathways subserving the depression. However, since brain trauma is associated with focal increases of IL-1,63 as are neurodegenerative disorders, the possibility should be considered that the behavioral effects observed in TBI and Parkinsonian patients are associated with heightened cytokine levels.
Biological aspects of dysthymia
Depression has been associated with a variety of neurochemical changes, including deficiencies of norepinephrine (NE),64 serotonin (5-HT)65, 66 and dopamine (DA),67, 68, 69 or to variations of DA autoreceptors, 5-HT2 receptors, α1-NE or β-NE receptors.70, 71, 72 Of course, depression is likely a biochemically heterogeneous disorder, such that the neurochemical underpinnings for the illness, as well as the symptom profile exhibited, vary across subjects.66, 73 Moreover, given the abnormal responses to various endocrine challenges (eg, the dexamethasone suppression test (DST), CRH, as well as thyroid releasing hormone (TRH) challenges), there is reason to suppose that hormonal variations contribute to the provocation or expression of depressive symptoms, and that considerable interindividual variability exists with respect to the contribution of these endocrine factors.74, 75, 76, 77
It has been suggested that stressful events or failure experiences are associated with depressive illness.78, 79, 80 Such an outcome may stem from the stressor experience provoking the formation of attributions, which give rise to negative expectancies of future performance and may result in the development of cognitive disturbances, such as helplessness.81, 82 Alternatively, stressor experiences may give rise to neurochemical alterations that favor depressed mood.83, 84
It is difficult, using animals, to adequately model depressive illness let alone to reflect dysthymia. Nevertheless, it may be productive to examine some of the neurochemical correlates of stressors. Although environmental insults initiate a series of neurochemical changes that may be of adaptive significance, when these neurochemical alterations are insufficient to deal with environmental demands (or neurochemical adaptation does not occur readily), vulnerability to pathology is increased.83 Indeed, animal studies indicated that stressors will induce many of the central neurotransmitter alterations (eg, variations of NE, DA and 5-HT turnover and levels) that have been proposed to subserve the depressive symptoms in humans. Specifically, in response to acute stressors the increased utilization of NE, DA and 5-HT is ordinarily met by adequate synthesis and hence transmitter levels remain stable. However, under conditions that favor amine utilization exceeding synthesis (ie, if the stressor is sufficiently severe and uncontrollable), amine levels may decline in several brain regions.83, 84, 85, 86 These amine alterations typically persist for only a few hours, depending on the stressor severity and several organismic variables (eg, age, species).84, 87, 88 However, re-exposure to a mild stressor enhances the utilization of hypothalamic NE89, 90 and mesocortical DA (sensitization effect),91, 92, 93 even if the re-exposure involves a different stressor.94, 95, 96
Since humans typically encounter chronic, intermittent and unpredictable stressors, it may be more relevant to assess the effects of such regimens in animal studies. In contrast to the amine reductions induced by acute stressors, transmitter levels equal or exceed control values following protracted or repeated stressors,93, 97, 98, 99, 100 owing to increased amine synthesis and/or moderation of excessive utilization.98, 101 Moreover, chronic insults affect the amine variations engendered by later stressors. Specifically, in addition to a sensitization with respect to amine utilization (as seen following acute stressors), chronic stressors also induce sensitization of amine synthesis, thereby assuring adequate transmitter levels upon later stressor encounters.99 In addition, a chronic stressor may result in the down-regulation of β-NE receptor activity and the NE-sensitive cAMP response.102 Interestingly, upon application of a chronic unpredictable stressor the neurochemical adaptation was slower to develop.83, 103 It has been suggested that when inadequate neurochemical coping mechanisms are generated (eg, in genetically vulnerable animals, and when the stressor occurs unpredictably and involves a series of different insults), depressive-like characteristics may evolve.104 In fact, Anisman and Merali105 indicated that a regimen of mild, unpredictable stressors may be precisely the antecedent events most closely aligned with dysthymic-like states. It is important to emphasize at this juncture that while some degree of behavioral and neurochemical adaptation may occur in response to chronic stressor experiences, the view has been expressed that the wear and tear induced by attempts to adapt to a chronic stressor (allostatic load), when sufficiently protracted and/or intense, may culminate in pathological outcomes.106 As will be discussed later, chronic insults may, in fact, promote persistent neurochemical alterations which favor the development of depressive characteristics.
