- Email: [email protected]
The efficacy of anti-inflammatory treatment interventions on depression in individuals with major depressive disorder and high levels of inflammation: A systematic review of randomized clinical trials
Physiology & Behavior 207 (2019) 104–112
Contents lists available at ScienceDirect
Physiology & Behavior journal homepage: www.elsevier.com/locate/physbeh
Review
The efficacy of anti-inflammatory treatment interventions on depression in individuals with major depressive disorder and high levels of inflammation: A systematic review of randomized clinical trials
T
David J. Allison , Bhanu Sharma, Brian W. Timmons ⁎
Child Health & Exercise Medicine Program, McMaster University, Hamilton, ON, Canada
ARTICLE INFO
ABSTRACT
Keywords: Major depressive disorder Depression Inflammation Cytokine Anti-inflammatory
Background: Studies evaluating the efficacy of anti-inflammatory treatment strategies for major depressive disorder (MDD) commonly include participants who do not demonstrate elevated concentrations of baseline inflammatory mediators, or simply fail to assess baseline inflammation. This may result in an underestimation of the efficacy of such treatment strategies. Sampling and methods: This systematic review included randomized controlled trials related to the use of antiinflammatory treatment strategies in individuals who demonstrate elevated concentrations of inflammatory mediators, identified by searching OVID MEDLINE, OVID EMBASE, and OVID PsychINFO. Results: Fifty-one randomized controlled trials were identified via a title and abstract screen, of which 43 were excluded following full-text screening. Of the 8 included trials (n = 437), 5 utilized anti-inflammatory pharmaceuticals, 2 utilized omega-3 fatty acids, and 1 utilized exercise and meditation. Among these studies, only the trial related to exercise and meditation demonstrated both an elevation in baseline inflammation, and a significant reduction in inflammation following intervention. Issues related to insufficient interventions and/or lack of reporting of inflammatory mediators at one or more time-points were common. Conclusion: Among the growing number of studies which examine the potential antidepressant benefit of antiinflammatory treatment strategies, few have studied populations which demonstrate elevated baseline concentrations. Further, studies commonly fail to induce significant changes in inflammation following intervention. Together, this may explain the low efficacy frequently reported with such interventions. Future studies which utilize both stronger interventions (sufficient to reduce levels of inflammation), and participants with elevated baseline concentrations, may produce a more substantial influence on symptoms of (MDD).
1. Introduction Numerous studies have demonstrated that major depressive disorder (MDD) is associated with an activation of the inflammatory response system and elevated concentrations of proinflammatory mediators such as interleukin (IL)-6, IL-1β, and tumour necrosis factor (TNF) α [1–4]. Elevations in proinflammatory mediators have been shown to be capable of influencing symptoms of MDD via several mechanisms related to monoamine neurotransmission, glutamate neurotransmission, and hippocampal functioning/neurogenesis [5,6]. The ability of inflammatory mediators to influence brain chemistry and functioning may contribute to the ineffectiveness of traditional antidepressant medications (which target monoamine pathways) in certain subsets of depressed patients [7]. Depressed patients with increased concentrations of inflammatory mediators have been shown to have a ⁎
higher likelihood of demonstrating treatment resistance [7]. Additionally, improvements in depression following treatment with traditional anti-depressant pharmaceuticals have been shown to be associated with a reduction in inflammation [7]. Growing evidence supporting an inflammatory etiology of MDD has led to numerous studies which have evaluated the efficacy of anti-inflammatory treatment strategies. Despite evidence pertaining to the various mechanisms by which proinflammatory mediators are likely to contribute to symptoms of MDD, previous studies often demonstrate mixed results related to efficacy. It is important to note however, that past studies have frequently failed to assess baseline levels of inflammation and/or performed anti-inflammatory interventions on depressed individuals who do not demonstrate signs of chronic inflammation. For example, a study by Müller et al. 2006 [8], demonstrated significantly greater improvements in depression
Corresponding author at: Department of Kinesiology, Faculty of Science, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4L8, Canada. E-mail address: [email protected] (D.J. Allison).
https://doi.org/10.1016/j.physbeh.2019.05.006 Received 10 September 2018; Received in revised form 25 April 2019; Accepted 8 May 2019 Available online 09 May 2019 0031-9384/ © 2019 Elsevier Inc. All rights reserved.
