Interleukin-6 Is Elevated in the Cerebrospinal Fluid of Suicide Attempters and Related to Symptom Severity

Interleukin-6 Is Elevated in the Cerebrospinal Fluid of Suicide Attempters and Related to Symptom Severity

Interleukin-6 Is Elevated in the Cerebrospinal Fluid of Suicide Attempters and Related to Symptom Severity Daniel Lindqvist, Shorena Janelidze, Peter ...

294KB Sizes 0 Downloads 88 Views

Interleukin-6 Is Elevated in the Cerebrospinal Fluid of Suicide Attempters and Related to Symptom Severity Daniel Lindqvist, Shorena Janelidze, Peter Hagell, Sophie Erhardt, Martin Samuelsson, Lennart Minthon, Oskar Hansson, Maria Björkqvist, Lil Träskman-Bendz, and Lena Brundin Background: Depressive disorders are associated with immune system alterations that can be detected in the blood. Cytokine concentrations in cerebrospinal fluid (CSF) and their relationship to aspects of suicidality have previously not been investigated. Methods: We measured interleukin-1␤, interleukin-6 (IL-6), interleukin-8, and tumor necrosis factor-␣ (TNF-␣) in CSF and plasma of suicide attempters (n ⫽ 63) and healthy control subjects (n ⫽ 47). Patients were classified according to diagnosis and violent or nonviolent suicide attempt. We evaluated suicidal ideation and depressive symptoms using the Suicide Assessment Scale and the Montgomery-Åsberg Depression Rating Scale (MADRS). We also analyzed the relation between cytokines and monoamine metabolites 5-hydroxyindoleacetic acid (5-HIAA), homovanillic acid (HVA), and 3-methoxy-4-hydroxyphenylglycol (MHPG) in CSF, as well as the integrity of the blood-brain barrier as reflected by the CSF:serum albumin ratio. Results: IL-6 in CSF was significantly higher in suicide attempters than in healthy control subjects. Patients who performed violent suicide attempts displayed the highest IL-6. Furthermore, there was a significant positive correlation between MADRS scores and CSF IL-6 levels in all patients. IL-6 and TNF-␣ correlated significantly with 5-HIAA and HVA in CSF, but not with MHPG. Cytokine levels in plasma and CSF were not associated, and patients with increased blood-brain barrier permeability did not exhibit elevated cytokine levels. Conclusions: We propose a role for CSF IL-6 in the symptomatology of suicidal behavior, possibly through mechanisms involving alterations of dopamine and serotonin metabolism. Key Words: 5-HIAA, attempted suicide, cytokines, HVA, interleukin-6, major depressive disorder

T

he term “sickness behavior” refers to a cluster of symptoms displayed during infection, including fatigue, problems with concentration, and lack of appetite. These features are also common in clinical depression, and therefore it was postulated that the immune system could be involved in depressive disorders (1). Depressed patients display increased blood levels of proinflammatory cytokines interleukin-6 (IL-6) (2), interleukin-1beta (IL-1␤) (3), tumor necrosis factor-␣ (TNF-␣) (4), and interleukin-8 (IL-8) (5). Other studies have reported immune alterations in the blood of suicide attempters (6,7), and epidemiological studies have revealed an association between allergy and suicidality (8,9). The source of the elevated blood cytokines in depressive disorders is as yet unknown. Detected cytokines may reflect immune system changes in peripheral tissues, but cytokines can also be produced within the central nervous system (CNS), mainly by astrocytes and microglia (10). Despite abundant evidence of peripheral immune alterations, little is known about From the Department of Clinical Sciences (DL, SJ, LT-B, LB), Section of Psychiatry, Lund University Hospital, and Department of Health Sciences (PH), Lund University, Lund, Sweden; Department of Physiology and Pharmacology (SE), Karolinska Institutet, Stockholm, Sweden; Faculty of Health Sciences (MS), Department of Clinical and Experimental Medicine, Psychiatry Section, Linköping University, Linköping, Sweden; Department of Clinical Sciences (LM, OH), Clinical Memory Research Unit, Lund University, SE-205 02 Malmö, Sweden; Department of Experimental Medicine (MB), Neuronal Survival Unit, Wallenberg Neuroscience Center, Lund University, Lund, Sweden. Address reprint requests to Lena Brundin, M.D., Ph.D., Department of Clinical Sciences, Section of Psychiatry, Kioskgatan 19 Lund University Hospital, SE-221 85 Lund, Sweden; E-mail: [email protected]. Received October 13, 2008; revised January 11, 2009; accepted January 28, 2009.

