Moderators Of The Depressionimmune Link

It is not known what accounts for the studywise heterogeneity in the relationship between depression and immunity. Stein et al. (1991) proposed that alterations in the immune system in major depression are not specific biologic correlates of this disorder, but rather occur in association with other variables that characterize depressed subjects such as age, hospitalization stress, and severity of depressive symptoms. Indeed, in the two meta-analyses conducted by Herbert and Cohen (1993a and b) and Zorrilla et al. (1998), subject characteristics such as age, gender, and symptom severity were evaluated as possible moderators of the relationship between depression and immunity. However, these moderator variables have been found to explain only partially the heterogeneity observed.

The role of other factors in moderating the effects of depression on the immune system has not been systematically explored. However, several exploratory studies have been conducted which indicate the possibility that additional subject characteristics and/or depression status variables may moderate the relation between depression and immunity.

In the present review, the role of several moderator variables will be considered in addition to age, gender, and symptom severity. For example, life stressors have been found to be associated with changes in the immune system (Herbert & Cohen, 1993b), and such stress-induced immune alterations are similar to the kinds of changes found in depressed subjects. Thus, we will discuss whether the subacute stress of hospitalization and/or chronic life stress often reported in depressed patients contribute to the immune changes found in diagnostic depression (Caldwell et al., 1991; Irwin et al., 1990b).

Second, symptom severity has been suggested as an important correlate of immune changes in depressed subjects, but what specific symptoms account for this relation requires further study to know whether there are certain subgroups or subtypes of depressed subjects who are most at risk for immune changes. Some evidence suggests that depressed patients with neurovegetative symptoms are more likely to show immune changes (Cover & Irwin, 1994) and a number of observations implicate sleep in the regulation of some aspects of immune function (Irwin, Mascovich, Gillin, Willoughby, Pike, & Smith, 1994; Irwin, McClintick, Costlow, Fortner, White, & Gillin, 1996; Irwin, Smith, & Gillin, 1992b).

Finally, co-morbidity in depression has emerged as a major concern in biologic research in psychiatry with recent epidemiological data showing, for example, that as many as 50% of depressed patients are co-morbid for an anxiety disorder such as panic (Stein & Uhde, 1988; Murphy, Oliver, Sobol, Monson, & Leighton, 1986). Moreover, the prevalence of alcohol and tobacco dependence is high in depressed subjects (Breslau, Peterson, Schultz, Chilcoat, & Andreski, 1998; Schuckit, 1986), but few studies have examined the contribution of alcohol intake or cigarette smoking on alterations of immune function in depressed subjects despite the well-recognized effects of these substances on immune parameters. The following will consider the effects of comorbidity for anxiety and/or substance dependence in the relation between depression and immunity.

3.1. Subject Characteristics

3.1.1. Age. Age has been proposed as a key moderator in the association between depression and immunity. In three reviews (Weisse, 1992; Stein et al., 1991; Herbert & Cohen, 1993a), it was concluded that older depressed patients are more likely to show immune differences as compared to younger depressed subjects. The findings of Schleifer et al. (1989) are most representative of this conclusion; controls showed an age-related increase in CD4 numbers and mitogen responses, whereas depressed patients did not show such changes in either T cell numbers or lymphocyte proliferation with advancing age. However, specific age group comparisons of depressives vs. controls have not been conducted and it is not known, for example, whether elderly depressed subgroups show greater immune differences than adolescent- or young adult samples. Furthermore, in studies of depressed patients, age, and hospitalization status are correlated, and as cautioned by Herbert and Cohen (1993a), the separate effects of age and hospitalization status on depression-immunity associations can not be determined without carefully matching depressed inpatients and outpatients with controls on both age and hospitalization status. Of course, another strategy would be to design a study that was restricted to either inpatient- or outpatient depressed subjects and then to evaluate the effects of age within that hospitalization status subgroup. For example, Irwin et al. (1990b) focused on inpatient depressed subjects and found that these subjects had lower NK activity than controls. However, within the inpatient depressed subjects, there was no effect of age or an age by depressive symptom interaction on NK activity. Zorrilla et al. (1998) also recently demonstrated using a random effects moderator analyses that age did not demonstrate a significant moderating effect with any of the immune variables found to be reliably altered in depression.

