Clinical Psychopharmacology and Neuroscience 2019; 17(1): 34-42
Values of Cytokines and Tryptophan Metabolites over a 12 Weeks Time Course in Patients with Depression and Somatoform Disorder
Daniela Krause1,*, Verena B. Kirnich2,*, Theresa M. Stapf2, Anika Hennings3, Sabine Riemer3, Michael Riedel2, Ralf Schmidmaier4, Francisco Pedrosa Gil5, Winfried Rief3, and Markus J. Schwarz2
1Department of Psychiatry and Psychotherapy, University Hospital, LMU Munich, Germany, 2Department of Laboratory Medicine, Ludwig Maximilians University, Munich, Germany, 3Division of Clinical Psychology and Psychotherapy, Philipps University of Marburg, Marburg, Germany, 4Department of Hematology and Oncology, Ludwig Maximilians University, Munich, Germany, 5Department of Psychiatry, Vogtland Hospital, Plauen, Germany
Correspondence to: Address for correspondence: Daniela Krause, MD, Department of Psychiatry, Ludwig Maximilians University, Nußbaumstr. 7, 80336 Munich, Germany, Tel: +49-89-4400-55511, Fax: +49-89-4400-73240, E-mail:, ORCID:
Received: April 24, 2017; Revised: August 31, 2017; Accepted: October 24, 2017; Published online: February 28, 2019.
© The Korean College of Neuropsychopharmacology. All rights reserved.

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.


Previous studies have suggested alterations in the kynurenine pathway as a major link between cytokine and neurotransmitter abnormalities in psychiatric disorders. Most of these studies used a cross-sectional case-control study design. However, knowledge is still lacking regarding the stability over time of kynurenine pathway metabolites and the functionally related cytokines. Therefore, we studied the stability of cytokines and tryptophan (TRP) parameters over a period of 12 weeks.


A total of 117 participants–39 with major depression, 27 with somatoform disorder, and 51 healthy controls–were enrolled. Four evaluations, including blood withdrawal and psychometric testing, were performed over a period of 12 weeks. We used ELISA to measure interleukin (IL)-6, IL-1 receptor antagonist (RA) and tumor necrosis factor α (TNF α). High-performance liquid chromatography was used to analyze neurotransmitter variables, i.e. TRP, 5-hydroxyindoleacetic acid (5-HIAA), kynurenine (KYN), 3-OH-kynurenine (3-HK), and kynurenic acid (KYNA).


We found no significant fluctuations of TRP, its metabolites (5-HIAA, KYN, KYNA, and 3-HK), or the cytokines (IL-1RA, IL-6, and TNF α) in any of the groups over the 12 weeks.


To our knowledge, this is the first longitudinal study performed in psychiatric patients to verify the stability and consequently the reliability of the biological parameters we investigated. Our data indicate that TRP metabolites and cytokines are reliable biological parameters in psychiatric research because they do not fluctuate significantly over time.

Keywords: Cytokines, Tryptophan, Depression, Somatoform disorders, Kynurenine, Kynurenic acid

In the field of psychoneuroimmunological research, evidence has been accumulating for reciprocal communication pathways between the nervous, endocrine, and immune systems.1) As a result, there is growing interest in whether and in which way these interactions are involved in the development of psychiatric disorders. Proinflammatory cytokines, as described recurrently in major depression (MD),24) are supposed to be related to an enhanced degradation of plasma tryptophan (TRP) via the kynurenine pathway.5,6) There is robust evidence that interactions of the immune system and serotonergic system, via indolamine 2,3-dioxygenase, the rate-limiting enzyme in the kynurenine pathway, play a key role in MD.7) Not only depression, also Alzheimer’s disease8) and especially somatoform disorder (SFD) has been shown to be associated with specific biomarkers of inflammation and cell-mediated immune activation, that are correlated with changes in the TRP catabolite pathway.9) Taken these findings together, it is further strengthened that depression and SFD are two frequently overlapping clinical conditions having both been related to immunological changes.10)

Recently, psychobiological studies in depression and SFD have had a huge impact on research. Especially depression research has lately very much focused on the pro- and anti-inflammatory trias of interleukin (IL)-6, tumor necrosis factor (TNF) α and IL-1 receptor antagonist (RA), as well as on the investigation of kynurenine pathway alterations with predominance of the neurotoxic pathway.11) However, the vast majorities of studies followed a cross-sectional case-control design and thus were unable to prove the stability of the investigated parameters over time. Therefore, the main aim of our study was to measure the most relevant biological features of depression and somatization in a longitudinal study. To our knowledge data about the longitudinal course and stability of pro-inflammatory cytokines and TRP metabolites is rare in patients suffering from MD and SFD and compared to healthy controls (HC). One recent study investigated the cytokines IL-6 and IL-18 already twice (baseline two weeks post stroke and then again one year later) in post-stroke depression. It was concluded that higher IL-6 and IL-18 values were associated with depressive disorders within two weeks and at one year after stroke.12) Furthermore there is another study that measured IL-6, TNF α and high sensitivity C-reactive protein (hsCRP) twice over the course of five years and calculated the relationship to depressive symptoms in men. For the two cytokines none was significantly associated with depression, only for hsCRP an association to depressive symptoms could be identified.13) However, knowledge about the long-term stability of these biological parameters is essential to validate the findings of previous and future cross-sectional studies Therefore, the main aim of the present study was to investigate the stability of the pro-inflammatory cytokines IL-6 and TNF α and the anti-inflammatory IL-1 RA in MD, SFD, and HC over a period of 12 weeks. Because of the known interactions between cytokines and neurotransmitter systems, we assessed the stability of both cytokines and TRP metabolites, thereby focusing on the kynurenine pathway.



