| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLE |
From the Department of Psychiatry, University of California, San Diego, and Department of Psychiatry, San Diego VA Medical Center, San Diego, California.
Address reprint requests to: Michael Irwin, MD, Department of Psychiatry V116A, San Diego VA Medical Center, 3350 La Jolla Village Drive, San Diego, CA 92161. Email: mirwin{at}UCSD.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: A sample of 245 men were stratified into four groups: control subjects who were not smokers, control subjects who were smokers, subjects with major depression who were not smokers, and subjects with major depression who were smokers. Blood samples were obtained for measurement of total white blood cell (WBC) counts, differential cell counts, and assay of NK activity.
RESULTS: Major depression and cigarette smoking interact and were together associated with changes in WBC counts and NK activity. Depressed subjects who were smokers had higher WBC counts (p < .001) and lower NK activity (p < .01) than depressed nonsmoking subjects. However, WBC counts and NK activity were similar in control smokers and nonsmokers. Backward elimination regression analyses showed that the interaction of depression and smoking significantly (p < .001) predicted WBC counts and NK activity.
CONCLUSIONS: This study extends previous findings of immune alterations in patients with major depression. Major depression and smoking interact and together contribute to an elevation of total WBC count and a decline of NK activity.
Key Words: major depression • cigarette smoking • immunity • depression • cancer • natural killer cell activity
Abbreviations: NK = natural killer cell; WBC = white blood cell count; MHCRC = Mental Health Clinical Research Center; DSM =Diagnostic and Statistical Manual of Mental Disorders; ANOVA =analysis of variance; ANCOVA = analysis of covariance.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Rather than there being a unitary link between depression and cancer, depression might interact with other characteristics, such as cigarette smoking or alcohol consumption, to affect health. Indeed, in a 12-year follow-up study of 2264 adult men and women, depressed mood was found to interact with cigarette smoking, and together depressed mood and cigarette smoking were associated with a marked increase in the relative risk of cancer (7). Compared with never smokers who were not depressed, smokers with depressed mood, as measured by elevated scores on the Clinical Epidemiological Scale for Depression, had a relative risk of 18.6 for cancers at sites associated with smoking and a relative risk of 2.9 for cancers at sites not associated with smoking. In contrast, smokers who were not depressed had a relative risk of only 4.2 for cancers at sites associated with smoking and no increase in relative risk of cancers at sites not associated with smoking.
The mechanisms that might account for the combined effect of smoking and depression on cancer development are far from understood. Nevertheless, substantial evidence indicates that dysfunction of certain aspects of the immune system, such as NK activity, might contribute to primary tumor development and metastatic cancer risk (8, 9). Patients with genetic deficiencies in NK function have an increased incidence of cancer that is mostly lymphoreticular in origin (9). Additionally, normal individuals with a high familial incidence of cancer have lower NK activity than individuals with a low incidence of cancer (10, 11). Indeed, NK activity varies inversely with the number of family members with cancer (10). Moreover, in patients who have various types of solid tumors, such as breast cancer or squamous cell carcinoma of the head and neck, low levels of NK activity at the time of diagnosis correlate with risk of metastatic disease development (12, 13). Finally, in two in vivo animal models, one of lung metastases of breast carcinoma and the other of human squamous cell carcinoma of the head and neck, the link between NK cells and cancer is clearly shown by evidence that adherent NK subsets are potent antitumor effector cells with potentially therapeutic value in the treatment of advanced solid tumors (14, 15).
Recent meta-analyses indicate that reduced NK cytotoxicity is one of the most reliable and reproducible alterations of immune function in depression (16, 17). However, the clinical significance of the decline of NK activity in depression is not known, and no data have shown that low NK activity in depression is correlated with increased cancer risk (18). In regard to cigarette smoking, effects on NK lytic activity are less clear, with decreases (19–23), increases (24), and no change of NK activity (25–27) reported to occur in current smokers as compared with nonsmokers.
In view of the prevalence of tobacco dependence in depressed individuals (28) and the possible influence of cigarette smoking on immune function, it is important to examine the relationship between depression, smoking status, and NK activity. However, of the nearly two dozen studies that have reported immune alterations in depression, only two reported smoking histories (29, 30). Although neither found that smoking consumption correlated with the changes of immune function in depression, moreover, neither assessed a possible synergistic relationship between depression and moderate (1–2 packs per day) cigarette use. To our knowledge, no study has compared immune function in depressed and control subjects stratified on the basis of smoking history and tested whether there is a possible interaction between depression and smoking on immune function.
