Changes in pain during a depressive episode and relationship to cytokine levels in major depressive disorder

Abstract Background Depressed patients have an increased incidence of pain. A pathophysiological connection between depression and pain is still not revealed. Immunological activation has been found in both depression and pain. There are few studies of pain and immune activation in patients with depression, without inflammatory and autoimmune disorders. Methods This is a naturalistic follow-up study of 50 patients with a major depressive disorder (MDD) depressive episode, without any inflammatory or autoimmune conditions. We have previously reported on the relationship between depression and cytokine levels. In this study, we obtained data of depression, pain and cytokine levels before and after 12 weeks of depression treatment. All patients were medication-free at inclusion. Results At inclusion three out of four patients experienced pain, and the pain scores correlated with the depression scores. After treatment, as depression was relieved, the pain scores dropped significantly and were no longer correlated to the depression scores. There were no correlations between pain scores and cytokine levels. Pain level at inclusion did not correlate with depression treatment outcome. Conclusion Our findings indicate that pain is a feature of depression. Pain levels and cytokine values didn’t correlate. Pain at inclusion did not predict depression treatment outcome.


Introduction
Depression is a disorder affecting how you feel, think, and behave.It is the most common mental disorder and is distributed over several diagnoses [1].According to the Diagnostic and Statistical Manual of Mental Disorders version 5 (DSM-5), the criteria for a depressive episode are sadness and/or loss of interest, in addition to several mental, cognitive and behavioral signs and symptoms.Although pain is not a symptom included in the diagnostic criteria of depression, the relation between depression and pain has been known for decades [2][3][4].Pain is defined as 'An unpleasant sensory and emotional experience associated with actual or potential tissue damage' [5].Pain is included as an item in the depression symptom severity scale, Inventory of Depressive Symptomatology (IDS) and numerous concepts have been developed to describe and explain the co-occurrence of depression and pain [6].
A multi-center study of 3568 out-clinic patients with a Major Depressive Disorder (MDD) and depressive episode found that almost 60% suffered from pain [7].Data from the Netherlands Study of Depression and Anxiety (NESDA) showed that patients with pain were more prone to a chronic course of depression and anxiety disorders [8].It has further been found that pain has a negative impact on depression treatment response [3], and that the severity of pain is a strong predictor of poor depression treatment outcome [9].Research has also shown that there is a bidirectional influence of pain and depression.A change in severity of either symptom predicted subsequent severity of the other symptoms [10,11].
Pain is one of the five cardinal clinical signs of inflammation, and research on the pathophysiology of pain has provided insight into the involvement of cells and signaling molecules of the immune system [12].Macrophages are immune cells resident in different tissues, or they may mature from circulating monocytes in the blood and migrate into the tissues and elicit tailored inflammatory tasks.Activated macrophages are dominant producers of cytokines, the signaling molecules of the innate immune system [13].Experimental enhancement of macrophages in peritoneum of mice has been found to cause an increase in nociceptive pain response [14].Furthermore, it has been shown that rats with a partial ligation of the sciatic nerve have enhanced pain sensitivity to thermal stimuli, and that depletion of Pain in depression; cytokines; follow-up study; treatment outcome prediction macrophages reduce this effect [15].A study by Morris and Esiri (1998) discovered expression of cytokines and the presence of their cellular receptors in normal human brain tissue [16].The notion of cytokine activity in the brain is supported by the findings of resident immune cells in the central nervous tissue, the microglia [17], and that these cells express cytokines [18].In addition, there is evidence for the involvement of cytokines and glia in a 'glial-cytokine-neuronal interactions' , and that this is part of the underlying mechanism of persistent pain [19].
Immune system activation has also been found in depression.Meta-analyses have shown that depression is associated with elevated levels of IL-6, TNF [20] and the soluble cytokine receptor sIL-2R [21].Previously we have found that a broader range of cytokines were elevated in a depressive episode and, normalized after recovery [22].This suggests that cytokines play a central role in the pathophysiology of depression, and that different cytokines may contribute to different symptoms.Cytokines are thus molecules that are found to be associated with the biological mechanisms of both pain and depression.
Chronic pain and depression are well-documented co-morbid conditions which co-occur between 30 and 50% of the time [23][24][25].A recent study found that chronic pain patients have an increased risk for depression, with an Odds Ratio (OR) of 1.86 (1.37-2.54)[26].However, to the best of our knowledge, there seems to be little updated data on how often and to which degree depressed patients experience pain.Although there is growing evidence showing that both pain and depression is related to immune abnormalities [27,28], there are few follow-up studies that have investigated the course and associations of depression, pain and cytokines in patients treated for a MDD depressive episode.
Here we investigate the prevalence of pain in a MDD depressive episode, how 12 weeks of treatment for the depressive episode is associated with a change in the prevalence of pain, and whether depression scores and pain are correlated at baseline and/or at follow-up.We also investigate whether pain at baseline is associated with treatment outcome in a MDD depressive episode, and finally, if cytokine levels are correlated with pain scores, or change in pain scores in MDD.