While there is considerable evidence supporting a relationship between major depression and central neurochemical disturbances,107, 108, 109, 110 scant information is available regarding the biological substrates of dysthymia. However, the frequent abnormal DSTs seen in major depressive patients were absent in dysthymia,2, 4 and the latter might actually be associated with hypocortisol responding.111 Indeed, it has been reported that the elevated salivary cortisol associated with exercise in pre-adolescent children was not apparent among dysthymic children of the same age.112 Paralleling these alterations of pituitary-adrenal activity, differences appeared between major depression and dysthymia in growth hormone secretion in response to physiological challenges, as well as TSH blunting in the TRH stimulation test.113
As part of a study assessing the role of psychosocial and biological variables in chronic and non-chronic major depression and dysthymia, a lower rate of DST non-suppression was observed in dysthymic patients (52% vs 8.5% non-suppression in major depression vs dysthymia). However, the rate of DST non-suppression was higher in the early-onset than in the late-onset dysthymics,113 although, when double depressives were excluded from the analysis (39 of the 75 dysthymic patients), there was no difference between early and late onset groups (9% and 8% respectively). Interestingly, paralleling the DST response, among late-onset dysthymics, the blunted TSH response to TRH administration was absent, whereas early-onset dysthymics (who parenthetically reported more traumatic and frustrating childhood backgrounds) had a higher rate of DST nonsuppression, and more frequently exhibited a blunted TSH response. In effect, these data suggest that early-onset dysthymia may represent a biologically distinct subgroup of chronically depressed patients.113 Indeed, it was noted that early-onset dysthymics responded preferentially to moclobemide relative to imipramine, while no such distinction was found among the late-onset dysthymics, suggesting that monoamine oxidase A might be more imbalanced among early-onset dysthymics.114
While limited attention has been devoted to the analysis of monoamine abnormalities in dysthymia, reduced levels of plasma NE coupled with elevated platelet and free 5-HT were evident in dysthymia.115 Further, following exercise, changes of epinephrine levels were relatively modest in dysthymic patients relative to control subjects. Thus, it was posited that dysthymia may be associated with altered adrenal responsivity to environmental challenges, as well as heightened sympathetic tone as reflected by the elevated free 5-HT levels. Ravindran et al111 observed reduced platelet MAO activity in primary, early-onset dysthymics relative to control subjects. Moreover, MAO activity prior to treatment was lower among nonresponders than among the drug responders. Along the same line, relative to endogenous depressives, MAO levels were low among neurotic depressives,116 while MAO activity correlated positively with clinical state in the endogenous group. Thus, low MAO activity may represent a marker for vulnerability to neurotic depression. Consistent with the potential involvement of serotonergic mechanisms in dysthymia, prior to treatment, lower urinary 5-hydroxyindoleacetic acid (5-HIAA) levels were observed among treatment responders relative to nonresponders, and the reduced levels in responders normalized following treatment.117
There have been few studies that assessed the electrophysiological correlates of dysthymia. However, it was reported that in anticipation of aversive stimuli, dysthymics exhibited hyporesponsiveness of skin conductance, and displayed subtle cognitive processing disturbances (possibly reflecting difficulties in the processing of complex information), as reflected by evoked potentials in response to task-relevant stimuli. It was posited that dysthymia may be associated with an impoverished ability to respond appropriately to external task demands, possibly owing to inappropriate allocation of processing resources. Furthermore, it was suggested that owing to their impaired resource allocation strategies, dysthymic subjects may be more generally impaired in their ability to cope with day-to-day stressors.118, 119
Akiskal et al120, 121 reported differences in the sleep architecture between subtypes of dysthymic patients. While subaffective dysthymics exhibited shortened REM latencies relative to controls, this was not the case among character spectrum disorder patients. In addition, dysthymia was associated with excessive and abnormal distribution of REM during the early part of the night121 as observed in major depression.2, 122 However, while the major depressive patients displayed reduced total sleep time, sleep latency, morning wake time and sleep efficiency, the sleep architecture of dysthymics in terms of stage percentages, and REM sleep features, were identical to those of major depressives. In effect, these data are consistent with the notion that the two disorders are variants of the same illness,123 or share common underlying mechanisms.
Pharmacological contributions to the analysis of dysthymia
The most persuasive data favoring a biological substrate for dysthymia originate from studies which assessed pharmacological agents in the treatment of dysthymic illness. The early pharmacological studies in dysthymia revealed that although MAO inhibitors and tricyclic antidepressants (TCAs) had superior therapeutic efficacy to placebo, their effects were not as marked as in major depression.4, 17, 124, 125, 126, 127, 128, 129, 130 However, it appears that reliable and impressive effects of antidepressant medications can be garnered in dysthymia. This stems from the development of medications, such as selective serotonin reuptake inhibitors (SSRIs), which have fewer side effects, thus permitting the use of higher doses. Moreover, it has been suggested that optimal drug effects would be obtained when administered primarily to patients with subaffective, rather than character spectrum disorder.4 In fact, studies which employed rigorous diagnostic criteria, established the efficacy of tricyclic agents, such as imipramine and desipramine,131, 132, 133, 134, 135, 136, 137 MAOIs,136, 138, 139 the reversible monoamine oxidase inhibitor, moclobemide,114, 137, 140, 141, 142, 143 SSRIs, such as fluoxetine and sertraline,22, 23, 111, 132, 134, 144, 145, 146 as well as other agents, such as the 5HT2 antagonist, ritanserin,147, 148 the selective norepinephrine reuptake inhibitor, reboxetine,149 and the serotonin/norepinephrine reuptake inhibitor (SNRI), venlafaxine150, 151 (Table 1). The use of well tolerated compounds, including moclobemide and sertraline, may be effective in the long-term management of dysthymia. This is particularly important since discontinuation of antidepressant treatment was found to be associated with an 89% rate of relapse in a 4-year maintenance study.18