A total of 117 subjects, comprising 39 outpatients with depression, 27 outpatients with SFD (defined according to Somatoform Symptom Index [SSI] 6 out of 8), and 51 age-matched HC, were recruited by public advertisements, press releases in local newspapers, and leaflets in clinics, pharmacies, and universities. This study was part of an experimental approach to investigate the effect of physical exercise on symptom severity in MD and SFD.14,15) The study protocol was approved by the ethics committees of the Ludwig Maximilian University, Munich, and the German Psychological Society. All participants provided written informed consent. As an incentive, all participants received a small financial compensation for completing the study. All participants were pre-screened in a short telephone interview. The effect size of the study was calculated according to Cohen’s d. For medium effect sizes at a 3×2×2×2×6 factor design (group×sex×intervention×order×time point) and an error probability of p< 0.05 a total of n=108 study participants is needed. During the 12 week course of the study the following drop-out rates were observed: 3 people in MD group, 2 people in the SFD group and 4 people in HC.

Psychological Measures

Suitable participants took part in a structured clinical interview (SCID; German version)16) for Diagnostic and Statistical Manual of Mental Disorders of the American Psychiatric Association, fourth edition (DSM-IV) to confirm present and lifetime diagnoses as well as co-morbid Axis I disorders. Interviewers at both of the study centers were physicians or psychologists familiar with conducting SCID and experienced in making psychiatric diagnoses. Sociodemographic variables, like body mass index (BMI) and physical activity (assessed with the “Freiburger Fragebogen zur körperlichen Aktivität,” FFKA [Freiburger Questionnaire on physical activity]), were recorded because of their possible influence on biological parameters. Further questionnaires about personality (SCID II) and social support (F-Sozu [German Questionnaire on Social Support]) were completed. Exclusion criteria were organic illnesses affecting the central nervous system or immune system, pregnancy, psychotic symptoms, alcohol or drug abuse, and current psychotherapy. During the study, we excluded also patients who developed a medical condition, e.g. a common cold. To reduce the number of missing values, we shifted the measurement for 4 weeks, so that women were again in the same period of their cycle. For the patient groups, co-morbidity of MD and SFD was an additional exclusion criterion. Furthermore, a weak somatization syndrome with four (men) or six (women) symptoms (SSI-4/6) was an exclusion criterion in the depression group. Oral contraceptives were permitted, but we excluded persons taking other immune-modulating medication (e.g., corticosteroids). A past or present psychiatric diagnosis according to DSM-IV was an exclusion criterion in the HC group.

SFD is characterized by bodily symptoms that cannot be explained by a clear organic pathology or known physiological mechanisms. Typically somatoform symptoms are long-lasting and disabling.17,18) We used the SSI-6/8 to define SFD, because of previous claims that the classification of SFDs according to DSM-IV is too strict and does not give a good representation of the relevance to healthcare of patients with multiple somatoform symptoms.19) By using an empirical approach to study the definition of SFD, Rief and Hiller20) found that a cut-off score of seven to eight symptoms is valid for differentiating between patients with high and low levels of disability. The SSI-6/8 requires that patients have at least eight (men only six) of the 33 complaints relevant for the classification of SFD according to DSM-IV.

MD was diagnosed strictly according to DSM-IV criteria, i.e. the presence of a period of at least 2 weeks of depressed mood or anhedonia or both in combination with at least five symptoms like somatic disturbances (change in sleep, appetite, energy level, and psychomotor functioning) and cognitive disturbances (feeling of inappropriate guilt, worthlessness, trouble concentrating and making decisions, and suicidal thoughts) that cause clinical significant distress or impairment in everyday functioning.16)


After completion of the diagnostic session, we arranged the first blood withdrawal (t1). Blood samples were taken between 8 AM and 9 AM exactly every four, eight, and twelve weeks thereafter (t2–t4). Patients were instructed to avoid alcohol and exercise for 24 hours before blood withdrawal. After blood sampling participants completed the following self-rating scales: Beck Depression Inventory (BDI) to assess dpression severity,21) Screening for Somatoform Symptoms-7 (SOMS-7) to examine somatoform symptom severity, and Symptom Checklist-90-Revised (SCL-90-R) to detect global psychopathological impairment.22) Nutrition and sleep were assessed with protocols to control for potential sleep disturbances and nutrition abnormalities that could affect cytokine and neurotransmitter levels. At t1 and t4, the level of chronic stress was measured with the Trier Inventory for Chronic Stress in its Shortform (TICS-K), and at t4, somatoform symptoms during the last two years were assessed with the SOMS-2.23) Because of the reported influence of physical activity on the immune system, we assessed the participants’ typical amount of physical activity at the onset of the study with the FFKA.24)

Analysis of blood samples

Blood samples (two 7.5 ml serum tubes) were collected between 8 AM and 9 AM. Immediately after collection, samples were allowed to clot for 30 minutes and then centrifuged; serum was stored at −80°C until being thawed for assay. All blood samples were assayed in a single run at the end of the data collection.

Analysis of TRP metabolites with high performance liquid chromatography

High performance liquid chromatography (HPLC; Waters Co., Milford, MA, USA) was used to measure serum TRP, 5-hydroxyindoleacetic acid (5-HIAA), kynurenine (KYN), 3-OH-kynurenine (3-HK), and kynurenic acid (KYNA). All TRP metabolites were measured by a previously described HPLC method.25) To avoid operator bias, tubes were coded only with numbers so that the laboratory technician was blinded to group.