Study design
This is a naturalistic observational follow-up study within a group of depressed MDD patients without any inflammatory or autoimmune diseases.All patients were un-medicated at baseline.Depression symptom severity, pain scores and 13 cytokines were assessed before (baseline) and at follow up after 12 weeks of treatment for depression.We measured depression scores using the Montgomery Aasberg Depression Rating Scale (MADRS) and the clinician rated 30-item version of Inventory of Depressive Symptomatology (IDS).Both were in the Norwegian version.We used IDS item no. 25 to assess pain.The psychometric properties and the validity of IDS is examined by Rush et al. [29,30].We have described item 25 of IDS and the answer options, English version, in Section 2.4 Pain assessment.We have earlier reported that the mean levels of various cytokines were elevated in the patient group at baseline and normalized after treatment [22].More information on the cytokine levels and their relation to depression and treatment in this group is elaborated in our former paper [22].In this paper, we have studied the prevalence of pain in Major Depressive Disorder (MDD), in a depressive episode and after depression treatment.We have also examined the correlation between pain scores and cytokine levels in MDD.

Study population
The study population (n = 50) was recruited from the outpatient clinic in Ringerike Psychiatric Center, Norway, between May 2009 and October 2011.All interviews, assessments and diagnosing were done by the same psychiatrist, Johan Dahl.To be included, the patients had to fulfill the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria for MDD and have a score of at least 22 on the IDS scale.They had to be aged 18-60 years, and only drug-free patient were included.The minimum duration of the drug washout period was 3 weeks.This is in line with previous studies of cytokines in depression [20,[31][32][33].The majority of the patients had a much longer drug-free period.Exclusion criteria were: presence of an autoimmune disorder, chronic inflammation, severe metabolic syndrome, body mass index (BMI; >34 kg/m 2 ), psychotic disorder, alcohol or drug dependence, receiving anti-inflammatory, antiviral, antibiotic or immune modulating drugs, and inability to provide informed consent to participate.Patients with chronic inflammation were excluded by collecting information from the general practitioners, from interview with the patients, and from the hospital records.In addition, the patients' C-reactive protein (CRP) values at baseline were used.A wide-screening blood analysis was used to reveal potential somatic comorbidity.This included liver enzymes, albumin, creatinine, glomerular filtration rate (calculated), alkaline phosphatase, blood lipids, hematologic status, erythrocyte sedimentation rate, CRP, ferritin, vitamin B12, methylmalonic acid, homocysteine, serum folate, thyroid status, cortisol, fasting blood glucoses, and glycated hemoglobin (HbA1c).This screening revealed one patient in need of vitamin B12 therapy, who received vitamin B12 injections and remained in the study.All other test results were within normal ranges We recruited 50 depressed MDD patients, 38 (76%) female and 12 (24%) male.The age at inclusion was 40.0 ± 12.0 years (mean ± SD) for the whole group (40.4 ± 12.8 years for females and 38.9 ± 9.7 years for males), and the BMI of this cohort was 25.8 ± 5.5 kg/m2.The data investigated in the present study are collected from the population in our earlier report of cytokine levels pre-and post-treatment in a MDD depressive episode compared to healthy controls [22].
At baseline all 50 patients fulfilled the DSM-IV criteria for a major depressive episode.The mean MADRS score was 27.8 (±5.8), and the IDS was 37.1 (±8.2).A total of 43 patients (86%) completed the 12 weeks of therapy.The MADRS was reduced to 13.4 (±9.2), and IDS was reduced 18.6 (±12.0).These are significant reductions with p < 0.001 for both.These data are presented in Table 1.

Treatment
The choice of treatment was decided by the therapist and patients on an individual basis, according to the Norwegian National Guidelines for treatment of depression [34].For mild depression, counseling and psychological intervention are the recommended interventions.Antidepressant medication should be considered when there is no response to non-pharmacological interventions.It should also be considered if the patient has had former moderate or severe depression.Also for moderate and severe depression, structured psychological treatment is recommended, but there is a stronger indication for the use of antidepressants.For severe depression, a combination of medication and structured psychological intervention is recommended.The guidelines underline the patient's own choice and collaboration.The use of medication in the 12 weeks treatment period was registered.After the medication-free period during baseline assessments the following medications were administered throughout the entire follow-up phase: escitalopram (n = 6), venlafaxine (n = 5), citalopram (n = 1), mirtazapine (n = 1), and sertraline (n = 1).Three patients received lamotrigine in combination with an antidepressant (venlafaxine, citalopram, or mirtazapine).One patient received benzodiazepines (oxazepam and zopiclone) in combination with mirtazapine.In addition, all patients received an average of one psychotherapy session per week, and for approximately two-thirds of the sample this was the only treatment.This is rendered in Table 2.