In addition to the effects of the aforementioned antidepressants, hormonal manipulations have also been shown to influence dysthymic symptoms. Specifically, in a crossover-study involving a small number of subjects, it was observed that administration of the adrenal androgen, dehydroepiandrosterone, alleviated dysthymic symptoms, primarily comprising anhedonia, loss of motivation and energy, inability to cope, worry, emotional numbness, and sadness.163 Interestingly, these effects were obtained after only 3 weeks of treatment. Furthermore, the thyroid hormone, thyroxine, potentiated the effects of a variety of antidepressant medications in dysthymic and treatment-resistant chronic depressive patients.164 Moreover, in a small study of five patients it was observed that chromium supplementation enhanced the antidepressant effects of more traditional therapeutic agents.165
As indicated earlier, we observed in a double-blind placebo-controlled study, that sertraline was generally more effective than group CBT in treating the symptoms of dysthymia, as measured by the Hamilton Depression Scale.22 Of course, these data need to be considered as highly provisional, since the trial was short-term (12 weeks), and the CBT consisted of group rather than individual treatment. It is possible that individual CBT, or a program designed specifically for dysthymia may be more conducive to the treatment of the disorder. Further, given that anhedonia is a characteristic and persistent feature of dysthymia, it may have been beneficial to employ cognitive techniques which focused specifically on the inability of patients to experience or perceive positive events. Indeed, using cognitive behavioral psychotherapy, which focuses on the helplessness and hopelessness associated with dysthymia, and also teaches adaptive coping skills, McCullough21 reported that nine of 10 dysthymic patients were still in remission after a 2-year period. Given the high rate of depressive relapse/recurrence ordinarily observed following cessation of treatment, it will be interesting to establish whether combination therapy minimizes recurrence of illness relative to that seen among patients who had received only pharmacotherapy. This is particularly the case since group CBT enhanced the effects of sertraline with respect to some functional behaviors (eg, cognitive coping styles, and several indices of quality of life) which, in turn, may have important implications with respect to illness recurrence.22
Owing to dysthymia's fluctuating and chronic nature, several studies evaluated the efficacy of antidepressants in the maintenance treatment of the illness. Kocsis et al166 indicated that the higher relapse rate in those dysthymic patients who were randomized to placebo exceeded that of patients who continued on the maintenance desipramine treatment. Paralleling these findings, symptom improvement was sustained, and the rate of relapse reduced, among dysthymic patients who were maintained on either trazodone or fluoxetine over a 40-week interval compared to those who discontinued medication.167 Commensurate with the notion that a long-standing illness, such as dysthymia, might require prolonged pharmacotherapy, continuous improvement was observed among dysthymic patients treated over a 6-month period,146 while Kocsis et al160 found a substantially reduced rate of relapse (11%) among dysthymic patients maintained on desipramine over a 2-year period, relative to the 52% relapse rate in the placebo group. Although these data do not necessarily speak to the mechanisms subserving dysthymia, the results from these controlled clinical trials are congruent with the proposition that antidepressants are effective for a substantial portion of dysthymic patients (primarily the subaffective variety), and that prolonged maintenance treatment may be beneficial.
While most antidepressant trials have focused on the effects of 5-HT and NE manipulations on the symptoms of dysthymia and major depression, there have been several studies implicating a role for dopamine (DA). Since DA has been thought to subserve reward processes,168 and anhedonia is a characteristic feature of depression, the view has been taken that reduced DA activity might contribute to the depressive profile.104, 169, 170 It will be recalled, however, that in contrast to major depression, anhedonia is not one of the fundamental symptoms of dysthymia according to the DSM-IV criteria. Yet, it has been suggested that anhedonia may be a cardinal feature of dysthymic individuals.4 Unfortunately, there have been few studies that evaluated the effects of DA manipulations in dysthymia. The administration of the selective D2 and D3 antagonist amisulpride, an agent most often used as an antipsychotic when administered in high doses (400–1200 mg), has agonistic DA properties at low doses (50 mg), likely owing to preferential presynaptic binding. Consistent with the proposition that DA may play a role in dysthymia, amisulpride was effective in attenuating the symptoms of both dysthymia and major depressive illness.154, 155, 156, 169, 171, 172 In fact, amisulpride was as effective as imipramine in alleviating depressive symptoms in dysthymia, and both agents were significantly better than placebo in this respect (Table 1).