0006-3223/09/$36.00 doi:10.1016/j.biopsych.2009.01.030

possible concurrent inflammatory changes in the CNS of depressive and suicidal patients. Two recent studies found microgliosis and elevation of pro-inflammatory cytokine messenger ribonucleic acids (mRNA) in postmortem brain sections from suicide victims (11,12), suggesting neuroinflammation as a possible trigger for suicidal behavior. In depressed patients, CSF levels of IL-1␤ were found to be elevated and IL-6 decreased (13) or unaltered (14). It is not clear whether the inflammatory changes in depressed and suicidal patients contribute directly to the generation of symptoms or if they constitute epiphenomena. Case reports describe an association between immune-modulating interferon therapies and suicidal behavior, which is suggestive of a causal relationship (15,16). Clinical trials employing cyclo-oxygenase-2 inhibitors against depression report a positive effect, providing further support for a causative link between inflammation and depressive symptoms (17). Experimental models have shown that proinflammatory cytokines can modulate neurotransmission and affect serotonin, noradrenaline and dopamine turnover in the CNS (18). For instance, systemic administration of IL-1␤ and IL-6 in rats increases the concentration of serotonin metabolite 5-hydroxyindoleacetic acid (5-HIAA) in different brain regions (19,20). Indeed, altered levels of CSF monoamine metabolites have been found in suicide attempters (21). We therefore hypothesized that some of the behavioral effects of cytokines in suicide attempters would be exerted through disturbance of monoamine systems. The aim of this study was to investigate associations between CSF levels of proinflammatory cytokines and multiple aspects of suicidality. We also analyzed the relationship among the cytokines, monoamine metabolites, and psychiatric symptoms.

Methods and Materials Patients This study was approved by the local ethic committees in Lund, Malmö, and Linköping, Sweden. All patients gave inBIOL PSYCHIATRY 2009;66:287–292 © 2009 Society of Biological Psychiatry

288 BIOL PSYCHIATRY 2009;66:287–292

D. Lindqvist et al.

Table 1. Main Axis I Psychiatric Diagnoses and Somatic Diagnoses in All Suicide Attempters (n ⫽ 63) and Violent Suicide Attempters (n ⫽ 16) All Suicide Attempters (n ⫽ 63)

Violent Suicide Attempters (n ⫽ 16)

Major Depressive Disorder (n)

19

7

Depression Not Otherwise Specified (n)

13

4

Adjustment Disorder (n)

12

1

4 4 3 1 7

0 0 1 1 2

Main Axis I Psychiatric Diagnosis

Substance Abuse Disorder (n) Dysthymia (n) Psychosis (n) Eating Disorder (n) No Diagnosis (n)

Somatic Diagnosis Gastritis (n ⫽ 1) Migraine (n ⫽ 1) Gastritis (n ⫽ 1) Migraine (n ⫽ 1) Diabetes (n ⫽ 1) Arthrosis (n ⫽ 1) Asthma (n ⫽ 1) N/A Gastritis (n ⫽ 1) N/A N/A N/A

N/A, not applicable.

formed consent and were enrolled in the study on admission to Lund University Hospital after a suicide attempt. Patients did not receive any antidepressant or antipsychotic medication during a washout period of 16 ⫾ 7 (mean ⫾ SD) days after the suicide attempt. At the end of the washout period, lumbar punctures were performed and patients were subjected to a general physical examination, blood sampling, and psychiatric evaluation. Sixty-three suicide attempters (33 men and 30 women), with a mean age of 39 ⫾ 14 (⫾ SD) years, without traces of antidepressant or neuroleptic medication in their blood samples were included. Diagnoses were set according to DSM-III (22) (Table 1). Depressive symptoms were evaluated using the Montgomery-Åsberg Depression Rating Scale (MADRS) (23). Patients were also evaluated using the Suicide Assessment Scale (SUAS), which is an interview-based scale consisting of 20 items (24). SUAS item number 16 rates “wish to die,” and SUAS item number 18 rates “preoccupation with suicidal thoughts.” Control Subjects The healthy control subjects were recruited through the Neuropsychiatric and Psychiatric Clinics at the University Hospitals in Malmö and Linköping, Sweden, and PrecisionMed, San Diego, California. They consisted of 40 men and seven women, somatically healthy, aged 37 ⫾ 20 (mean ⫾ SD), with no previous or ongoing psychiatric condition. They were thoroughly checked for psychiatric comorbidity by an evaluating psychiatrist and with the Structured Clinical Interview for DSM-IV Axis I and II Disorders (25,26). Somatic Comorbidity All patients and healthy control subjects went through a general physical examination in conjunction to the lumbar puncture. A complete medical history was taken, including questions about current and previous diseases. Control subjects were free of medication. Eight patients had received somatic diagnoses (Table 1). One control subject had asthma but was medication-free. To identify subjects with potential infections at the time of the lumbar punctures, blood samples were analyzed for inflammatory and hematological markers (white blood cell count, erythrocyte sedimentation rate or C-reactive protein, hemoglobin, glucose or glycosylated hemoglobin, thyroid-stimulating hormone, liver enzymes, and creatinine), and the subjects were checked for fever. No evidence of infection was found. www.sobp.org/journal