105
60/20/20/20
Jazayeri et al. [23]
54/27/27
Gharekhan et al. (2014)
128/64/64
40/20/20
Abbasi et al. [17]
Inamdar et al. [19]
Randomized (Total/ Treatment/ Control)
No. of patients
Patient population
Source
Study
42/14/14/14
101/51/50
45/25/20
37/19/18
Analyzed (Total/ Treatment/ Control)
29%/29%/33%
Not stated
56.8 ± 13.09/ 57.2 ± 15.19
52%/60%
65%/70% 35.1 ± 8.0/ 34.2 ± 6.9
% Male & Age (yrs) (Treatment/ Control)
Characteristics
Fluoxetine (20 mg/daily/
None/None
Placebo (7.5 mg/daily)
Placebo (paraffin oil (1800 mg daily/ 4 months) ACE inhibitors; ARB; Statins Hemodialysis patients Losmapimod (7.5 mg daily)
Sertraline + Celecoxib (400 mg daily/ 6 weeks) Sertraline + Placebo (400 mg daily/6 weeks) Sertraline (constant dose) No Comorbidities Omega-3 (1800 mg daily/ 4 months)
Treatment/ Comedication/ Comorbidity
Fluoxetine: IL-6: 2.12 ± 1.44
IL-10:1.71 ± 1.07 pg/ml TNF-RII:1682 ± 1090.02 pg/ml
Fluoxetine: IL-6: 1.83 ± 1.51
IL-6: 6.07 ± 7.65 pg/ml TNF-alpha: 18.01 ± 73.88 pg/ml IL-10:1.78 ± 1.33 pg/ml TNFRII:1535.06 ± 1025.18 pg/ml
IL-6: 7.10 ± 11.59 pg/ml TNF-alpha: 11.67 ± 36.19 pg/ml
Placebo IL-6 2.56 ± 0.64 pg/ml Omega-3 CRP: 8.0 ± 6.78 mg/l IL-6: 136.85 ± 183.56 ng/l IL-10: 132.06 ± 94.81 pg/ml TNFα: 116.57 ± 172.32 ng/l Placebo CRP: 12.69 ± 14.02 mg/l IL-6: 106.48 ± 136.51 ng/l IL-10: 107.24 ± 63.01 pg/ml TNFα: 77.43 ± 59.82 ng/l
Placebo IL-6 2.78 ± 0.72 pg/ml Omega-3 CRP: 9.24 ± 8.79 mg/l IL-6: 144.38 ± 167.66 ng/l IL-10: 123.94 ± 71.91 pg/ml TNFα: 122.09 ± 171.69 ng/l Placebo CRP: 7.73 ± 4.89 mg/l IL-6: 103.54 ± 145.18 ng/l IL-10: 111.76 ± 67.39 pg/ml TNFα: 80.93 ± 64.89 ng/l
Losmapimod IL-6: 5.21 ± 9.71 pg/ml TNF-alpha: 5.46 ± 2.02 pg/ml IL-10: 2.03 ± 1.36 pg/ml TNF-RII: 1496.63 ± 987.15 pg/ml Placebo
2.16 ± 0.60 pg/ml
2.79 ± 0.76 pg/ml
Losmapimod IL-6: 5.77 ± 9.01 pg/ml TNF-alpha: 5.98 ± 2.58 pg/ml IL-10: 4.10 ± 10.94 pg/ml TNF-RII: 1841.07 ± 1292.82 pg/ml Placebo
Celecoxib IL-6
Post-treatment
Celecoxib IL-6
Baseline
Inflammatory mediators
Table 1 Study characteristics of identified clinical trials investigating anti-inflammatory interventions for depression in individuals with elevated inflammation.
Placebo 20.33 ± 7.56
Placebo 21 ± 4.72
HamD
HAMD-17 = 24.5 QIDS-SR16 = 18.0”
BDI = 12.6 IDS-C = 45.0
Placebo
(continued on next page)
BDI = 5.94 HAMD-17 = 11.94 IDS-C = 22.93 IDS-SR = 25.35 QIDS-SR16 = 8.74” HamD
IDS-SR = 29.34 QIDS-SR16 = 10.31 Placebo
Change Scores Losmapimod BDI = 7.05 HAMD-17 = 14.90 IDS-C = 28.44
13.44 ± 5.66
23.52 ± 7.49
Losmapimod BDI = 12.9 IDS-C = 45.3 HAMD-17 = 24.49 QIDS-SR16 = 17.8
Placebo 11.25 ± 1.25 BDI Omega-3
8.8 ± 2.1
HamD Celecoxib
Post-treatment
Placebo 21.3 ± 1.9 BDI Omega-3
22.2 ± 1.7
HamD Celecoxib
Baseline
Depression
D.J. Allison, et al.