Genetic factors in dysthymia
Since the prevalence rates of various affective illnesses differ in families with dysthymic, major depressive and double depressive probands, it was suggested that dysthymia and major depression are independent disorders.2, 47 While dysthymia may represent a trait factor predicting increased risk for major depression, it may be important to distinguish between early- and late-onset illness. In fact, while major depression and dysthymia appear to be distinct illnesses, relatives of probands with early-onset depression were at increased risk for both major depression and dysthymia.173 Likewise, Goodman and Barnhill174 reported the results of a study comparing the rates of dysthymia in relatives of probands with either panic disorder, major depression, or both conditions (a subset of 33 patients were also dysthymic). Increased rates of dysthymia were observed in relatives of early-onset major depressives, and among relatives of dysthymic probands, thus supporting a relationship between early-onset major depression and dysthymia. Unfortunately, the small number of subjects tested makes it difficult to discern whether the risk of dysthymia varies as a function of early- vs late-onset of the disorder in the dysthymic proband. Paralleling these findings, rates of major depression in relatives of early-onset dysthymics compared to relatives of controls, confirmed a familial association between dysthymia and major depression. It was also observed that relatives of the dysthymics had higher rates of chronic depression than relatives of episodic depressives. Thus, there appears to be support for familial aggregation in dysthymia, as well as for the validity of dysthymia as a distinct diagnostic category.175
Donaldson et al176 found higher rates of dysthymia among relatives of pure dysthymics and of double depressives, than among relatives of major depressive probands and normal controls. Furthermore, the rates of pure dysthymia did not differ between relatives of pure dysthymics and those of double depressives, nor did they differ between relatives of major depressive and normal control probands. Once again, these data are consistent with the notion that while dysthymia may be distinct from major depressive disorder, dysthymia and double depression may be more closely related. The conclusions are clouded, however, by the finding that there was a higher rate of pure major depression among the relatives of pure major depressive probands, as well as among the relatives of double depressives, than among normal controls. These investigators suggested that dysthymia may be associated with two distinct etiological profiles. That is: (a) increased vulnerability to depression occurs in all relatives of unipolar depressive illness, irrespective of subtype; and (b) that risk for dysthymia may be particularly notable among relatives of dysthymic patients and those suffering from double depression.
The high comorbidity of dysthymia with personality disorders has consistently been noted.177 It has also been reported that the relatives of dysthymic patients, regardless of the presence of cluster B personality disorder (antisocial, borderline, histrionic, narcissistic) in the proband, exhibited increased frequency of dysthymia with and without cluster B personality disorder, as well as cluster B personality disorder without dysthymia. Thus, these results supported the notion that dysthymia and cluster B personality disorder share etiological factors such as genetic or familial factors, or early home environment.178
The contribution of genetic factors to dysthymia prompted Akiskal6 to categorize patients, in part, on the basis of genetic history. It was suggested that subaffective dysthymics frequently had a family history of depression, whereas character spectrum dysthymics tended to have a significant family history of alcoholism/drug abuse, but not of depressive disorder. Partial support for Akiskal's classification of subaffective vs character spectrum disorder was obtained from the finding that there was a higher rate of alcoholism among the relatives of the character spectrum disorder dysthymics, while the subaffective dysthymics exhibited higher rates of depressive symptoms, as well as personality and cognitive features.34 Unlike Akiskal's classification, however, these investigators did not observe differences between groups with respect to early home environment, family history of mood disorders, gender, or personality disorder.
Few studies have examined the genotypic expression of factors that might be related to dysthymia. However, it was demonstrated in a Japanese sample, that patients diagnosed with depressive disorders exhibited higher rates of genotypes coding for low activity catechol-o-methyltransferase relative to non-depressed controls.179 Although only five dysthymic patients were represented in the sample, the results obtained were consistent with those observed in a larger set of major depressive patients (n = 66) in this study. Of course, given the small number of subjects tested, these data must be considered cautiously. Nevertheless, they are suggestive of disturbances of enzyme activity related to catecholamine function in dysthymic patients, just as such effects may occur in major depressive disorder.
While the preceding studies supported a genetic contribution to dysthymia, other studies challenged this conclusion. For instance, monozygotic and dizygotic twins did not differ in their concordance rates for dysthymic illness (7.4% vs 8.7% respectively) as they did with respect to major depression.180 It was concluded that the shared or family environment may contribute more to the etiology of dysthymia than to major depression. It was argued that severity of depression and early-onset of the illness may be aligned with a genetic association, while the milder depressive illness spectrum (including dysthymia) may be more closely tied to environmental factors. As the dysthymic patients were not subdivided into character spectrum vs subaffective, it is unclear from these data whether the conclusion applies to both subtypes equally.