Lumbar Punctures and Analyses of Biochemical Markers CSF was drawn from the L4 –L5 interspace between 8 and 11 AM after a night of fasting and bed rest, using a previously described protocol (27). Sampling procedures for control subjects and patients were similar, although control subjects were not hospitalized. The samples were collected during all four seasons for both patients and control subjects. They were stored in aliquots and immediately frozen at – 80°C. IL-6, IL-1␤, IL-8, and TNF-␣ were quantified in CSF and plasma using multiplex sandwich enzyme-linked immunosorbent assays. We employed Discovery assays (MesoScale, Gaithersburg, Maryland) as per the manufacturer’s protocol on a SECTOR 6000 instrument (http://www.mesoscale.com). The respective detection limits in our analysis were IL-6: .1 pg/mL; IL-1␤: .2 pg/mL; IL-8: .3 pg/mL, and TNF-␣: .1 pg/mL. The cytokine analyses were undertaken in 2008, 10 ⫾ 5 (mean ⫾ SD) years after lumbar puncture. Concentrations of 5-HIAA, homovanillic acid (HVA), and 3-methoxy-4-hydroxyphenylglycol (MHPG) in CSF from patients were analyzed with gas chromatography mass spectrometry (28). The absolute values have been described elsewhere (29). As a measure of the integrity of the blood-brain barrier, the ratio between albumin in CSF and serum was determined (30). For 15 patients, information regarding blood-brain barrier permeability was incomplete. Statistical Analyses We used the Statistical Package for Social Sciences (SPSS, Chicago, Illinois) program, version 15.00. For groupwise comparisons, Student t test was applied. Mann-Whitney U test was used for small number of subjects. Nonnormally distributed variables were transformed into natural logarithms (ln) before analysis. For multiple parametric comparisons, one-way analysis of variance (ANOVA) was used, correcting for covariates when appropriate (ANCOVA). Post hoc tests were performed using the Bonferroni-Dunn method. We used Pearson’s r and the Spearman’s rho for correlations. The cytokines were measured in triplicates and duplicates, and the analysis repeated twice. Mean values are reported from the second experiment.

Results CSF Cytokines: Demographics The data for IL-1␤, IL-6, and 5-HIAA displayed skewness above 2 and were transformed into normal distribution using

BIOL PSYCHIATRY 2009;66:287–292 289

D. Lindqvist et al.

Table 2. IL-1␤, IL-L6, IL-8, and TNF-␣ Concentrations in CSF and Plasma (pg/mL), Mean ⫾ SEM in Suicide Attempters and Control Subjects and Classified According to Diagnosis and Violent Suicide Attempt

All Patients Control All Patients MDD AD Depression NOS Violent Attempt Nonviolent Attempt

Medium

n

IL-1␤

IL-6

IL-8

TNF-␣

CSF CSF Plasma CSF CSF CSF CSF CSF

63 47 33 19 12 13 16 47

.08 ⫾ .01 .07 ⫾ .01 1.35 ⫾ .39 .06 ⫾ .01 .11 ⫾ .03 .07 ⫾ .01 .07 ⫾ .01 .08 ⫾ .01

1.97 ⫾ .83a .64 ⫾ .09 5.19 ⫾ .64 3.76 ⫾ 2.65b .83 ⫾ .16 1.94 ⫾ 1.14 5.26 ⫾ 3.21c .85 ⫾ .08

23.70 ⫾ .85 23.10 ⫾ .97 3.72 ⫾ .41 25.07 ⫾ 1.64 25.16 ⫾ 2.08 23.14 ⫾ 2.08 21.57 ⫾ 1.77 24.43 ⫾ .95