Analysis of cytokines with quantitative enzyme-linked immunosorbent assay

The serum samples were analyzed for the cytokines IL-1 RA, IL-6, and TNF α. All assays were carried out with quantitative enzyme-linked immunosorbent assay (ELISA) kits in accordance with the manufacturers’ instructions (Human IL-6 Quantikine HS, Human TNF α Quantikine HS; Human IL-1 RA Quantikine, all from R&D Systems, Minneapolis, MN, USA). A highly sensitive ELISA was used for IL-6 and TNF α. According to the manufacturer, the intra-assay coefficient of variation was between 6.9% and 7.8% for IL-6, between 3.1% and 8.5% for TNF α, and between 3.7% and 7.3% for IL-1 RA, with a minimum detectable amount of 0.039 pg/ml for IL-6, 0.106 pg/ml for TNF α, and 6.26 pg/ml for IL-1 RA.

Statistical Analysis

SPSS Statistics software (version 23; IBM Corp., Armonk, NY, USA) was used for statistical procedures. The chi-square test was used to check for potential demographic group differences in categorical variables; ANOVA or Kruskal-Wallis test, for potential demographic group differences in ordinal and interval variables. Further, all data were tested for normal distribution by plotting the values and with the Kolmogorov-Smirnov test. Since most psychometric variables and immunological parameters were distributed non-parametrically, we used mainly non-parametric tests. We applied the Friedman rank sum test and Spearman’s correlations to analyze the stability of cytokines and TRP metabolites during the 12 weeks. A Pearson product-moment correlation coefficient was computed to assess the relationship between the kynurenine system and cytokines. Interactions between biological parameters and socio-demographic characteristics and between biological parameters and psychometric results were analyzed with Spearman rank coefficients. A p< 0.05 was considered as statistically significant. Because of the explorative character of our study, we did not use a Bonferroni adjustment.26)


Sample Characteristics

Thirty-nine patients with MD (age, 32.6±12.6 years), 27 patients with SFD (age, 38.8±14.29 years), and 51 HC (age, 36.51±13.18 years) were included in the study. In all groups, more than 60% of the participants were female and over 72% had at least high school education. The groups did not differ in terms of age, sex distribution, smoking, BMI, physical activity, or nutrition (kcal and proteins). In total, 23.1% of patients with MD were taking antidepressants (mainly SSRIs). The SFD group was treated only in 14.8% with medication (antidepressants, lithium and anticonvulsant). The most frequent comorbidities were social phobia and panic disorder.

Detailed characteristics of the study sample are presented in Table 1.

Longitudinal Examination of Cytokines and TRP Metabolites

The Friedman rank sum test demonstrated stability of all investigated biological parameters over the course of 12 weeks: significant fluctuations were not observed in either of the patient groups or in the HC group. Detailed results for TRP and its metabolites are presented in Table 2. The immune markers IL-6, TNF α, and IL-1 RA were also stable over the 12 weeks (Table 3). To check for unsystematic variations, we also performed Spearman’s correlations between the single measurements. These tests (Table 4) also showed no fluctuations of the investigated parameters. Further, we performed a subgroup analysis of the parameters to verify their stability (in females and in participants without psychiatric medication), which were in line with our results (data not shown). Box plot analysis also showed high stability. Only the HC group contained sporadic outliers in the cytokines IL-1 RA and IL-6 as well as in the serotonin degradation product 5-HIAA. Detailed results are shown in Tables 2 to 4.


As expected, somatization scores were higher in SFD patients and depression scores higher in MD patients. Statistical analysis of BDI scores showed significant group differences between the patient groups and HC at all time points. A comparison of patient groups with the t test revealed significant differences in depressive symptoms measured with the BDI at evaluations t1, t2, and t3; differences at t4 showed a trend towards significance. Statistical analysis of somatic symptoms (SOMS-7 sum) showed significant group differences between patient groups and HC at all evaluations. A comparison of SOMS-7 scores between patient groups revealed significant differences at t1 and t3; differences showed a trend towards significance at t2, but at t4 there was no such trend. SCL-90-R scores did not differ significantly between patient groups, although there was a trend towards higher scores in MD. Psychometric results are shown in detail in Table 5. Analysis of the psychometric scores with the Friedman rank sum test showed clear fluctuations over the 12 weeks in each of the groups. The only stable parameter was SOMS-7 in MD (p=0.62) and HC (p=0.55).

Relationship between the Kynurenine System and Cytokines

There was a positive correlation between the kynurenine/TRP ratio and TNF α (r=0.297, n=84, p=0.007). For the other investigated cytokines IL-1 RA (r=0.177, n=84, p=0.107) and IL-6 (r=0.167, n=84, p=0.129), no correlation to kynurenine/TRP ratio was found. Overall, increases in the KYN/TRP ratio were correlated with increases in TNF α.

Correlations between Biological Parameters and Psychometric Test Results

We performed a correlation analysis between the biological parameters and BDI, SCL-90-R, SOMS-7, and TICS-K. None of the investigated groups showed a stable association between a biological parameter and a psychometric variable at any of the assessment points.


The present study aimed to assess whether previous inconclusive psychobiological findings could be due to fluctuations of neurotransmitter and cytokine levels. In our study sample, we did not find significant fluctuations that might explain earlier incongruent psychobiological findings. The pro-inflammatory cytokines IL-6 and TNF α and the anti-inflammatory cytokine IL-1 RA were stable in all three groups (MD, SFD, and HC) over the 12-week study period, as were TRP, its kynurenine metabolites, and 5-HIAA. Concerning the stability of cytokines, a study in dialysis patients with chronic kidney disease, another illness associated with chronic inflammation, showed high fluctuations in IL-6 over a period of 12 weeks.27) But a study on women with obesity showed a good within stability in a period of 3 years.28) Previous studies have often only measured cytokines at one time point, whereas the psychopathology was assessed over years. Zalli et al.29) showed that IL-6 and CRP levels at baseline could predict the persistence of depressive symptoms over 5 years.