Pain assessment
Pain was assessed with item 25 of IDS.The IDS starts with the instruction: Please mark one response to each item that best describes the patient for the last seven days.
For item 25, assessment of pain, the answer options are: 0. States there is no somatic complaints or pains.1. Complains of headaches, abdominal, back, or joint pains that are intermittent and not disabling.2. Complains that the above pains are present most of the time.
3. Functional impairment results from above pains.

Cytokine measurements
A broad screening panel (human cytokine 27 plex, catalog no.171A11127, Bio-Rad) was used to screen a selection of plasma samples (11 patients and 7 normal controls).Based on the screening and previous findings, a custom made 13 plex targeted against interleukin (IL)-1β, IL-1Ra, IL-2, IL-5, IL-6, IL-7, IL-8, IL-10, IL-15, granulocyte colony-stimulating factor (G-CSF), macrophage inflammatory protein 1 alpha (MIP-1α), tumor necrosis factor (TNF), and interferon gamma (IFN-γ) (Bio-Rad, Hercules, California, USA) was further used to determine the plasma targeted cytokine concentrations.All samples were thawed on ice, vortexed, and then spun down at 14,000 ×g for 10 min at 4 °C prior to dilution (1:3), and fifty microliters was loaded onto plate.Individual sets of patient samples were loaded in duplicate on the same plate.The assays were optimized for low level detection with performing a tenfold dilution of the standards and extending incubations.All wash steps were performed using the Bio-Plex Pro™ Wash Station and the multiplex analysis were performed with a Luminex IS 100 (both from Bio-Rad, Hercules, CA, USA).An in-house spiked control was used to determine intra-assay and inter-assay percent of coefficient of variation (% CV).The cytokine values are rendered in Table 3.

Statistics
To compare the pre-and post-treatment levels of the depression scores, a nonparametric two-paired-samples test (Wilcoxon test with repeated measures) was used.The same test was used for comparison of the pain scores pre-and post-treatment.The nonparametric Spearman's correlations test was used for the correlation between pain and depression scores, and for the correlation between pain and the change in depression scores.This test was also used for the correlation between pain scores and cytokine levels.For the comparison of pain and IDS, the pain score was omitted from the IDS total score.

Pain severity scores at baseline and after 12 weeks of treatment
The median pain score at baseline (N = 50) was 1.54 (±1.15) and was significantly reduced to 0.93 (±0.91) at follow up (p < 0.001) These data are presented in Table 4.The pain scores at baseline were distributed as follows: 12 patients had a score of 0, 13 had a score of 1, 11 had a score of 2, and 14 had a score of 3.After 12 weeks of treatment (N = 43) 17 patient scored 0, 14 scored 1, 10 scored 2 and only two scored 3.These data are rendered in Figure 1.After 12 weeks of depression treatment (N = 43), 24 patients (55.8%) reported a lower pain score, 15 patients (34.9%) reported no change and 4 (9.3%) reported enhanced pain score.For the seven patients who dropped out, the pain scores at baseline were not significantly different from the other participants.

Correlation between pain and depression
At baseline the depression scores were correlated with pain scores.The Spearman correlation coefficient (r) for pain score and MADRS at baseline was 0.363 (p = 0.01), while for pain score and IDS it was 0.303 (p = 0.03).After 12 weeks of treatment, pain scores and depression scores were no longer significantly correlated, with r = 0.155 (p = 0.32) for MADRS and r = 0.235 (p = 0.13) for IDS.The change in pain scores from baseline until after 12 weeks (Δ pain) were not significantly correlated with the change in depression scores (Δ MADRS and Δ IDS).See Table 5.

Pain and depression treatment outcome
The reduction in depression scores after 12 weeks of depression treatment was not significantly correlated to the pain level at baseline.For baseline, pain scores and the Δ MADRS the r was ÷0.067, with a p-value of 0.67.For baseline pain scores and the Δ IDS the r was ÷0.117 with a p-value of 0.45.The data are rendered in Table 5.We also compared change in MADRS and IDS from baseline till follow-up between those with low pain at baseline (score 0 and 1) and those with high pain at baseline (score 2 and 3).There was no difference between the two groups; the p-values were respectively 0.990 and 0.894.

Correlation between cytokines and pain
There was no correlation between cytokine levels and pain scores, neither at baseline nor after 12 weeks.The change in pain scores (significant lower after 12 weeks of treatment), was not correlated to the cytokine levels, neither at baseline nor at follow-up.We omitted IL-2 and IL-15 in this test because these cytokines were detectable in less than 50% of the samples.