In this study, we examined the influence of current cigarette smoking on NK activity in depressed and control subjects. Using data from a previously described large series (N = 245) (29, 31–35), we stratified depressed and control subjects on the basis of nonsmoking or current smoking status. Values of total WBC counts, differential cell counts, and NK activity in the four groups were compared.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Subjects with major depression were self-referred from the San Diego community to the University of California, San Diego, MHCRC for evaluation of depressive symptoms and possible inclusion in MHCRC nonintervention and/or treatment trials. All depressed subjects underwent immunological assessment for the present study before entry into a MHCRC treatment protocol. Controls subjects, also identified and recruited by the MHCRC, volunteered in response to community educational outreach efforts (ie, lectures, meetings with community groups, etc.) or responded to advertisements placed in local newspapers or campus publications.
Procedures
Both depressed and control subjects underwent a medical evaluation that involved a medical history, review of systems, physical examination, and screening laboratory evaluation (ie, complete blood count, chemistry panel, thyroid tests, and HIV antibody test). They were in good medical health, did not report any recent (<10 days) viral illness, and were not taking any immunosuppressive medications (eg, corticosteroids or cancer chemotherapeutic drugs). Depressed subjects discontinued use of antidepressant and/or anxiolytic medications for at least 14 days before immunological assessment.
To ascertain the psychiatric diagnosis of all subjects, the Schedule for Clinical Interview and Diagnosis DSM-IIIR or DSM-IV was administered (36–38). Diagnoses were then made during a consensus meeting of MHCRC psychiatrists, research fellows, and nursing staff. Depressed subjects fulfilled DSM-IIIR or DSM-IV criteria for a current episode of major depressive disorder. Control subjects fulfilled criteria for lifetime history of never mentally ill (37, 38), consistent with methods used in previous studies of immune alterations in depression (16).
On the day of immunological assessment, tobacco- and substance-dependence histories were obtained using the previously validated Alcohol Research Center Intake Interview (39), which gives detailed information about consumption histories, dependence symptoms, and clinical course of substance use (36). We focused on current smoking status rather than past cigarette use, because there is evidence that any possible decline of NK activity reverses within 30 days of smoking abstinence and that former smokers have an average level of NK activity comparable to that of nonsmokers (21). Severity of depressive symptoms was measured using the Hamilton Depression Rating Scale (40).
Blood sampling for assay of total WBC counts, differential cell counts, and NK activity was obtained between 8 and 10 AM. All subjects rested in a reclining chair for 15 to 30 minutes before blood sampling, which was conducted with a 21-gauge intravenous catheter. In a subsample of subjects (N = 46), the catheter was placed and subjects continued to rest for an additional 20 minutes before the sample was collected. We previously showed, using a within-subjects design, that similar levels of NK activity are found in blood samples obtained after various intervals of rest from 15 to 30 minutes (32, 41). In the present study, which used a between-subjects design, neither WBC counts, differential values, nor NK activity showed systematic changes related to differences in blood sampling pro- tocols.
Immune Assays
WBC counts were obtained using standard automated procedures (Coulter Counter S+4, Coulter Electronics, Hialeah, FL) as previously reported (29). Differential cell counts were manually determined using a modified Wright’s solution by a trained hematology technician within 2 hours of sample acquisition.
For the assay of NK cell activity, mononuclear cells were sedimented on Ficoll-Hypaque (Pharmacia, Piscataway, NJ). Cells were then collected at the interface, washed twice with phosphate- buffered saline, incubated on plastic plates to remove adherent cells, and resuspended to yield a cell suspension that was more than 99% viable with less than 2% monocytes when stained with Wright’s stain. Lymphocytes were then suspended in RPMI with 10% fetal calf serum at a concentration of 2 x 106 cells per milliliter and incubated with 51Cr-labeled K562 target cells at effector-to-target cell ratios of 40:1, 20:1, 10:1, and 5:1 in triplicate. After 3 hours of incubation in a 37°C incubator with 5% carbon dioxide, the plate was removed and spun at 200g. An aliquot (100 µl) was removed, and chromium release was measured in a gamma counter. Assay results for NK activity were expressed as the percentage of specific cytotoxicity across the four effector-to-target cell ratios.