We assume that differences might rather be related to characteristics of the study sample, like illness severity, illness duration, illness classification, co-morbidities, or sociodemographic parameters, i.e. BDI or physical activity level.3033) IL-1 RA has been found to be strongly elevated in people with obesity.34) Accordingly the SFD group showed a significant correlation between BMI and both IL-6 and IL-1 RA. HC showed correlations between IL-6 and both BMI and age. This finding is concordant with that of an epidemiologic study on cytokines in the Swiss population, which showed that age and BMI correlate positively with IL-6 and TNF α. Further, the study showed higher TNF α in men who were smoking, but did not demonstrate any effect of physical activity in leisure time on cytokines like TNF α and IL-6.35) In our study, we could also show an effect of smoking on TNF α in HC. Furthermore TNF α was the only cytokine that was positively correlated with KYN/TRP ratio that indicates the activity of the first degrading enzyme of TRP to kynurenine. This finding is in line with previous studies suggesting that increased TNF α correlates with increased monocyte, macrophage and microglial activation and therefore with the activity of the enzyme that breaks down TRP to kynurenine.36)

In addition to examining the biological parameters, we simultaneously measured the psychological state of health. We did not find any persistent correlations between biological parameters and depression (BDI), somatoform symptoms (SOMS-7), or global psychopathological impairment (SCL-90-R). But it has to be taken into account that immunological research is susceptible to artifacts and interfering variables.

This study has some limitations that might affect the generalizability of our results. First, the sample size of 117 (51 HC, 39 MD, and 27 SFD) was relatively small. Although the smallest group was that of patients with SFD, this group showed the highest stability of the investigated biological parameters. Second, patients in both patient groups were treated with antidepressants, which could have affected levels of neurotransmitters and cytokines. Moreover there are parameters in the kynurenine pathway, like quinolinic acid, which were not analysed in this study.

One strength of our study is the methodological approach; during the study period of 12 weeks, we specified an exact timeframe of four weeks between the measurements to consider possible effects of the menstrual cycle. Second, we excluded patients from the measurement if they had a common cold or any other medical condition. Third, we examined the interactions of cytokines and TRP-related neuroactive amines, instead of assessing only a single biological system. Fourth, we strictly separated MD and SFD, even though the two have a high co-morbidity rate. Fifth, we recorded a large number of confounding factors.

Although many cross-sectional studies have focused on the “cytokine hypothesis” of depression as well as on its association with the monoamine hypothesis, in particular the kynurenine pathway, to our knowledge this study is the first to examine the longitudinal stability of TRP metabolites and cytokines in psychiatric patients in comparison to HC. Our data indicate that TRP metabolites and cytokines are reliable biological parameters in psychiatric research, because they do not fluctuate significantly in the course of time. This is also particular important because recent research found associations between depression and structural brain damage with increased risk for neurodegenerative illnesses like Alzheimer disease.37) This validates biological studies performed so far and will help to assess future studies in this field.

In our opinion, the major finding of our study is that fluctuations in cytokines such as IL-6, TNF α, and IL-1 RA, and in TRP and its metabolites, 5-HIAA, 3-HK, and KYNA, are to a large extent negligible in MD and SFD.


This study was funded by a grant from the German Research Foundation (DFG). The DFG had no further role in the planning or execution of the study.

The authors thank Jacquie Klesing, ELS, for editing assistance with the manuscript. Further, the authors express their gratitude to all the participants for their great commitment to the study. The authors thank Johanna Zach for performing the analyses, Anina Burgbacher, Katharina Jacobi, and Benedikt Bedenk for data management, and Sebastian Meyer for statistical advice.


Characteristics of patients with somatoform disorder (SFD) or major depression (MD) and healthy controls (HC)

Characteristic SFD group (n=27) MD group (n=39) HC group (n=51) Significance level
Sociodemographic data
 Age (yr) 38.8±14.29 32.64±12.58 36.51±13.18 F=1.88; NS
 Women 20 (74.1) 24 (61.5) 33 (64.7) Chi2=1.16; NS
 High school education 18 (66.7) 30 (76.9) 43 (84.3) Chi2=5.59; NS
 Occupation 24 (88.9) 36 (92.3) 43 (82.4) Chi2=0.50; NS
 Smoking status 4 (14.8) 15 (38.5) 16 (31.4) Chi2= 4.62; NS
 Adequate physical activity 11 (40.7) 12 (30.8) 26 (51.0) Chi2= 8.67; NS
 Antidepressants 4 (14.8) 9 (23.1) 0 (0) Chi2=0.13; NSa
 Body mass index (kg/m2) 24.11±4.80 24.72±4.83 24.05±5.47 Chi2=0.97; NS
Psychiatric comorbidity
 Generalized anxiety disorder 2 (7.4) 2 (5.1) - Chi2=0.12; NSa
 Obsessive-compulsive disorder 3 (11.1) 1 (2.6) - Chi2=1.94; NSa
 Panic disorder 4 (14.8) 6 (15.4) - Chi2=0.01; NSa
 Posttraumatic stress disorder 0 (0) 5 (12.8) - Chi2=3.79; NSa
 Social phobia 4 (14.8) 7 (17.9) - Chi2=0.14; NSa
 Specific phobia 2 (7.4) 2 (5.1) - Chi2=0.12; NSa
Psychopathological measures
 F-Sozu 59.5±9.3 56.8±12.95 65.2±8.0 Chi2=13.8; p=0.001***MD vs. SSI: T=0.69; NS
Depression, BDI 13.78±8.57 21.63±9.52 3.06±4.25 Chi2=69.16; p=0.000***
Psychological distress, SCL 90-R, GSI 0.78±0.52 1.01±0.50 0.13±0.17 Chi2=60.18; p=0.000***
Somatization severity indices
 Total SOMS-7 29.31±17.45 8.00±9.52 5.86±5.63 Chi2=47.90; p=0.000***
 Total SOMS-2 26.00±9.93 23.88±9.52 5.74±7.68 Chi2= 107.33; p=0.001***
Stress Evaluation
 TICS-K 21.73±9.62 26.43±7.76 9.53±6.48 Chi2=103.48; p=0.009**

Values are presented as mean±standard deviation or number (%).