.15 ⫾ .01 .13 ⫾ .01 6.16 ⫾ .53 .16 ⫾ .01 .14 ⫾ .01 .13 ⫾ .01 .15 ⫾ .02 .15 ⫾ .01

Patients were diagnosed with MDD, AD, and depression NOS. The data for IL-1␤ and IL-6 were nonnormally distributed, and the natural logarithms were used for all statistical analyses. AD, adjustment disorder; CSF, cerebrospinal fluid; IL, interleukin; MDD, major depressive disorder; NOS, not otherwise specified; TNF, tumor necrosis factor. a Student t test, all patients (CSF) vs. control subjects (CSF), p ⬍ .001. b One-way analysis of variance, Bonferroni-Dunn post hoc tests, MDD vs. control subjects, p ⬍ .002. c One-way analysis of variance, Bonferroni-Dunn post hoc tests, violent vs. nonviolent suicide attempters, p ⬍ .004; violent suicide attempters vs. control subjects, p ⬍ .001.

natural logarithms before statistical analysis. IL-8, TNF-␣, MHPG, and HVA measurements were normally distributed. Patient and control subject age correlated positively with IL-8 and TNF-␣, and analyses including these cytokines were corrected for age. There was no impact of sex, smoking, body mass index, sample age, or length of the washout period on cytokine levels (Student t test and Pearson’s r, ns). One patient and one control subject had mild asthma (non–steroid treated), and the results did not differ with or without these individuals. The results presented include all patients (n ⫽ 63) and control subjects (n ⫽ 47). CSF Cytokines and Psychiatric Diagnoses Suicide attempters (n ⫽ 63) displayed significantly higher levels of IL-6 compared with control subjects (n ⫽ 47; Student t test, p ⫽ .001). Levels of IL-1␤, IL-8, and TNF-␣ did not differ between the groups (Table 2). When the suicide attempters were divided on the basis of main Axis I diagnosis, major depressive disorder (MDD; n ⫽ 19), depression not otherwise specified (n ⫽ 13), and adjustment disorder (n ⫽ 12), the group with MDD had significantly higher levels of IL-6 than healthy control subjects (one-way ANOVA, Bonferroni-Dunn post hoc test, p ⫽ .002). The IL-6 levels did not differ between the other diagnostic groups and control subjects. For TNF-␣, IL-8, and IL-1␤, there were no significant differences between the diagnostic groups and control subjects (one-way ANCOVAS and ANOVA, ns; Table 2). CSF Cytokines and Psychiatric Symptoms Patients who performed a violent suicide attempt (n ⫽ 16) had significantly higher levels of IL-6 in CSF compared with patients who performed nonviolent attempts (n ⫽ 47) and healthy control subjects (n ⫽ 47; one-way ANOVA, BonferroniDunn post hoc tests; p ⫽ .004 and p ⬍ .001 respectively; Table 2). TNF-␣, IL-8, and IL-1␤ did not differ significantly between the groups (one-way ANCOVAs and ANOVA, ns). There was a significant positive correlation between total MADRS scores and IL-6 in all suicide attempters (n ⫽ 63; Pearson’s r ⫽ .31; p ⫽ .016). There were no correlations between any of the cytokines and total SUAS scores among the suicide attempters, either when analyzed all together or when divided into violent and nonviolent suicide attempters (Pearson’s r, ns). Among the violent suicide attempters (n ⫽ 16), CSF IL-6 corre-

lated with scores on SUAS items 16; “preoccupation with suicidal thoughts” (Spearman’s rho ⫽ .51, p ⫽ .044) and 18; “wish to die” (Spearman’s rho ⫽ .60, p ⫽ .014). Relationship Between CSF Cytokines and Monoamine Metabolites In all suicide attempters, we found positive significant correlations among CSF 5-HIAA, HVA, IL-6, and TNF-␣ (Pearson’s r, see Table 3; Figure 1A and 1B). MHPG did not correlate with any cytokine (Table 3). When divided into violent and nonviolent suicide attempters, these correlations remained significant in the violent group. Blood-Brain Barrier Increased permeability of the blood-brain barrier was defined in relation to reference limits of CSF:serum albumin ratios derived from a group of healthy control subjects described previously (31). There was no correlation between the CSF: serum albumin ratios and the CSF cytokines in the suicide attempters (Pearson’s r, ns). Cytokine Levels in Plasma Plasma samples were available from a subset of the suicide attempters, taken at the same occasion as the lumbar punctures (n ⫽ 33). Measurements for all cytokines in plasma were normally distributed, and mean levels were higher than in CSF for IL-1␤, IL-6, and TNF-␣ (Table 2). Mean IL-8 was 10-fold higher in CSF than in plasma. Cytokine levels in CSF and plasma did not correlate (Pearson’s r, ns).