Treatment without medication
The choice of treatment was not an object of this study, and was decided by therapist and patients, as recommended in the Norwegian national treatment guidelines.Out of the 43 who completed the 12 weeks of treatment, 29 did not use any medication and 14 were treated with antidepressants.There were no significant differences between these groups for MADRS, IDS, and pain scores at baseline or follow up (data not shown).

Discussion
Our main finding is that as many as three out of four depressed MDD patients experienced pain, nearly 30% experienced functional impairment from the pain, and there was a significant drop in pain scores after 12 weeks of treatment as depression was relieved.We did not find that pain was associated with treatment effect, nor with cytokine levels.
Our finding of a high occurrence of pain in depression is in line with former findings [7,24,35,36].We also found that pain scores were correlated to depression severity before treatment, and that both depression and pain scores were significantly reduced after 12 weeks of depression treatment.However, the scores were no longer correlated after ended treatment.Altogether our findings indicate that pain is a feature of a MDD depressive episode.This may have an important clinical implication.In the RESPECT trial Kroenke et al. showed that 'recognizing and optimizing the management of comorbid pain that commonly coexists with depression may be important in enhancing depression response and remission rates' [36].
We found that neither pain score at baseline nor change in pain score after treatment was correlated to depression treatment outcome.Further, no difference in depression treatment outcome between those with high and low pain scores at baseline was found.This indicates that pain does not seem to predict depression treatment outcome.These findings contrast with other findings [3,9].
Cytokines have been found to be involved in a 'glial-cytokineneuronal interaction' , the underlying mechanism of persistent pain [19].We have earlier shown that the levels of various cytokines are elevated in a MDD depressive episode and normalized after recovery [22].In our present study, we found a significant reduction of pain as depression was reduced.Approximately two third did not use any medication during the 12 weeks of therapy.Several studies have shown that antidepressants have an anti-inflammatory effect [37,38].It has even been hypothesized that the anti-inflammatory effect of antidepressants is crucial to their anti-depressant effect [39,40].However, in 2016, we published a paper on cytokine levels related to non-pharmacological treatment of MDD.The levels of various cytokines were elevated before treatment and reduced to normal levels after recovery [41].This indicates that normalization of cytokine levels in depressed patients may be connected to recovery from depression per se.
In this present study, we found indications for pain being a feature of depression.On the other hand, we found no correlations between cytokine levels and pain scores, neither at baseline nor after treatment.Further, cytokine levels were not correlated to the change in pain scores.Thus, we were not able to provide evidence for a possible causal relation between depression, cytokines, and pain.This is in contrast with former findings [28].Our finding may be a false negative finding (type 2 error) caused by a restricted number of patients (n = 50) included in our study.This means that we cannot make any strong conclusion regarding the involvement of cytokines in the pathophysiology of pain in depression based on our findings.There is a need for more research into the relation between cytokines and pain in depression.In a recent paper,  the support for a Psycho-Neuro-Endocrine-Immune basis for a placebo mediated analgesia and anxiolytic response is reviewed [42].

Limitations
One possible limitation of the present study might be that we assessed cytokines in peripheral blood, and pain is constituted of mechanisms both in the periphery and in the central nervous system.However, there are findings indicating a connection between peripherally assessed cytokines and cytokine activity in the brain.Former studies have shown that the blood brain barrier (BBB) is impeded by immune activation [21], and there is evidence of a strong correlation between plasma and CSF levels of certain cytokines [43,44].
Cytokines produced and assessed elsewhere in the body are able to bind to receptors on brain immune cells, such as microglia, and further induce the expression of cytokines [45,46].In addition, an increase in serum cytokine levels have also been observed after non-neurological surgery that was accompanied by even higher cytokine levels in the cerebrospinal fluid (CSF) [47].
We have used item 25 of IDS to assess pain.This captures pain during the last week without defining whether it is daily pain, chronic pain, etc.We do not know of any previous studies using item 25 IDS as a main indicator of pain, or how it could relate to standard tools, as VAS or SF-36 Another possible limitation is the relatively small number of individuals (N = 50), making the study vulnerable to type II error.A further limitation may be that the patients were enrolled from a population referred with MDD to a specialized psychiatric outpatient clinic, making these results less relevant to other depressed patients.
One strength of this study is that the patients were thoroughly screened for inflammatory and immunological diseases.Further, the patients were all free of medication at inclusion.This rules out the potential influence from other conditions and medication on the experienced pain and cytokine levels at baseline.

Table 1 .
description of the patient population; sub classification and depression severity quantified by Madrs and ids.

Table 3 .
Plasma cytokine concentrations in patients at baseline and at week 12.

Table 4 .
symptom scores of depression and pain, pre-and post-treatment.

Table 5 .
correlations between pain scores and depression scores.