Cytokines and depressive illness
Increasing evidence has indicated that depressive illness is accompanied by immune dysregulation. While it had typically been assumed that depression promoted immunosuppression, it has been argued that the compromised immunity may actually be secondary to an initial immune activation. Furthermore, this notion has led to the possibility that products of an activated immune system may come to promote central neurochemical changes, hence provoking depressive symptoms.208 Commensurate with this view, depressed patients exhibited signs of immune activation, including increased plasma concentrations of complement proteins, C3 and C4, and immunoglobulin (Ig) M, as well as positive acute phase proteins, haptoglobin, α1-antitrypsin, α1 and α2 macroglobulin, coupled with reduced levels of negative acute phase proteins. Also, depression was accompanied by an increased number of activated T cells (CD25+ and HLA-DR+), secretion of neopterin, prostaglandin E2 and thromboxane. Furthermore, it appears that depression may be associated with variations of either circulating cytokine levels (ie, cell signalling factors released from activated macrophages), or cytokine production from mitogen-stimulated lymphocytes, including interleukin-2 (IL-2), soluble IL-2 receptors (sIL-2R), IL-1β, IL-1 receptor antagonist (IL-1Ra), IL-6, soluble IL-6 receptor (sIL-6R), and γ-interferon (IFN).209, 210, 211, 212, 213, 214, 215, 216 While there have been reports that the elevated levels of IL-1β, IL-6 and α1-acid glycoprotein normalized with antidepressant medication,217 the unregulated production of sIL-2R, IL-6 and sIL-6R was not attenuated with antidepressant agents, leading to the suggestion that the latter factors may be trait markers of the illness.208
Although severity of depressive illness is likely fundamental in determining cytokine levels,208 the possibility cannot be ignored that chronic depression (or chronic stress) may induce cytokine changes to a greater extent than those observed following acute episodes (as in the case of major depression). Consistent with reports in melancholic patients,208 levels of mitogen-stimulated IL-1β production were enhanced in dysthymia.55 However, in this particular study, IL-1β production was stimulated by the T cell mitogen, phytohemagglutinin (PHA), and thus may have reflected primarily T cell rather than macrophage-produced IL-1β. While not excluding the possibility that illness severity may be a pertinent feature in promoting the enhanced IL-1β production, it seems likely that illness chronicity or age of onset may also be important in this respect. Indeed, it was observed that age of onset was inversely related to IL-1β production, while duration of illness was directly related to production of this cytokine. Additionally, the altered IL-1β production was evident irrespective of whether a typical or atypical (reversed neurovegetative) profile was evident. Thus, it is unlikely that the altered cytokine production was related to the neurovegetative alterations that may appear in depression. It might be noted, as well, that IL-1β production was not markedly reduced with the alleviation of dysthymic symptoms following 12 weeks of SSRI treatment.55, 218 However, given that dysthymia is a long-standing illness, it is certainly possible that more prolonged treatment would have been necessary to realize changes of IL-1β, just as relatively protracted treatment was previously reported to promote variations of circulating natural killer cells.219
In contrast to the elevated IL-1β in supernatants of mitogen-stimulated lymphocytes, we observed that circulating serum IL-1β, presumably derived from macrophages and T cells, was not increased in either typical major depressive or in dysthymic patients. However, among atypical major depressive patients, circulating IL-1β levels were greatly increased, and normalized with treatment response.220 It could be assumed that the elevated levels of serum IL-1β in atypical depression were secondary to the neurovegetative features of this depressive subtype. However, animal studies have shown that IL-1β provokes some symptoms characteristic of atypical depression (including, increased sleep and fatigue),99 and thus it is just as likely that elevated circulating IL-1β contributes to the neurovegetative features of this depressive subtype. The finding that illnesses involving atypical depressive features (eg, chronic fatigue syndrome) may be associated with HPA disturbances (eg, reduced plasma cortisol, increased ACTH, and reduced ACTH release following oCRH challenge),221, 222 coupled with the fact that IL-1β is a potent stimulator of CRH release,223, 224 raises the possibility that elevated circulating IL-1β levels contribute to the pathophysiology of atypical depressive symptoms.
In support of the immune activation view of depressive illness, Maes225 indicated that in addition to the altered cytokine, acute phase protein, and hormonal changes ordinarily elicited as part of the inflammatory response, elevations of IL-1β may be evident in depression associated with a variety of medical illnesses. These include not only infectious diseases (influenza, herpes virus, HIV, Borna virus), but also numerous noninfectious illnesses, such as neurodegenerative disorders, autoimmune disorders and brain injury.61, 225 Thus, it was suggested that the cytokine activation associated with these illnesses and/or injuries may have provoked variations of HPA activity, as well as central neurochemical alterations, which may then have favored the development of depression. In effect, these comorbid conditions may have contributed to the illness owing to the cascade of cytokine, hormone and transmitter alterations engendered. It will be recalled that dysthymia likewise is associated with a large number of comorbid conditions, such as neurodegenerative disorders,1, 61 traumatic brain injury,62 and illnesses potentially involving viral components, such as fibromyalgia syndrome and chronic fatigue syndrome.53, 54 As such, the possibility ought to be considered that such coexisting illnesses are not simply correlates of the mood disorder, but may actually act to either precipitate or aggravate dysthymia.
The data presently available concerning cytokine changes in depressive illness (ie, studies showing elevations of the cytokines in severe major depression) are largely correlational. Thus, it is unclear whether the cytokine alterations seen in affective disorders are secondary to the illness (or the stress associated with the illness), or play an etiological role in the provocation of the disorder. Yet, administration of high doses of IL-2, IFN-α and tumor necrosis factor-α (TNF-α) in humans undergoing immunotherapy have been shown to induce neuropsychiatric symptoms, including depression, and these effects were related to the cytokine treatment rather than to the primary illness.226, 227, 228, 229 Of course, the doses administered in these studies were in the pathophysiological range, and thus their relevance to depression per se must be interpreted cautiously. However, as indicated by Meyers,230 even when administered at relatively low doses, cytokines such as IFN-α may elicit depressive-like symptoms.