Discussion We examined four proinflammatory cytokines in CSF of healthy control subjects and suicide attempters with welldefined clinical psychiatric profiles and found that suicide attempters had elevated levels of IL-6. The highest levels were detected in patients who had performed a violent suicide attempt. Furthermore, CSF IL-6 correlated positively with the severity of depressive symptoms in all suicide attempters. When divided on the basis of the four most prevalent psychiatric Axis I diagnoses, suicide attempters with MDD had elevated CSF IL-6 levels compared with control subjects. IL-1␤, TNF-␣, and IL-8 did not differ between any of the groups and control subjects. www.sobp.org/journal

290 BIOL PSYCHIATRY 2009;66:287–292

D. Lindqvist et al.

Table 3. Pearson Correlations Between CSF Cytokines IL-1␤, IL-6, IL-8, TNF-␣, and CSF Monoamine Metabolites 5HIAA, HVA, and MHPG (n ⫽ 63) IL-6 IL-6 IL-1␤ TNF-␣ IL-8 5-HIAA HVA MHPG

⫺.09 .56b .15 .40b .45b ⫺.08

IL-1␤

TNF-␣

⫺.09

.56b .09

.09 .07 ⫺.25 ⫺.12 ⫺.07

.28b .27a .28a .00

IL-8 .15 .07 .28b .12 ⫺.06 .13

5-HIAA

HVA

MHPG

.40b ⫺.25 .27a .12

.45b ⫺.12 .28a ⫺.06 .77b

⫺.08 ⫺.07 .00 .13 .25a .12

.77b .25a

.12

The data for IL-1␤, IL-6 and 5-HIAA were non-normally distributed and their natural logarithms were used in the statistical analyses. 5-HIAA, 5-hydroxyindoleactic acid; CSF, cerebrospinal fluid; HVA, homovanillic acid; IL, interleukin; MHPG, 3-methoxy-4-hydroxyphenylglycol; TNF, tumor necrosis factor. a Pearson’s r, p ⬍ .05. b Pearson’s r, p ⬍ .005.

The suicide attempters displayed no signs of infectious, neurological, or autoimmune disorder when subjected to a thorough physical examination and routine laboratory screening. Therefore, such conditions are unlikely to have caused the rise in CSF IL-6. Elevated plasma IL-6 has been reported in patients with MDD (32), despite any known underlying inflammatory condition. In the periphery, the main source of IL-6 is immune cells, but skeletal muscle and adipose tissue can also secrete this cytokine (33,34). In our study, there were no significant correlations between patient body mass index and CSF cytokine levels. Furthermore, patients were hospitalized, fasting, and in bed the night and morning before the lumbar punctures, minimizing any possible effect of cytokine release from muscle. A previous study showed increased production of IL-6 from peripheral blood mononuclear cells in depressed patients compared with control subjects (35). Under normal conditions, IL-6 can be actively transported across the blood-brain barrier (36). Hence, it is possible that the IL-6 detected in the CSF of suicide attempters could derive from peripheral blood. However, plasma and CSF cytokine concentrations did not correlate in this study. Moreover, the degree of permeability of the blood-brain barrier was not associated with the cytokine levels, speaking against a leakage from the periphery as the source of the elevated CSF IL-6. Our results rather suggest that the IL-6 measured in CSF originates from the CNS, where microglia and astrocytes are known to produce cytokines (10). This would be in line with studies

reporting microgliosis and increased levels of pro-inflammatory cytokine-mRNA in brain tissue from suicide victims (11,12). On the basis of the findings in this study, we suggest that CSF IL-6 is of importance for the development of depressive and suicidal symptoms. The group of suicide attempters with MDD stood out with higher levels compared with control subjects. In addition, CSF IL-6 was positively correlated with total MADRS scores, which rate depression. Experimental models support the theory of IL-6 as an etiological factor in depression. Mice with a transgenic overexpression of IL-6 in the CNS display enhanced stress-induced glucocorticoid levels (37), possibly reflecting an increased vulnerability for developing depressive-like behaviors. In contrast, IL-6-deficient mice are resistant to helplessness and display reduced despair in forced swimming (38). The results in our study also suggest that IL-6 has a specific relevance for suicidal behavior. The group with the highest levels was the violent suicide attempters, containing not only patients with MDD but several other main diagnoses. Interestingly, in the violent suicide attempters CSF IL-6 measurements correlated positively with symptoms dealing specifically with suicidality— namely, “wish to die” and “preoccupation with suicidal thoughts.” To test fully the idea of whether IL-6 directly can generate depressive and suicidal symptoms in humans, it would be necessary to administer the cytokine peripherally or into the CNS and assess subsequent behavioral and affective symptoms. This approach has been used in rodents in which peripheral and