Group differences were calculated with analyses of variance and chi-square tests as appropriate.

NS, not significant; SSI, Somatoform Symptom Index; BDI, Beck Depression Inventory; SCL-90 R, Symptom Check List-90 Revised; GSI, Global Severity Index; SOMS, Screening for Somatoform Symptoms; TICS-K, Trier Inventory for Chronic Stress in its Shortform.

The right column includes significant results (p<0.05);




comparison between patient groups.

Longitudinal analysis of tryptophan and kynurenine metabolites

Biological parameter Group t1 t2 t3 t4 Friedman rank sum test
Tryptophan (ng/mg) MD (n=37) 16.80±3.98 16.82±4.60 16.93±4.13 16.17±3.83 Chi2=1.24; df=3; p=0.74
SFD (n=26) 15.78±3.00 15.67±3.38 15.33±3.04 15.09±2.85 Chi2=1.99; df=3; p=0.58
HC (n=43) 16.85±4.59 16.92±4.41 16.84±4.65 16.17±3.55 Chi2=1.80; df=3; p=0.62
Total (n=106) 16.57±4.03 16.58±4.24 16.50±4.14 15.91±3.50 Chi2=3.52; df=3; p=0.32
5-HIAA (ng/mg) MD (n=37) 2.66±3.03 2.75±3.34 2.67±3.11 2.45±3.17 Chi2=1.45; df=3; p=0.69
SFD (n=25) 4.68±6.35 6.10±10.65 4.21±5.25 3.81±4.59 Chi2=5.68; df=3; p=0.13
HC (n=43) 3.97±5.30 5.35±13.08 4.96±13.31 3.52±5.76 Chi2=1.02; df=3; p=0.80
Total (n=105) 3.68±4.95 4.61±10.06 3.97±9.07 3.22±4.70 Chi2=4.21; df=3; p=0.24
Kynurenine (ng/mg) MD (n=37) 433.17±93.36 440.55±102.08 442.24±91.84 436.23±111.49 Chi2=0.80; df=3; p=0.85
SFD (n=26) 458.69±151.51 435.04±110.75 448.32±140.24 435.78±124.36 Chi2=1.25; df=3; p=0.74
HC (n=43) 464.70±107.99 455.28±91.93 456.89±96.25 459.58±103.12 Chi2=0.29; df=3; p=0.96
Total (n=106) 452.22±115.42 445.18±99.74 449.67±106.40 445.59±111.05 Chi2=1.19; df=3; p=0.76
3-OH-kynurenine (ng/mg) MD (n=37) 8.64±3.20 8.47±3.0 9.20±3.4 8.75±4.26 Chi2=3.38; df=3; p=0.34
SFD (n=25) 10.47±5.10 10.17±4.02 9.98±4.17 10.54±5.40 Chi2=0.16; df=3; p=0.98
HC (n=43) 9.65±3.84 9.73±3.70 9.82±4.16 9.69±4.03 Chi2=0.50; df=3; p=0.92
Total (n=105) 9.35±3.88 9.40±3.57 9.59±3.83 9.54±4.50 Chi2=2.07; df=3; p=0.56
Kynurenic acid (ng/mg) MD (n=37) 11.06±10.15 11.51±10.02 11.72±6.83 12.68±13.60 Chi2=3.67; df=3; p=0.30
SFD (n=25) 11.02±4.87 9.86±5.49 10.26±5.19 10.39±4.93 Chi2=2.13; df=3; p=0.55
HC (n=43) 10.41±6.65 10.42±5.30 10.30±5.79 11.18±11.23 Chi2=0.98; df=3; p=0.81
Total (n=105) 10.78±7.69 10.67±7.31 10.79±6.03 11.52±11.02 Chi2=0.18; df=3; p=0.98
Kynurenic acid/kynurenine MD (n=37) 0.026±0.028 0.026±0.022 0.026±0.015 0.031±0.041 Chi2=3.29; df=3; p=0.35
SFD (n=25) 0.025±0.011 0.023±0.012 0.024±0.010 0.024±0.011 Chi2=2.71; df=3; p=0.44
HC (n=43) 0.022±0.013 0.023±0.010 0.023±0.012 0.024±0.017 Chi2=1.72; df=;3 p=0.63
Total (n=105) 0.024±0.019 0.024±0.016 0.024±0.013 0.027±0.027 Chi2=0.19; df=3; p=0.98
3-OH-kynurenine/kynurenine MD (n=37) 0.020±0.01 0.020±0.01 0.021±0.01 0.020±0.01 Chi2=2.68; df=3; p=0.44
SFD (n=25) 0.023±0.01 0.024±0.012 0.023±0.01 0.025±0.014 Chi2=3.67; df=3; p=0.30
HC (n=43) 0.021±0.01 0.021±0.01 0.021±0.01 0.021±0.01 Chi2=3.64; df=3; p=0.30
Total (n=105) 0.021±0.01 0.021±0.01 0.022±0.01 0.022±0.01 Chi2=3.08; df=3; p=0.38
Kynurenine/tryptophan MD (n=37) 27.03±8.45 27.98±10.51 27.45±8.08 28.26±8.98 Chi2=1.35; df=3; p=0.72
SFD (n=26) 30.20±13.36 28.87±10.16 30.72±13.50 30.13±12.84 Chi2=0.37; df=3; p=0.95
HC (n=42) 29.44±10.49 28.44±9.57 28.91±9.66 29.33±8.38 Chi2=1.29; df=3; p=0.73
Total (n=105) 28.78±10.62 28.38±9.96 28.85±10.24 29.16±9.78 Chi2=0.26; df=3; p=0.97

Values are presented as mean±standard deviation.

t1, month 1; t2, month 2; t3, month 3; t4, month 4.