Statistical Analyses
A 2 x 2 ANOVA (group: control and depressed subjects; smoking status: nonsmokers and smokers) was used to establish the main effects of depression, smoking, and their interaction in predicting age, severity of depressive symptoms, alcohol consumption, and WBC and differential cell counts. Differences in cigarette smoking histories between control and depressed smokers were tested by an independent t test. To evaluate the central hypothesis that depression and smoking produce a more marked decline of NK activity than depression or smoking status alone, a planned comparison using a repeated-measures ANCOVA covarying for age was conducted. NK activity across the four effector-to-target cell ratios was compared between the depressed smokers and depressed nonsmokers, control smokers, and control nonsmokers. Planned comparisons are considered the most appropriate test for evaluating specific hypotheses and allow for the extraction of information critical to the status of the research question responsible for initiating the present study (42). Additional pairwise comparisons tested whether NK activity differed between depressed smokers and depressed nonsmokers and between control smokers and control nonsmokers. Finally, regression models were used to test the relative contribution of depression, smoking, and their interaction on WBC counts, differential values, and NK activity after control for confounders, such as age and alcohol use. Backward elimination regression analyses were used in which the following predictor variables were entered: age, alcohol consumption (days per month of drinking), smoker or nonsmoker as a binary variable, and depressed or nondepressed as a binary variable. The set of regression analyses was then repeated with the addition of the smoking x depression interaction to identify the importance of this interaction in predicting the immune outcome measures.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
|
For NK activity, a planned comparison covarying for age demonstrated that depressed smokers had a marked reduction of NK activity compared with depressed nonsmokers, control smokers, and control nonsmokers (F = 13.0, p < .001; Table 3 and Figure 1 ). An additional pairwise comparison demonstrated that depressed smokers had lower NK activity than depressed nonsmokers (F = 7.8, p < .005). In contrast, within control subjects, smoking was not associated with a change in levels of NK activity; control smokers and control nonsmokers had similar levels of NK activity (F = .02, p = .89).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The immunological changes found in depressed smokers do not appear to be due to the simple effects of smoking. Only depressed subjects who were smokers had elevated WBC counts and declines of NK activity; control subjects who were current smokers showed no change in immune variables. Although depressed and control smokers were statistically different in the reported consumption of cigarettes, both groups reported, on average, light to moderate cigarette consumption. Heavy smokers, who report cigarette use nearly twice the level of that in our control and depressed smokers, show a decline of NK activity (22), whereas light to moderate smokers, who report cigarette histories similar to those of our subjects, typically show no change of cytotoxicity (21, 22). The immune alterations in depressed smokers, along with the finding that depression and smoking interact to uniquely contribute to elevated WBC counts and reduced NK activity, suggest that cigarette smoking, even in the moderate range, combines with depression to have synergistic effects on WBC counts and NK activity.
There are several limitations to this study. First, the control group was self-selected in response to recruitment drives and may not be representative of the general population. Second, consistent with the methods of prior studies that have described immune alterations in depressed patients (16), the control subjects had no lifetime history of a major psychiatric disorder. Thus, the immunological differences between the control and depressed subjects might be viewed as related to the absence of psychiatric morbidity in controls rather than positively associated with depression. Indeed, other psychiatric samples show immune alterations similar to those found in depressed patients. Stein et al. (18) proposed that immune changes in major depression are not specific biological 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. Third, the study sample included only men. Findings of reduced NK activity in depression may be gender specific, as suggested by Evans et al. (44). Fourth, prevalence rates of smoking are elevated in depressed patients as compared with normal individuals (28), and the control group in the present study had a low percentage of smokers (8%). Even in a study as large as the one performed here, only a small number of control smokers could be identified (N = 11). Thus, the lack of effect of cigarette smoking on NK activity in control subjects requires cautious interpretation because it may be due to Type II error. Fifth, self-reports of smoking history do not take into account differences in nicotine delivery, which depends on the cigarette and how it is used. Moreover, nicotine has been found to inhibit NK cytotoxicity (45), and determination of plasma levels of nicotine and cotinine would have permitted further objective assessment of the relationship between smoking and changes of immune function. Finally, assay of NK activity involved a heterogeneous population of peripheral blood mononuclear cells depleted of adherent cells. It is possible that a selective loss of NK cells in the peripheral blood, rather than a defect in the function of this subset, accounts for the decline of NK activity in depressed smokers. In the present study, number of NK cells, as measured by phenotypic expression of the surface antigens CD16, 56, was evaluated in a subset of subjects (N = 54). There was no difference in percentage of CD16, 56 cells across the groups (F = .88, p = .46), and NK activity was not correlated with CD16, 56 cells (r = -.11, p = .44), consistent with the findings of Evans et al. (44).
The results of this study have several implications. First, smoking status is a critical variable to assess in studies examining the relationship between depression and immunity. The combined effects of depression and smoking appear to predict changes in numbers of WBCs and NK activity independent of the effects of depression and smoking alone. Because of the interaction between depression and smoking on WBC counts and NK activity, inconsistent findings are likely to be reported if different studies have samples of depressed subjects who differ in the prevalence of smoking. Cigarette use alone and/or in combination with depression might also contribute to the suppression of other nonspecific measures of immune function, such as mitogen-induced lymphocyte proliferation. Moreover, cigarette smoking is associated with immune activation (46), and it is important to address whether smoking status alters the reported relationship between depression and increases in serum levels of interleukin-6 and other acute-phase proteins (47). However, compared with the effects of smoking on nonspecific measures of immunity, smoking may have less of an impact on disease-specific measures of immune function. A marked decline of varicella zoster-specific lymphocyte responses has been found in nonsmoking depressed subjects compared with age-matched control subjects (43).