3.1.2. Gender. The stability of the depression immune relationship across gender also deserves further attention due to the well-recognized effects of reproductive hormones such as progesterone to antagonize the suppressive effects of glucocorticoids on cellular immune function (Dietrich, Chasserot-Golaz, Beck, & Bauer, 1986). Many studies have included only male depressed subjects and eliminated the possible contribution of gender (Irwin et al., 1990a; 1990b). Other studies have included male and female depressives and evaluated effects of gender using correlational analyses (Schleifer et al., 1989). The latter have typically not found any effects of gender on either enumerative or functional immune measures in depression, a finding consistent with the conclusion reached by Herbert and Cohen (1993a). However, only a few studies have stratified the sample and made direct comparisons between depressives and controls matched on the basis of gender. Miller, Asnis, Lackner, Halbreich, & Norin (1991) found in outpatient depressives that female subjects had higher NK activity than controls, whereas male depressives were similar to controls. In contrast, Evans et al. (1992) studied both inpatient and outpatient depressed subjects and found that men with major depression had a marked reduction in number of NK cells (Leu-11) and in NK activity compared with levels in controls. By contrast, depressed women did not differ significantly on any of these immune measures. Similar comparisons on the basis of gender have not been conducted for other immune measures, although there is evidence that studies with a higher percentage of females are more likely to show greater deficits in Con A responsiveness along with smaller decreases in NK cell cytotoxicity (Zorrilla et al., 1998).

3.1.3. Stress. Both life stress and depression have been associated with similar changes of immune measures (Herbert & Cohen, 1993a and b), and it is possible that these two conditions together might interact and influence immunity more than the presence of life stress or depression alone. Depressed subjects are more likely to experience severe life events during their depressive episode due in part to the functional impairments associated with depression (Brown, Harris, & Hepworth, 1994). For example, a depressive episode may be associated with deterioration in work performance such that the person loses his/her job. Likewise, impairment in social interactions during depression may affect a spousal relationship leading to threatened divorce or separation. To our knowledge, the joint contribution of threatening life events and depression on immune function has only preliminarily been examined in one study.

In a sample of 36 subject pairs of depressed patients and comparison controls, the presence of severe threatening life stress was assessed using the objective methods of Brown and Harris (1978). Subjects in both groups were classified as having low- or high stress and NK activity was measured. Findings demonstrate a reduction of NK activity in subjects with major depression as compared with persons who are neither stressed nor depressed. Second, individuals who are undergoing severe life stress show a similar reduction of NK activity even though they do not have clinically significant depressive symptoms. The combination of severe stress and depression did not result in any further reduction of NK activity. Thus, the immune changes found in life stress parallel those found in depression. However, the changes found in depressed subjects can not be explained simply by the occurrence of severe life stress (Irwin et al., 1990b). In other words, life stress influences cellular immunity, but it does not appear to moderate the association between depression and immunity. Similar conclusions were also reached in the evaluation of the effects of acute hospitalization stress on depression related impairments of NK activity (Caldwell et al., 1991).

3.2. Depression Status

3.2.1. Ambulatory Status. A decrease of mitogen responses in hospitalized patients with major depression but not in ambulatory patients with depression suggests that alterations of immunity in depression are related to hospitalization status or to severity of depressive symptoms (Schleifer et al., 1984; 1985). While subsequent study of both inpatient and outpatient depressed subjects failed to identify an effect for hospitalization status (Schleifer et al., 1989), the meta-analyses conducted by Herbert and Cohen (1993a) and by Zorrilla et al. (1998) showed that hospitalization status indeed moderated the effects of depression on mitogen responses but had no effect on other immune measures. Suppression of mitogen-induced lymphocyte proliferation in depressed subjects tends to be greater in inpatient than in outpatient samples. However this meta-analytic result requires caution. The set of studies that was used included hospitalized subjects who were more severely depressed than the ambulatory subjects. Thus, the meta-analytic effect of ambulatory status may be more explained by the moderating effects of depressive symptom severity than by hospitalization status per se.