A provisional model of chronic depressive illness
It is clear from the preceding sections that limited data are available concerning the mechanisms underlying dysthymia. Because of dysthymia's chronic, low-grade nature, animal models of the illness have yet to be developed. Nevertheless, data derived from animal studies offer some clues as to the potential persistent effects of stressor experiences that may be relevant to dysthymia. In particular, it seems that in addition to any immediate consequences, stressors may also proactively influence the neurochemical response to subsequently encountered aversive stimuli (sensitization), hence favoring long-term behavioral repercussions. Post and Weiss231 indicated that although the variations of certain peptides persist for relatively brief periods following a single stressor session, with repeated challenges the release of some peptides will be more readily induced and will be more persistent (eg, CRH, and to a greater extent TRH). It was suggested that depressive illness may initially stem from the neuroendocrine alterations provoked by a stressor. However, with each subsequent stressor experience, or with each episode of depression, the sensitization becomes more pronounced, such that progressively less intense psychosocial stressors are required to provoke the onset of a depressive episode. Ultimately, episodes of depression may occur in the absence of obvious stress triggers. In fact, it was reported that unlike the initial episode, recurrence was less likely to be preceded by antecedent stressors196, 232 and even occurred spontaneously.233, 234 Of course, it is often difficult to identify significant or meaningful stressors that may be pertinent to a given individual, thus conclusions concerning the presence or absence of stressful precipitants of depression may be difficult to validate. It is also conceivable that in addition to stressors of a psychological nature, a physiological stressor, such as a virus, may be interpreted by the CNS in the same way as a psychosocial stressor, hence triggering the cascade of events resulting in neuroendocrine, cytokine, and mood alterations.235
Although it is often thought that HPA disturbances are likely only a reflection of depression, it has been proposed that alterations of HPA activity could be the primary abnormality in depression, rather than simply an illness response.236, 237 Moreover, it has been suggested that among biologically predisposed individuals, chronic stressors may come to promote sustained HPA activation which leads to adverse effects.74 As alluded to earlier, the view has even been offered that the monoamine variations often associated with depression may actually stem from endocrine alterations.236, 237 In this respect, it was suggested that stressful events promote CRH variations in the central amygdala, which in turn may affect forebrain serotonin alterations. The former may reflect a basic stress-response, while the latter, presumably, entails the appraisal of the stressor situation.238
While the model developed by Post and Weiss231 was meant to accommodate recurrent depression, it may also be applicable to the analysis of dysthymia. In this respect, however, it is important to underscore that sensitization effects are not limited to the neuroendocrine factors discussed by Post, nor is such a sensitization limited to antecedent stressors. In fact, it has been demonstrated that several neurochemical alterations associated with stressors, psychostimulant use (amphetamine and cocaine), and electrical stimulation of the amygdala or the piriform cortex, may permanently enhance the neuronal response to subsequent manipulations (sensitization).239 As will be seen shortly, this applies to the effects of cytokine treatments as well.
In modelling dysthymic disorders, several features of this illness need to be considered, and it is important to distinguish these from other types of depression. Depressive disorders may be associated with profound interindividual differences in the symptoms subserving the illness, the response to pharmacotherapy, as well as the neuroendocrine correlates of the disorder. Further, subtypes of depression may differ in terms of their response to specific pharmacological treatments, and with respect to their neuroendocrine factors. For instance, major depression is characterized by HPA alterations, including elevated plasma ACTH and cortisol levels, nonsuppression of cortisol release following dexamethasone challenge, and a blunted ACTH response to CRH challenge.222, 240 While limited data are available, it appears that in illnesses involving atypical features (eg, bulimia, seasonal affective disorder, and chronic fatigue syndrome) the elevated ACTH levels are accompanied by reduced cortisol, a blunted ACTH response to CRH challenge,221, 222, 240, 241 and absence of the CRH hypersecretion characteristic of typical depression.240 Among dysthymic patients the profile is different yet again, and as indicated earlier, there is reason to believe that cortisol levels may actually be reduced,4 although a contradictory finding has been reported with respect to plasma cortisol and CRH concentrations.242 Moreover, in response to a stressor challenge, the cortisol response may be minimal in dysthymic patients relative to that seen in other types of depression and in nondepressed subjects.112 It remains to be established whether subaffective and character-spectrum dysthymia can be distinguished on the basis of these parameters. In any event, in providing a model of the mechanisms subserving dysthymia it needs to be understood why this disorder is not associated with cortisol abnormalities like those seen in major depressive illness. Furthermore, with respect to double depression, it would be of obvious benefit to determine why, following successful treatment, patients typically return to their dysthymic states rather than to an euthymic state.