Figure 1. (A) Cerebrospinal fluid (CSF) logarithms (LN) interleukin-6 (IL-6) plotted against CSF LN 5-hydroxyindoleactic acid (5-HIAA). Circles represent the individual suicide attempters. The correlation was significant (Pearsons’s r ⫽ .40, p ⫽ .001). (B) CSF ln IL-6 plotted against CSF homovanillic acid (HVA). Circles represent the individual suicide attempters. The correlation was significant (Pearsons’s r ⫽ .45, p ⬍ .001).

www.sobp.org/journal

D. Lindqvist et al. central injections of IL-1␤ and IL-6 leads to depressive-like symptoms, as evaluated using the forced swim and tail suspension tests (39,40). Both IL-6 and TNF-␣ correlated positively with 5-HIAA (a metabolite of serotonin) and HVA (a metabolite of dopamine) in the CSF of violent suicide attempters. This points to a potential mechanism by which proinflammatory cytokines may affect suicidal behavior. Previous studies show that increased monoamine metabolism is related to aggression (41,42), although lower levels of CSF 5-HIAA and HVA have been found in violent suicide attempters and violent offenders (43– 46). These findings may reflect different stages in a dynamic process. Longitudinal studies are warranted to study potential fluctuations in CSF monoamine and cytokine levels and their relation to behavior. We hypothesize that the changes in monoamine metabolism observed in suicide attempters could depend on episodes of low-grade CNS inflammation. Proinflammatory cytokines may also activate the hypothalamic-pituitary-adrenal (HPA) axis (47), which in turn modulates monoamines in the CNS (48). Because changes in cortisol levels is a common finding in depressed and suicidal patients (49,50), the interactions between cytokines, monoamines, and the hypothalamic-pituitary-adrenal axis will be of great interest to elucidate in future studies. In conclusion, we found that CSF IL-6 is significantly elevated in suicide attempters compared with healthy control subjects. We observed the highest levels of IL-6 in the CSF of patients with MDD and those who had made a violent suicide attempt. CSF IL-6 was related to the severity of depressive and suicidal symptoms. Furthermore, we show that proinflammatory cytokines are associated with monoamine metabolite levels in violent suicide attempters. Because plasma levels of IL-6 and the blood-brain barrier status did not have an impact on CSF IL-6, we propose a possible role for neuroinflammation in suicidal behavior. This study was supported by the Swedish Research Council (Grant No. 14548); the Lundbeck, Sjöbring and O.M. Persson Foundations; and the province of Scania state grants (ALF). None of these organizations had any further role in the study or in the decision to submit the work for publication. All authors report no biomedical financial interests or potential conflicts of interest. 1. Smith RS (1991): The macrophage theory of depression. Med Hypotheses 35:298 –306. 2. Berk M, Wadee AA, Kuschke RH, O’Neill-Kerr A (1997): Acute phase proteins in major depression. J Psychosom Res 43:529 –534. 3. Thomas AJ, Davis S, Morris C, Jackson E, Harrison R, O’Brien JT (2005): Increase in interleukin-1beta in late-life depression. Am J Psychiatry 162: 175–177. 4. Hestad KA, Tonseth S, Stoen CD, Ueland T, Aukrust P (2003): Raised plasma levels of tumor necrosis factor alpha in patients with depression: Normalization during electroconvulsive therapy. J ECT 19:183–188. 5. Mikova O, Yakimova R, Bosmans E, Kenis G, Maes M (2001): Increased serum tumor necrosis factor alpha concentrations in major depression and multiple sclerosis. Eur Neuropsychopharmacol 11:203–208. 6. Mendlovic S, Mozes E, Eilat E, Doron A, Lereya J, Zakuth V, et al. (1999): Immune activation in non-treated suicidal major depression. Immunol Lett 67:105–108. 7. Kim YK, Lee SW, Kim SH, Shim SH, Han SW, Choi SH, et al. (2008): Differences in cytokines between non-suicidal patients and suicidal patients in major depression. Prog Neuropsychopharmacol Biol Psychiatry 32:356 –361. 8. Postolache TT, Stiller JW, Herrell R, Goldstein MA, Shreeram SS, Zebrak R, et al. (2005): Tree pollen peaks are associated with increased nonviolent suicide in women. Mol Psychiatry 10:232–235.