3.2.2. Depression Severity. To examine the contribution of depressive symptom severity on immune measures, Schleifer et al. (1989) measured severity of depressive symptoms (total HDRS scores) and assessed B, T, and T, cell subset numbers, mitogen-induced lymphocyte stimulation and NK activity in a sample of 30 inpatient and 61 outpatient depressed subjects. Severity of depressive symptoms was found to be associated with a decline of ConA and PHA lymphocyte responses independent of the effects of gender, age, and hospitalization status. However, neither the enumerative measures or NK activity was correlated with severity of depressive symptoms in this large sample. Evans and colleagues also found that neither NK cell numbers nor NK activity correlated with severity of depressive symptoms, even though diagnostic depression was associated with a decline in numbers of NK cells and NK activity as compared to levels found in comparison controls (Evans et al., 1992). While these negative findings contrast with those of Irwin and colleagues who found that severity of depressive symptoms is a distinct correlate of NK activity in three independent samples (Irwin et al., 1987, 1990a, 1990b), it appears that the moderating effect of severity of depressive symptoms on immune measures is small. In the recent meta-analyses conducted by Zorrilla et al. (1998), total score on the HDRS was not reliably correlated with either of the functional measures, lymphocyte proliferation or NK activity. Moderator analyses also did not reveal correlations between severity of depressive symptoms and other immune measures such as total white blood cell count or percentages of neutrophils and lymphocytes.

The lack of relationship between depressive symptom severity and measures of lymphocyte proliferation and NK activity is surprising considering the robust association between these immune measures and diagnostic depression. However, the discrepancy between these two observations raises an important question concerning the link between depression and immune changes. If there is no association between depressive symptom severity and immune alterations in depressed subjects, then it is possible that immune alterations might persist even into remission and be indicative of a trait rather than a state marker of depression.

To our knowledge, one longitudinal case-control study has been conducted to examine state vs. trait issues in the depression-immune link (Irwin et al., 1992a). This longitudinal design also provided an opportunity to test the temporal relationship between symptom severity and immunity beyond the prior correlational analyses conducted in cross-sectional samples. Assessment of NK activity was obtained in the second week of hospitalization and repeated at six months after discharge from the hospital in 20 depressed subjects. Matched comparison controls were studied at the same interval and on the same day as their depressed subject pair. The depressed subjects showed a decrease in HDRS and an increase of NK activity from intake to follow-up. There was no change in either depression scores or NK activity in the controls.

The temporal relationship between depressive symptom severity and reduced NK activity contrasts with the meta-analytic findings of Zorrilla et al. (1998) who found no association between depressive symptom severity and NK activity. It might be that the state of current depression is actually a higher order variable. Depression is correlated with symptom severity, but the association between acute diagnostic depression and altered immunity is related to an interaction and/or combination of depressive symptom severity with other correlates of the state of depression.

3.2.3. Depression Subtype. That a constellation of subject characteristics and depressive symptoms explains the decline of cellular immunity is depression is also consistent with findings generated in the study of immune alterations in melancholic depression. Melancholic depressed patients are typically older, hospitalized, and have more severe depressive symptoms (Parker, Hadzi-Pavlovic, Boyce, Wilhelm, Brodaty, 1990; Mitchell, Hickie, & Eyers, 1990). While none of these characteristics alone have been found to moderate the association between depression and immunity, these characteristics together with the diagnosis of melancholia have considerable predictive significance in determining impaired cell mediated immunity in depression. Melancholic depressed subjects are reported to be at increased risk for immune alterations and to show lower levels of mitogen-induced lymphocyte proliferation and of NK activity as compared to depressed subjects without melancholia (Cosyns et al., 1989; Maes et al., 1989; 1991; 1995). In addition, patients with melancholia show impaired in vivo immune responses measured by delayed type hypersensitivity skin responses whereas non-melancholic depressed subjects have responses similar to controls (Hickie, Hickie, Lloyd, Silove, & Wakefield, 1993). Moreover, the decrement of delayed type hypersensitivity responses was predicted by the diagnosis of melancholia independent of the separate effects of age, hospitalization status, and depression severity.