The effects of smoking on NK activity in depressed but not control subjects might also suggest that NK cells of depressed subjects are more sensitive or responsive to the immunosuppressive effects of nicotine than those of controls. Indeed, differential responsivity to nicotine, although not examined in immune cells, has been reported in depressed subjects as compared with controls. Sleep studies have revealed that transdermal administration of nicotine increases rapid eye movement sleep in nonsmoking depressed subjects but decreases rapid eye movement and the percentage of rapideye-movement sleep in control subjects (48). Other data also show that nicotine administration and cigarette smoking are associated with disordered sleep (49, 50). These data, together with observations that sleep disturbance is correlated with a reduction of NK activity (34, 51, 52), suggest a role of disordered sleep in mediating the relationship between depression, smoking, and a reduction of lytic activity. An alternative mechanism might involve the adrenocortical axis. Both depression and cigarette smoking are associated with activation of the adrenocortical axis (53, 54), and it is possible that smoking and depression interact to produce a greater elevation of cortisol that inhibits NK activity. However, in a subset of subjects (N = 86), plasma levels of cortisol were coincidentally measured with NK activity. No difference in cortisol levels were found between the groups of depressed and control smokers and nonsmokers (F = 1.9, p = .13), and there was no correlation between cortisol and NK activity in the total sample (r = -.03, p = .81).
The health implications of reduced NK activity in depressed smokers are uncertain, and the decrease of NK activity in depressed smokers may not be of etiological significance with respect to the increase in cancer that occurs in these patients. However, experimental and clinical data point to NK cells as effector cells in immune surveillance with a demonstrated ability to lyse target cells that have undergone malignant transformation (8). Furthermore, it is important to note that the degree of suppression of NK activity found in depressed patients is comparable to the relative decline of cytotoxicity in individuals with a high familial incidence of cancer, who are at increased risk of cancer development (9, 10). In addition, inescapable stress, a putative animal model of depression, induces a suppression of NK activity that compromises host resistance to NK-sensitive tumors (55). Because of the interaction of depression and smoking on cancer risk (7), prospective studies that examine NKs and the development of cancer at sites associated and not associated with smoking are necessary to establish a causal link.
In summary, the present study shows that the combination of depression and cigarette smoking is associated with a decline of NK activity. These immunological findings are compatible with epidemiological data that show an interaction between depressed mood and smoking in the risk of development of cancer.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication April 27, 1998.
Revision received December 11, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. J. Motivala, A. Sarfatti, L. Olmos, and M. R. Irwin Inflammatory Markers and Sleep Disturbance in Major Depression Psychosom Med, March 1, 2005; 67(2): 187 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Vines, S. Gupta, T. Whiteside, D. Dostal-Johnson, and A. Hummler-Davis The Relationship between Chronic Pain, Immune Function, Depression, and Health Behaviors Biol Res Nurs, July 1, 2003; 5(1): 18 - 29. [Abstract] [PDF]
|
|
|
|
 |
|
 S. Greer Healing the Mind/Body Split: Bringing the Patient Back Into Oncology Integr Cancer Ther, March 1, 2003; 2(1): 5 - 12. [Abstract] [PDF]
|
|
|
|
 |
|
 S. T. Orr, S. A. James, and C. Blackmore Prince Maternal Prenatal Depressive Symptoms and Spontaneous Preterm Births among African-American Women in Baltimore, Maryland Am. J. Epidemiol., November 1, 2002; 156(9): 797 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Kiecolt-Glaser, L. McGuire, T. F. Robles, and R. Glaser Psychoneuroimmunology and Psychosomatic Medicine: Back to the Future Psychosom Med, January 1, 2002; 64(1): 15 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bargellini, A. Barbieri, S. Rovesti, R. Vivoli, R. Roncaglia, and P. Borella Relation between immune variables and burnout in a sample of physicians Occup. Environ. Med., July 1, 2000; 57(7): 453 - 457. [Abstract] [Full Text]
|
|
|
|
|
|
J. K. Kiecolt-Glaser and R. Glaser Psychoneuroimmunology and Immunotoxicology: Implications for Carcinogenesis Psychosom Med, May 1, 1999; 61(3): 271 - 272. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