BIOL PSYCHIATRY 2009;66:287–292 291 9. Timonen M, Viilo K, Hakko H, Sarkioja T, Meyer-Rochow VB, Vaisanen E, et al. (2004): Is seasonality of suicides stronger in victims with hospital-treated atopic disorders? Psychiatry Res 126:167–175. 10. McGeer PL, McGeer EG (1995): The inflammatory response system of brain: Implications for therapy of Alzheimer and other neurodegenerative diseases. Brain Res Brain Res Rev 21:195–218. 11. Steiner J, Bielau H, Brisch R, Danos P, Ullrich O, Mawrin C, et al. (2008): Immunological aspects in the neurobiology of suicide: elevated microglial density in schizophrenia and depression is associated with suicide. J Psychiatr Res 42:151–157. 12. Tonelli LH, Stiller J, Rujescu D, Giegling I, Schneider B, Maurer K, et al. (2008): Elevated cytokine expression in the orbitofrontal cortex of victims of suicide. Acta Psychiatr Scand 117:198 –206. 13. Levine J, Barak Y, Chengappa KN, Rapoport A, Rebey M, Barak V (1999): Cerebrospinal cytokine levels in patients with acute depression. Neuropsychobiology 40:171–176. 14. Carpenter LL, Heninger GR, Malison RT, Tyrka AR, Price LH (2004): Cerebrospinal fluid interleukin (IL)-6 in unipolar major depression. J Affect Disord 79:285–289. 15. Rifflet H, Vuillemin E, Oberti F, Duverger P, Laine P, Garre JB, et al. (1998): [Suicidal impulses in patients with chronic viral hepatitis C during or after therapy with interferon alpha]. Gastroenterol Clin Biol 22:353–357. 16. Nickel T, Sonntag A, Backmund M, Pollmacher T (2005): Depression during therapy with interferon alpha— how long should an antidepressant treatment last? Pharmacopsychiatry 38:102–104. 17. Muller N, Schwarz MJ, Dehning S, Douhe A, Cerovecki A, GoldsteinMuller B, et al. (2006): The cyclooxygenase-2 inhibitor celecoxib has therapeutic effects in major depression: Results of a double-blind, randomized, placebo controlled, add-on pilot study to reboxetine. Mol Psychiatry 11:680 – 684. 18. Dunn AJ (2006): Effects of cytokines and infections on brain neurochemistry. Clin Neurosci Res 6:52– 68. 19. Dunn AJ (1992): Endotoxin-induced activation of cerebral catecholamine and serotonin metabolism: Comparison with interleukin-1. J Pharmacol Exp Ther 261:964 –969. 20. Wang J, Dunn AJ (1998): Mouse interleukin-6 stimulates the HPA axis and increases brain tryptophan and serotonin metabolism. Neurochem Int 33:143–154. 21. Mann JJ (2003): Neurobiology of suicidal behaviour. Nat Rev Neurosci 4:819 – 828. 22. American Psychiatric Association (1987): Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. Washington, DC: American Psychiatric Press. 23. Montgomery SA, Asberg M (1979): A new depression scale designed to be sensitive to change. Br J Psychiatry 134:382–389. 24. Stanley B, Traskman-Bendz L, Stanley M (1986): The suicide assessment scale: A scale evaluating change in suicidal behavior. Psychopharmacol Bull 22:200 –205. 25. First MB (1997): Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I). Washington, DC: American Psychiatric Press. 26. First MB (1997): Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II). Washington, DC: American Psychiatric Press. 27. Traskman-Bendz L, Ekman R, Regnell G, Ohman R (1992): HPA-related CSF neuropeptides in suicide attempters. Eur Neuropsychopharmacol 2:99 –106. 28. Swahn CG, Sandgarde B, Wiesel FA, Sedvall G (1976): Simultaneous determination of the three major monoamine metabolites in brain tissue and body fluids by a mass fragmentographic method. Psychopharmacology (Berl) 48:147–153. 29. Engstrom G, Alling C, Blennow K, Regnell G, Traskman-Bendz L (1999): Reduced cerebrospinal HVA concentrations and HVA/5-HIAA ratios in suicide attempters. Monoamine metabolites in 120 suicide attempters and 47 controls. Eur Neuropsychopharmacol 9:399 – 405. 30. Bayard-Burfield L, Alling C, Blennow K, Jonsson S, Traskman-Bendz L (1996): Impairment of the blood-CSF barrier in suicide attempters. Eur Neuropsychopharmacol 6:195–199. 31. Blennow K, Fredman P, Wallin A, Gottfries CG, Karlsson I, Langstrom G, et al. (1993): Protein analysis in cerebrospinal fluid. II. Reference values derived from healthy individuals 18-88 years of age. Eur Neurol 33:129 –133. 32. Sluzewska A, Rybakowski J, Bosmans E, Sobieska M, Berghmans R, Maes M, et al. (1996): Indicators of immune activation in major depression. Psychiatry Res 64:161–167.