MD, major depression; SFD, somatoform disorder; HC, healthy control; df, degree of freedom.

Longitudinal analysis of cytokines

Biological parameter t1 t2 t3 t4 Friedman rank sum test
IL-1 RA (pg/ml) MD (n=37) 359.9±320.9 416.3±410.5 324.9±224.1 310.4±230.1 Chi2=1.07; df=3; p=0.78
SFD (n=26) 308.1±155.5 330.1±203.3 319.0±143.0 316.9±150.2 Chi2=0.26; df=3; p=0.97
HC (n=45) 346.7±288.5 343.3±204.0 384.9±268.6 391.99±299.2 Chi2=4.7; df=3; p=0.19
Total (n=108) 341.9±273.7 365.1±291.3 348.5±228.7 373.6±295.4 Chi2=0.77; df=3; p=0.86
IL-6 (pg/ml) MD (n=37) 1.17±0.74 1.78±2.57 1.08±0.71 1.12±0.83 Chi2=4.87; df=3; p=0.18
SFD (n=26) 1.55±2.12 1.48±2.17 1.63±2.29 1.38±2.08 Chi2=6.51; df=3; p=0.09
HC (n=45) 1.38±1.59 1.47±1.85 1.55±2.04 1.62±2.00 Chi2=0.66; df=3; p=0.88
Total (n=108) 1.35±1.51 1.58±2.18 1.41±1.78 1.39±1.72 Chi2=2.59; df=3; p=0.46
TNF α (pg/ml) MD (n=37) 1.41±0.44 1.54±0.51 1.48±0.48 1.51±0.63 Chi2=4.28; df=3; p=0.23
SFD (n=26) 1.45±0.76 1.45±0.76 1.46±0.72 1.43±0.65 Chi2=0.70; df=3; p=0.87
HC (n=45) 1.45±1.17 1.36±0.54 1.45±0.75 1.39±0.63 Chi2=1.26; df=3; p=0.74
Total (n=108) 1.44±0.87 1.44±0.59 1.46±0.66 1.44±0.34 Chi2=1.92; df=3; p=0.59

Values are presented as mean±standard deviation.

t1, month 1; t2, month 2; t3, month 3; t4, month 4.

IL, interleukin; RA, receptor antagonist; TNF, tumor necrosis factor; MD, major depression; SFD, somatoform disorder; HC, healthy control; df, degree of freedom.

Correlation coefficients of cytokines and tryptophan metabolites at different time points


t1–t2 t1–t3 t1–t4 t1–t2 t1–t3 t1–t4 t1–t2 t1–t3 t1–t4
Tryptophan 0.41** 0.65** 0.57** 0.70** 0.54** 0.42* 0.81** 0.71** 0.70**
5-HIAA 0.81** 0.88** 0.71** 0.80** 0.96** 0.73** 0.90** 0.90** 0.87**
Kynurenine 0.71** 0.64** 0.64** 0.86** 0.86** 0.88** 0.74** 0.61** 0.55**
3-OH-kynurenine 0.68** 0.70** 0.37* 0.67** 0.79** 0.82** 0.69** 0.71** 0.53**
Kynurenic acid 0.87** 0.77** 0.81** 0.64** 0.82** 0.81** 0.62** 0.76** 0.70**
IL-1-RA 0.62** 0.77** 0.61** 0.63** 0.79** 0.66** 0.74** 0.68** 0.63**
IL-6 0.48** 0.68** 0.67** 0.65** 0.61** 0.73** 0.53** 0.53** 0.66**
TNF α 0.85** 0.86** 0.82** 0.95** 0.88** 0.87** 0.76** 0.74** 0.62**

MD, major depression; SFD, somatoform disorder; HC, healthy controls; t1, begin of examination; t2, examination 4 weeks later; t3, examination 8 weeks later; t4, examination 12 weeks later; 5-HIAA, 5-hydroxyindoleacetic acid; IL, interleukin; RA, receptor antagonist; TNF, tumor necrosis factor.

Spearman coefficient of correlation, r;