The sensitization model described earlier introduces an important facet of stressor actions that may be relevant to understanding the relationship between stressor-induced neurochemical alterations and chronic depressive illness. Tilders and his associates243, 244, 245, 246 indicated that in response to repeated stressor experiences, or with the passage of time following a stressor or IL-1β challenge,247 phenotypic variations may occur within hypothalamic neurons that are ordinarily responsive to stressors. In particular, increased coexpression of CRH and AVP was observed within CRH containing neurons originating in the paraventricular nucleus (PVN) and having terminals in the external zone of the median eminence. As the co-released peptides act synergistically to promote ACTH secretion, the chronic stressor regimen increases the potential for elevated HPA functioning.243, 246 It is of particular significance, as well, that the altered coexpression of CRH and AVP was exceptionally long lasting, and was even evident as long as 60 days following stressor exposure.
Given the chronic nature of dysthymia, characterized by increased stress perception and inadequate coping styles, it is conceivable that this illness would be accompanied by the neuroendocrine characteristics ordinarily associated with chronic stressors. For instance, dysthymia may be associated with increased CRH and AVP coexpression within the external zone of the median eminence, as occurs with chronic stressors,246 and this may represent a permanent (or persistent) characteristic. The fact that dysthymic patients do not exhibit increased ACTH and cortisol, however, raises the possibility that the protracted CRH/AVP may have given rise to the down-regulation of pituitary and adrenal sensitivity. Clearly, this suggestion is highly speculative given that experiments have not been performed to assess, in detail, the characteristics of HPA functioning in dysthymic patients. Since an abnormal DST response has typically not been noted in dysthymia, it was taken for granted that this illness is not accompanied by any HPA disturbances. Yet, it may be the case that dysthymia is associated with adrenal hypofunctioning (as observed in atypical depression), rather than the hyperfunctioning seen in major depressive disorder.4 There are few neuroendocrine studies, however, that observed distinctive differences between dysthymic and non-depressed subjects. In part, this may stem from the subtle pathophysiological disturbances in dysthymia, and the use of neuroendocrine analyses that tap circulating hormonal levels rather than the dynamic, temporal patterns of hormone release.1 Also, assessment of HPA functioning in dysthymia requires evaluation of the effects of various challenges (eg, ACTH, CRH, AVP, as well as serotonergic acting agents) in order to identify the nature of any dysregulation that may exist.107, 248 Further, it is essential to evaluate these processes independently in the character-spectrum and subaffective variants of the illness.
Inasmuch as dysthymia is a chronic illness, often of mild-moderate severity and typically without clear precipitating events, it is unlikely that it is related to a single strong stressor experience. It is more reasonable to speculate that dysthymia reflects the actions of more sustained, variable, and probably less intense stressors, coupled with the use of inadequate or inappropriate methods of coping, culminating in the phenotypic CRH/AVP coexpression. Thus, it would not be altogether surprising to find that double depression may be related to the superimposition of a further stressor on the backdrop of dysthymia, which would then promote the increased release of these peptides (and the ensuing neurochemical cascade). Given this scenario, it might further be expected that after treatment of double depression, the CRH/AVP coexpression would persist and hence the dysthymic profile would be maintained. In these individuals the risk for further major depressive episodes would, of course, be heightened. In effect, we are suggesting that dysthymia may reflect a chronic state of altered endocrine and central neurotransmitter functioning which may be related to sustained stressor experiences together with inadequate coping. Indeed, even with the remission of symptoms, the persistent neuropeptide disturbances would increase the likelihood of symptom recurrence.249 Obviously, it would be of particular interest to establish whether the alleviation of double depressive symptoms would be accompanied by abnormal responses to CRH challenge, and whether such an effect differed from that seen following recovery from a major depressive episode. These factors, coupled with the long standing nature of the disorder, and the comorbid features discussed earlier, may necessitate a more sustained regimen of pharmacotherapy. Moreover, given the personality disturbances and the maladaptive cognitive coping strategies characteristic of dysthymia, in the absence of cognitive therapy or psychotherapy (as adjunctive or maintenance treatment) susceptibility to recurrence of illness may be increased. Indeed, we have shown previously that in spite of clinical improvement following pharmacotherapy, functional disturbances (as reflected by compromised quality of life, anhedonia) may persist in dysthymia. It was hypothesized that these residual features may actually be predictive of illness recurrence following cessation of pharmacotherapy.22
In relating stressful events to the mechanisms underlying dysthymia, we have defined stressors in a fairly broad way. As discussed earlier, it has been posited that, among other things, the immune system acts like a sensory organ informing the brain of antigenic challenge.250, 251 Furthermore, given the nature of the neurochemical changes elicited by antigens and cytokines, it was suggested that immune activation may be interpreted by the CNS as a stressor.44, 223, 235, 251 To be sure, the effects of systemic stressors (eg, those associated with viral insults, bacterial endotoxins, cytokines) are not entirely congruous with those elicited by processive stressors (ie, those involving higher order sensory processing).252 Nevertheless, cytokines may be part of a regulatory loop that, by virtue of their effects on CNS functioning, might influence behavioral outputs and may even contribute to the symptoms of behavioral pathologies, including mood and anxiety-related disorders.44, 235, 253 It is certainly the case that both processive and systemic stressors effectively increase HPA activity. However, while processive stressors do so via limbic circuits, the HPA alterations elicited by systemic stressors may result from limbic-independent processes.245 Yet, it ought to be underscored that systemic stressors, including IL-1β, IL-2 and TNFα have all been shown to influence central monoamine activity at both hypothalamic and extrahypothalamic sites, including hippocampal 5-HT activity, as well as that of NE and DA in hypothalamus, locus coeruleus and mesolimbic regions.235, 254, 255, 256 Thus, the possibility exists that cytokine elevations, by virtue of these monoamine effects, may come to promote or exacerbate depressive disorders, quite apart from any actions involving the HPA axis. It remains to be determined whether the IL-1β variations seen in dysthymia are secondary to the illness or, in fact, play an etiological role. Yet, as indicated earlier, this cytokine provokes behavioral changes, some of which are reminiscent of the characteristics of atypical depression, including increased sleep, lethargy and reduced locomotor activity,257 and may provoke anxiety.235 Given that the production of IL-1β in mitogen-stimulated lymphocytes is greater among dysthymic than among major depressive patients, particularly in those reporting early onset of the illness,55 the possibility exists that dysthymia may be associated with excessive cytokine reactivity. Increased IL-1β activation would then stimulate CRH functioning, and the ensuing neuroendocrine cascade. In effect, it may be that in dysthymic patients, stressors in the form of viral or bacterial challenges, may be particularly potent in provoking major depressive symptoms and hence promoting double depression.