3.2.4. Neurovegetative Symptoms. Rather than the sum of depressive symptoms as reflected in total HDRS scores, certain symptoms may be relatively more important in moderating the association between diagnostic depression and immunity. Indeed, the association between melancholic depression and impaired cellular immunity may be due in part to these patients' increased prevalence of neurovegetative symptoms such as sleep disturbance, appetite disturbance with associated nutritional impairments, and weight loss, and/or increased psychomotor retardation. Thus, to examine whether certain depressive symptoms account for some of the immune changes in depression, Cover and Irwin (1994) extended previous analyses that examined the relationship between depressive symptoms and immunity by evaluating the contribution of clusters of depressive symptoms.

Six symptom clusters, defined by factor analysis of items from the HDRS, were identified: anxiety/somatization, weight loss, cognitive disturbance, diurnal variation, retardation, and sleep disturbance (Cleary & Guy, 1977). In a sample of depressed inpatients, no correlation was found between total HDRS scores and NK activity. Likewise, there was no correlation between four symptom clusters (anxiety/somatization, weight loss, cognitive disturbance, or diurnal variation) and NK activity. However, two symptom clusters, retardation, and sleep disturbance, were correlated with NK activity, and together these two symptom variables accounted for over 16% of the variance in NK activity (Cover & Irwin, 1994).

3.2.5. Nutritional status. In regards to the lack of relationship between weight loss and NK activity as described above, Schleifer et al. (1989) also found that statistical control for weight and recent weight loss was not associated with depression-related declines in lymphocyte proliferation. However, further investigation of the potential effects of nutritional factors in the depression-immune relation is clearly needed before excluding nutritional status as a variable that might account for immune alterations in depressed subjects.

3.2.6. Sleep disturbance. Insomnia is one of the most common complaints of depressed subjects, but its role in moderating and/or mediating immune alterations in depression has been relatively unexplored. However, with evidence that subjective insomnia correlates with NK activity in depression (Cover & Irwin, 1994), the hypothesis emerged that disordered sleep may be a distinct factor accounting for some of the observed immune alterations found in depression. Consequently, a series of studies have now been conducted to test more carefully the role of sleep in the modulation of multiple aspects of the immune system and to determine the moderating effects of sleep disturbance on immune measures in depressed subjects.

First, observations regarding subjective complaints of insomnia and immunity were extended by assessment of disordered sleep by EEG (Irwin et al., 1992b). Many aspects or parameters of sleep are assessed during all-night EEG studies, thus the initial approach was to identify those measures of EEG that are altered in association with insomnia or subjective sleep disturbance and then to characterize the relationship of these EEG measures to immunity. Self-report of sleep disturbance or insomnia is characterized by disturbances of EEG sleep continuity measures with increases of sleep latency (the interval from lights out to onset of sleep) and decreases in total sleep time and sleep efficiency (the ratio of sleep to the amount of time in bed) (Benca, Obermeyer,Thisted, & Gillin, 1992). Considering the relation between subjective sleep disturbance and NK activity, we hypothesized that measures of sleep continuity measures would correlate with NK activity in depressed subjects. In addition, if sleep has a distinct role in the modulation of immune function independent of diagnostic depression and other depressive symptoms, then similar correlations between sleep and NK activity should also be identified in control subjects who differ in the amount and quality of their sleep. Indeed, both total sleep time and sleep efficiency were positively correlated with NK activity in separate groups of depressed subjects and controls who underwent all night EEG study with assessment of NK activity in the morning upon awakening (Irwin et al., 1992b). These data were some of the first to indicate that sleep amounts are associated with immune function and to suggest that disordered sleep may be a key variable in understanding the behavioral mechanisms underlying the link between depression and immune alterations. Recent studies in bereaved subjects have replicated this correlation between EEG sleep and immunity and shown by way of causal statistical analyses that disordered sleep also mediates the relationship between severe life stress and a decline of NK responses (Hall, Baum, Buysse, Prigerson, Kupfer, & Reynolds, 1998).