Psychometric results at each evaluation

Group t1 t2 t3 t4
BDI MD (n=35–38) 21.63±9.52 18.45±9.23 15.64±10.94 15.24±9.73
SFD (n=25–27) 13.78±8.57 11.62±7.56 19.96±7.07 10.77±8.05
HC (n=45–51) 3.06±4.25 1.77±3.16 1.41±2.93 1.80±3.78
MD vs. SFD T=3.41; p=0.001* T=3.12; p=0.003* T=2.04; p=0.05* T=1.93; p=0.06
SCL-90R MD (n=35–38) 90.57±45.06 81.64±47.57 67.23±46.85 61.95±40.67
SFD (n=25–27) 70.13±47.13 56.59±41.08 51.04±35.61 42.81±35.72
HC (n=45–51) 11.49±15.29 9.69±12.33 8.35±12.43 8.80±12.95
MD vs. SFD T=1.52; p=0.14 T=1.92; p=0.06 T=1.39; p=0.17 T=1.94; p=0.06
SOMS-7 MD total (n=30–37) 19.36±15.00 13.05±16.48 14.21±18.62 18.59±18.91
MD women (n=17–23) 20.86±14.88 22.55±17.97 17.61±12.07 18.61±12.37
MD men (n=12–15) 16.75±15.49 19.62±19.50 14.08±14.54 18.57±27.04
SFD total (n=19–27) 29.31±17.45 21.78±19.70 25.59±19.30 22.16±14.68
SFD women (n=14–19) 34.37±17.23 29.83±18.31 31.47±15.93 24.32±15.54
SFD men (n=5–7) 15.57±8.77 11.71±5.15 17.00±9.95 15.33±9.52
HC total (n=43–45) 5.86±5.63 2.94±6.84 2.78±7.20 4.40±8.74
HC women (n=29–31) 6.09±5.65 4.77±6.62 4.90±6.42 4.14±5.95
HC men (n=13–16) 5.38±5.75 3.60±4.78 3.29±3.25 4.93±13.05
MD vs. SFD T=−2.4; p=0.02* T=−2.0; p=0.055 T=−2.4; p=0.02* T=−0.79; p=0.43

Values are presented as mean±standard deviation.

Group differences were calculated with the t-test;

p<0.05 (statistically significant).

t1, month 1; t2, month 2; t3, month 3; t4, month 4; BDI, Beck Depression Inventory; SCL-90-R, Symptom Checklist-90-Revised; SOMS-7, Screening for Somatoform Symptoms-7; MD, major depression; SFD, somatoform disorder; HC, healthy control.