The paucity of data from human studies, coupled with the lack of a suitable animal model for dysthymia, have limited the conclusions that can be drawn concerning the etiology of this disorder. Nevertheless, the available data have made it clear that elucidation of the mechanisms underlying dysthymia, and the development of adequate treatment strategies, will require that several fundamental features be included in experimental analyses. Foremost in this respect is the need to subtype subjects according to definite criteria. In particular, it will be of obvious advantage to distinguish between pure dysthymia, double depression, and other forms of chronic depression. Additionally, patients need to be characterized into homogeneous subgroups (eg, early- vs late-onset; subaffective vs character-spectrum), and the symptom profile of the dysthymic patients ought to be considered (vis-à-vis the presence of typical or atypical neurovegetative symptoms).
Although genetic factors likely contribute to the expression of dysthymia, there is also reason to believe that experiential factors play a cogent role in this respect. There is no information, however, as to whether early-life experiences contribute to the biological (subaffective) type of dysthymia. It is interesting, however, that studies in rodents have indicated that early life maternal deprivation may give rise to a cascade of neurochemical alterations much like those purported to occur in dysthymia. These include increased CRH mRNA expression in the amygdala and CRH concentrations in the median eminence, as well as increases of CRH receptors in the prefrontal cortex, amygdala, hypothalamus and cerebellum. As adults, rats that had undergone maternal deprivation display increased stress-elicited arousal and elevated HPA functioning.258, 259 The possibility ought to be explored that in humans, early-life stress or ‘neglect’ may give rise to these neurochemical disturbances, hence increasing vulnerability to later stressor-induced neuroendocrine and neurotransmitter alterations and ultimately the dysthymic profile. Of course, the failure to establish appropriate coping strategies (and this includes affiliation, attachment and support systems, particularly with parents) may augment stressor effects, thereby encouraging the development of dysthymia.
Finally, the identification of subtypes of dysthymia may be an important feature in determining the optimal treatment strategy employed. From the outset, it must be acknowledged that because dysthymia is a chronic condition, relatively sustained treatment may be required to alleviate the symptoms.260 Thus, drugs that are relatively well tolerated will be most efficacious in treatment, particularly when these agents do not elicit sexual dysfunction or somatic complaints, symptoms which themselves typically are not characteristic of dysthymia.18 Further, it is likely, as indicated earlier, that the effectiveness of various treatment strategies may be related to factors such as age of onset, and the presence of characterological features. In this respect, determining whether a given episode is associated with neuroendocrine disturbances (eg, reduced cortisol secretion secondary to excessive CRH activation) may offer insights into whether pharmacotherapy (and the type of agents used) would be most efficacious in treating the disorder. Of course, family history of psychiatric illness and the effectiveness of specific pharmacotherapy therein, would be of obvious value in planning a treatment strategy. Finally, it ought to be underscored that the effectiveness of cognitive therapy has not been extensively evaluated in dysthymic patients. Nevertheless, it would appear that this therapeutic modality may be useful in treating some dysthymic patients. It remains to be established what characteristics of the illness might be predictive of those patients who would benefit most from this form of therapy. In this respect, it may be useful to consider functional parameters related to quality of life (eg, cognitive disturbances, social interaction, life satisfaction), as opposed to relying simply on clinical indices of depression. Given the particularly high rate of relapse in dysthymia upon cessation of pharmacotherapy,18 the possibility ought to be considered that cognitive therapy, particularly when focusing on residual functional disturbances, would be useful as an adjunctive or maintenance treatment strategy.
This work was supported by the Medical Research Council of Canada. HA is an Ontario Mental Health Foundation Senior Research Fellow.