www.sobp.org/journal

292 BIOL PSYCHIATRY 2009;66:287–292 33. Hiscock N, Chan MH, Bisucci T, Darby IA, Febbraio MA (2004): Skeletal myocytes are a source of interleukin-6 mRNA expression and protein release during contraction: Evidence of fiber type specificity. FASEB J 18:992–994. 34. Keller C, Keller P, Marshal S, Pedersen BK (2003): IL-6 gene expression in human adipose tissue in response to exercise— effect of carbohydrate ingestion. J Physiol 550:927–931. 35. Maes M (1995): Evidence for an immune response in major depression: A review and hypothesis. Prog Neuropsychopharmacol Biol Psychiatry 19:11–38. 36. Banks WA, Kastin AJ, Broadwell RD (1995): Passage of cytokines across the blood-brain barrier. Neuroimmunomodulation 2:241–248. 37. Raber J, O’Shea RD, Bloom FE, Campbell IL (1997): Modulation of hypothalamic-pituitary-adrenal function by transgenic expression of interleukin-6 in the CNS of mice. J Neurosci 17:9473–9480. 38. Chourbaji S, Urani A, Inta I, Sanchis-Segura C, Brandwein C, Zink M, et al. (2006): IL-6 knockout mice exhibit resistance to stress-induced development of depression-like behaviors Neurobiol Dis 23:587–594. 39. Dunn AJ, Swiergiel AH (2005): Effects of interleukin-1 and endotoxin in the forced swim and tail suspension tests in mice. Pharmacol Biochem Behav 81:688 – 693. 40. Wu TH, Lin CH (2008): IL-6 mediated alterations on immobile behavior of rats in the forced swim test via ERK1/2 activation in specific brain regions. Behav Brain Res 193:183–191. 41. Engelborghs S, Vloeberghs E, Le Bastard N, Van Buggenhout M, Marien P, Somers N, et al. (2008): The dopaminergic neurotransmitter system is associated with aggression and agitation in frontotemporal dementia. Neurochem Int 52:1052–1060.

www.sobp.org/journal

D. Lindqvist et al. 42. Alia-Klein N, Goldstein RZ, Kriplani A, Logan J, Tomasi D, Williams B, et al. (2008): Brain monoamine oxidase A activity predicts trait aggression. J Neurosci 28:5099 –5104. 43. Traskman L, Asberg M, Bertilsson L, Sjostrand L (1981): Monoamine metabolites in CSF and suicidal behavior. Arch Gen Psychiatry 38:631– 636. 44. Placidi GP, Oquendo MA, Malone KM, Huang YY, Ellis SP, Mann JJ (2001): Aggressivity, suicide attempts, and depression: Relationship to cerebrospinal fluid monoamine metabolite levels. Biol Psychiatry 50:783–791. 45. Traskman-Bendz L, Alling C, Oreland L, Regnell G, Vinge E, Ohman R (1992): Prediction of suicidal behavior from biologic tests. J Clin Psychopharmacol 12:21S–26S. 46. Virkkunen M, Eggert M, Rawlings R, Linnoila M (1996): A prospective follow-up study of alcoholic violent offenders and fire setters. Arch Gen Psychiatry 53:523–529. 47. O’Brien SM, Scott LV, Dinan TG (2004): Cytokines: abnormalities in major depression and implications for pharmacological treatment. Hum Psychopharmacol 19:397– 403. 48. Chaouloff F (2000): Serotonin, stress and corticoids. J Psychopharmacol 14:139 –151. 49. Bao AM, Meynen G, Swaab DF (2008): The stress system in depression and neurodegeneration: Focus on the human hypothalamus. Brain Res Rev 57: 531–553. 50. Lindqvist D, Isaksson A, Traskman-Bendz L, Brundin L (2008): Salivary cortisol and suicidal behavior—a follow-up study. Psychoneuroendocrinology 33:1061–1068.