  1. Dantzer, R, Bluthé, RM, Castanon, N, Kelley, KW, Konsman, JP, and Laye, S (2007). Cytokines, sickness behavior, and depression. Psychoneuroimmunology, Robert, A, ed. Burlington: Academic Press, pp. 281-318
  2. Leonard, BE (2010). The concept of depression as a dysfunction of the immune system. Curr Immunol Rev. 6, 205-212.
    Pubmed KoreaMed CrossRef
  3. Dowlati, Y, Herrmann, N, Swardfager, W, Liu, H, Sham, L, and Reim, EK (2010). A meta-analysis of cytokines in major depression. Biol Psychiatry. 67, 446-457.
    Pubmed CrossRef
  4. Howren, MB, Lamkin, DM, and Suls, J (2009). Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosom Med. 71, 171-186.
    Pubmed CrossRef
  5. Capuron, L, Ravaud, A, Neveu, PJ, Miller, AH, Maes, M, and Dantzer, R (2002). Association between decreased serum tryptophan concentrations and depressive symptoms in cancer patients undergoing cytokine therapy. Mol Psychiatry. 7, 468-473.
    Pubmed CrossRef
  6. Wichers, MC, Koek, GH, Robaeys, G, Verkerk, R, Scharpé, S, and Maes, M (2005). IDO and interferon-alpha-induced depressive symptoms: a shift in hypothesis from tryptophan depletion to neurotoxicity. Mol Psychiatry. 10, 538-544.
    Pubmed CrossRef
  7. Maes, M, Leonard, BE, Myint, AM, Kubera, M, and Verkerk, R (2011). The new ‘5-HT’ hypothesis of depression: cell-mediated immune activation induces indoleamine 2,3-dioxygenase, which leads to lower plasma tryptophan and an increased synthesis of detrimental tryptophan catabolites (TRYCATs), both of which contribute to the onset of depression. Prog Neuropsychopharmacol Biol Psychiatry. 35, 702-721.
    Pubmed CrossRef
  8. Demirci, S, Aynalı, A, Demirci, K, Demirci, S, and Arıdoğan, BC (2017). The serum levels of resistin and its relationship with other proinflammatory cytokines in patients with alzheimer’s disease. Clin Psychopharmacol Neurosci. 15, 59-63.
    Pubmed KoreaMed CrossRef
  9. Anderson, G, Berk, M, and Maes, M (2014). Biological phenotypes underpin the physiosomatic symptoms of somatization, depression, and chronic fatigue syndrome. Acta Psychiatr Scand. 129, 83-97.
    Pubmed CrossRef
  10. Rief, W, Hennings, A, Riemer, S, and Euteneuer, F (2010). Psychobiological differences between depression and somatization. J Psychosom Res. 68, 495-502.
    Pubmed CrossRef
  11. Eyre, H, and Baune, BT (2012). Neuroimmunological effects of physical exercise in depression. Brain Behav Immun. 26, 251-266.
    Pubmed CrossRef
  12. Kang, HJ, Bae, KY, Kim, SW, Kim, JT, Park, MS, and Cho, KH (2016). Effects of interleukin-6, interleukin-18, and statin use, evaluated at acute stroke, on post-stroke depression during 1-year follow-up. Psychoneuroendocrinology. 72, 156-160.
    Pubmed CrossRef
  13. Tully, PJ, Baumeister, H, Bengel, J, Jenkins, A, Januszewski, A, and Martin, S (2015). The longitudinal association between inflammation and incident depressive symptoms in men: the effects of hs-CRP are independent of abdominal obesity and metabolic disturbances. Physiol Behav. 139, 328-335.
    Pubmed CrossRef
  14. Euteneuer, F, Schwarz, MJ, Hennings, A, Riemer, S, Stapf, T, and Selberdinger, V (2012). Psychobiological aspects of somatization syndromes: contributions of inflammatory cytokines and neopterin. Psychiatry Res. 195, 60-65.
    Pubmed CrossRef
  15. Euteneuer, F, Schwarz, MJ, Hennings, A, Riemer, S, Stapf, T, and Selberdinger, V (2011). Depression, cytokines and experimental pain: evidence for sex-related association patterns. J Affect Disord. 131, 143-9.
    Pubmed CrossRef
  16. Wittchen, HU, Zaudig, M, and Fydrich, T (1997). Göttingen: Hogrefe Verlag
  17. Hiller, W, Rief, W, and Brähler, E (2006). Somatization in the population: from mild bodily misperceptions to disabling symptoms. Soc Psychiatry Psychiatr Epidemiol. 41, 704-712.
    Pubmed CrossRef
  18. Rief, W, Henningsen, P, and Hiller, W (2006). Classification of somatoform disorders. Am J Psychiatry. 163, 746-747.
    Pubmed CrossRef
  19. Escobar, JI, Rubio-Stipec, M, Canino, G, and Karno, M (1989). Somatic symptom index (SSI): a new and abridged somatization construct. Prevalence and epidemiological correlates in two large community samples. J Nerv Ment Dis. 177, 140-146.
    Pubmed CrossRef
  20. Rief, W, and Hiller, W (1999). Toward empirically based criteria for the classification of somatoform disorders. J Psychosom Res. 46, 507-518.
    Pubmed CrossRef
  21. Schmitt, M, Altstötter-Gleich, C, Hinz, A, Maes, J, and Brähler, E (2006). Norms for the simplified Beck Depression Inventory (BDI-V) in a non-clinical population. Diagnostica. 52, 51-59.
  22. Derogatis, LR (1994). SCL-90-R: Administration, scoring and procedures manual. Minneapolis, MN: National Computer Systems
  23. Pedrosa Gil, F, Nickel, M, Ridout, N, Schwarz, MJ, Schoechlin, C, and Schmidmaier, R (2007). Alexithymia and interleukin variations in somatoform disorder. Neuroimmunomodulation. 14, 235-242.
    Pubmed CrossRef
  24. Frey, I, Berg, A, Grathwohl, D, and Keul, J (1999). Soz Praventivmed. 44, 55-64.
  25. Oades, RD, Myint, AM, Dauvermann, MR, Schimmelmann, BG, and Schwarz, MJ (2010). Attention-deficit hyperactivity disorder (ADHD) and glial integrity: an exploration of associations of cytokines and kynurenine metabolites with symptoms and attention. Behav Brain Funct. 6, 32.
    Pubmed KoreaMed CrossRef
  26. Perneger, TV (1998). What’s wrong with Bonferroni adjustments. BMJ. 316, 1236-1238.
    Pubmed KoreaMed CrossRef
  27. LaClair, R, O’Neal, K, Ofner, S, Sosa, MJ, Labarrere, CA, and Moe, SM (2008). Precision of biomarkers to define chronic inflammation in CKD. Am J Nephrol. 28, 808-812.
    Pubmed KoreaMed CrossRef
  28. Kaplan, RC, Ho, GY, Xue, X, Rajpathak, S, Cushman, M, and Rohan, TE (2007). Within-individual stability of obesity-related biomarkers among women. Cancer Epidemiol Biomarkers Prev. 16, 1291-1293.
    Pubmed CrossRef
  29. Zalli, A, Jovanova, O, Hoogendijk, WJ, Tiemeier, H, and Carvalho, LA (2016). Low-grade inflammation predicts persistence of depressive symptoms. Psychopharmacology (Berl). 233, 1669-1678.
    Pubmed KoreaMed CrossRef
  30. Maes, M, Berk, M, Goehler, L, Song, C, Anderson, G, and Gałecki, P (2012). Depression and sickness behavior are Janus-faced responses to shared inflammatory pathways. BMC Med. 10, 66.
    Pubmed KoreaMed CrossRef
  31. Dantzer, R, O’Connor, JC, Lawson, MA, and Kelley, KW (2011). Inflammation-associated depression: from serotonin to kynurenine. Psychoneuroendocrinology. 36, 426-436.
    Pubmed KoreaMed CrossRef
  32. Gabbay, V, Liebes, L, Katz, Y, Liu, S, Mendoza, S, and Babb, JS (2010). The kynurenine pathway in adolescent depression: preliminary findings from a proton MR spectroscopy study. Prog Neuropsychopharmacol Biol Psychiatry. 34, 37-44.
    Pubmed KoreaMed CrossRef
  33. Rothermundt, M, Arolt, V, Fenker, J, Gutbrodt, H, Peters, M, and Kirchner, H (2001). Different immune patterns in melancholic and non-melancholic major depression. Eur Arch Psychiatry Clin Neurosci. 251, 90-97.
    Pubmed CrossRef
  34. Meier, CA, Bobbioni, E, Gabay, C, Assimacopoulos-Jeannet, F, Golay, A, and Dayer, JM (2002). IL-1 receptor antagonist serum levels are increased in human obesity: a possible link to the resistance to leptin?. J Clin Endocrinol Metab. 87, 1184-1188.
    Pubmed CrossRef
  35. Marques-Vidal, P, Bochud, M, Bastardot, F, Lüscher, T, Ferrero, F, and Gaspoz, JM (2011). Levels and determinants of inflammatory biomarkers in a Swiss population-based sample (CoLaus study). PLoS One. 6, e21002.
    Pubmed KoreaMed CrossRef
  36. Kwidzinski, E, and Bechmann, I (2007). IDO expression in the brain: a double-edged sword. J Mol Med (Berl). 85, 1351-1359.
    Pubmed CrossRef
  37. Moylan, S, Maes, M, Wray, NR, and Berk, M (2013). The neuroprogressive nature of major depressive disorder: pathways to disease evolution and resistance, and therapeutic implications. Mol Psychiatry. 18, 595-606.
    Pubmed CrossRef

This Article

Cited By Articles
  • CrossRef (0)

Author ORCID Information

Funding Information

Social Network Service