A second line of studies to investigate the contribution of sleep on immune function has focused on the immunological assessment of a group of insomniac subjects who report disordered sleep but who are not depressed or suffering from some other psychiatric disorder. Primary insomnia is diagnostically characterized by the presence of subjective sleep difficulties that persist unrelated to another mental disorder or a known organic factor, such as a physical condition, psychoactive substance use disorder, or a medication (American Psychiatric Association, 1994). In a recent study of such primary insomniac patients who have no current or lifetime history of another mental disorder, EEG sleep and immune measures of NK activity, NK cell numbers and lymphokine activated killer (LAK) cell activity were compared with sleep and immune measures obtained in depressed subjects and controls. Both psychiatric groups, insomniac and depressed subjects, showed significant disturbances in sleep continuity with prolonged sleep latency and decreases in total sleep time and sleep efficiency as compared to the controls. Furthermore, the insomniac and depressed groups showed similar alterations in NK and LAK responses. The insomniac subjects showed a reduction of NK activity and LAK activity as compared to levels in controls. The declines of NK and LAK responses in insomniac subjects were similar to those observed in depressed subjects (Irwin & Gillin, 1998).

A third strategy used to evaluate the relationship between disordered sleep and immunity has involved an experimental approach, sleep deprivation. In an effort to mimic the kind of disordered sleep found in depressed subjects, the effect of partial night sleep loss on lymphocyte function and NK and LAK responses has been tested. Depressed subjects often report symptoms of early or late insomnia with loss of sleep during only parts of the night. Thus, it was thought that a partial night sleep deprivation paradigm would more closely parallel the kinds of sleep continuity disturbance reported in depression. In contrast, previous studies have examined the effects of prolonged, total sleep deprivation on immunity (Palmblad, Petrini, Wasserman, & Akerstedt, 1979; Moldofsky, Lue, Davidson, & Gorczynski, 1989; Dinges, Douglas, Zaugg, Campbell, McMann, Whitehouse, Orne, Kapoor, Icaza, & Orne, 1994). The effects of partial sleep loss on immunity had not been previously explored.

In two separate studies, one that examined sleep loss during the late part of the night (i.e. awake from 3 am to 7 am) (Irwin, et al., 1994) and the other that tested the effects of sleep loss during the early part of the night (i.e., awake from 11 pm to 3 am) (Irwin et al., 1996), substantial alterations of functional immune measures occurred in association with partial sleep deprivation. This modest sleep loss, typical of depressed subjects and other psychiatric and non-psychiatric groups who report sleep disturbance, was associated with declines of NK activity (Figure 1), LAK activity (Figure 2), and stimulated IL-2 production (Figure 3). The reduction of NK and LAK responses appeared to be due to impairments in the activity of these cells, rather than a selective redistribution of circulating populations of NK cells and LAK precursors. For example, calculation of NK lytic activity per number of NK cells (CD16,56) and of LAK cytotoxicity per number of LAK precursors (CD16,56 + CD25) revealed impairments of individual effector cell function. Furthermore, partial night sleep loss induced a marked decline in the stimulated production of IL-2 by peripheral blood mononuclear cells. Interestingly, this effect of sleep loss on the release and/or utilization of IL-2 was due to effects on both lymphocyte and adherent, antigen-presenting cell populations. In experiments that involved mixing adherent and nonadherent cells obtained after base-

Cigarette Crusher

Cigarette Crusher

Get All The Support And Guidance You Need To Permanently STOP Being A Slave To Nicotine And Cigarettes. This Book Is One Of The Most Valuable Resources In The World When It Comes To Easy Ways To Eliminate Smoking Addiction And Revitalize Your Body.

Get My Free Ebook


Post a comment