Physiology & Behavior 207 (2019) 104–112
57/14/13/15/ 15
Mehta et al. [20]
106
40/20/20
45/15/15/15
60/30/30
Musil et al. [18]
Prakhinkit et al. [24]
Raison et al. [21]
*treatment & controls groups divided into responders and non- responders
Randomized (Total/ Treatment/ Control)
No. of patients
Patient population
Source
Study
Table 1 (continued)
60/30/30
40/13/14/13
32/18/14/20
53/12/12/15/14
Analyzed (Total/ Treatment/ Control) 8 weeks) + EPA placebo EPA (1000 mg/daily/ 8 weeks) + fluoxetine placebo EPA (1000 mg/daily/ 8 weeks) + Fluoxetine (20 mg/ daily/8 weeks) None/None Infliximab (5 mg/kg at baseline, 2 weeks, 6 weeks) Placebo None/Constant antidepressant regimen (types not specified)
37 ± 8.49/ 34 ± 8.46/ 33.86 ± 10.85
33%/33%
74.8 ± 1.7/ 74.0 ± 1.9/ 81.0 ± 1.7
0%
Placebo
44.8 ± 12.2/ 44.3 ± 11.1/ 40.0 ± 10.4
Traditional Walking (20 min: 3×/week) Sedentary None/None Infliximab (5 mg/kg at
None/ Benzodiazepine & Reboxetine at variable doses Meditative Walking (20 min: 3×/week)
Celecoxib (400 mg/daily/ 6 weeks)
7.6/ 8.0/ 10.0/ 8.6
61%/36%/75%
33.3%/ 33.3% 38.9 ± 46.0 ± 42.1 ± 46.6 ±
35.7%/38.5%/
Treatment/ Comedication/ Comorbidity
% Male & Age (yrs) (Treatment/ Control)
Characteristics
Not reported (no time x group interaction)
IL-6: 1.2 ± 0.4 (pg/ml)
IL-6: 0.8 ± 0.2 (pg/ml) Infliximab: CRP: 6.3 ± 8.9 mg/l
Traditional Walking CRP: 6.2 ± 2.2 (mg/l) IL-6: 1.1 ± 0.5 (pg/ml) Walking + Meditation CRP: 4.7 ± 1.5 (mg/l) IL-6: 0.7 ± 0.2 (pg/ml) Sedentary CRP: 4.2 ± 1.9 (mg/l)
Celecoxib: MIF: 49.2 ± 30.0 TGF-β: 9968 ± 2311 sCD14: 1441 ± 187 Placebo: MIF: 23.11 ± 8.3 TGF- β: 9968 ± 2311 sCD14: 1104 ± 772
CRP not assessed post Post Cytokine Transcripts not provided
IL-1B: 4.73 ± 3.37 EPA: IL-6: 2.38 ± 2.21 IL-1B: 5.31 ± 5.22 Fluoxetine+EPA: IL-6: 4.56 ± 7.28 IL-1B: 5.03 ± 4.01
Post-treatment
Traditional Walking CRP: 8.1 ± 2.4 (mg/l) IL-6: 0.7 ± 0.1 (pg/ml) Walking + Meditation CRP: 6.3 ± 1.8 (mg/l) IL-6: 0.9 ± 0.2 (pg/ml) Sedentary CRP: 3.8 ± 1.9 (mg/l)
CRP: 7.4 mg/l Placebo (R): CRP: 7.8 mg/l Placebo (NR): CRP: 3.1 mg/l Celecoxib: MIF: 32.7 ± 16.2 TGF-β: 9484 ± 3081 sCD14: 1474 ± 285 Placebo: MIF: 30.8 ± 11.7 TGF- β:8554 ± 2266 sCD14: 1279 ± 608
Infliximab (R):
Infliximab (NR): CRP: 6.0 mg/l
IL-1B: 4.21 ± 2.97 EPA: IL-6: 2.11 ± 1.44 IL-1B: 6.12 ± 3.56 Fluoxetine+EPA: IL-6: 4.21 ± 8.37 IL-1B: 2.55 ± 2.84
Baseline
Inflammatory mediators
(continued on next page)
Sed. Control: 18.6 HamD
Walking+Medit. 8.6*
Walking+Medit. 16.8
Sed. Control: 17.9 HamD Inflaximab: 24.1
GDS Trad. Walking: 15.5
Placebo: 16.9 ± 8.6
Infliximab (R): 12.7 ± 6.8 Placebo (R): 20.2 ± 6.7 Placebo (NR): 8.1 ± 3.8” HamD Celexcoxib: 11.8 ± 7.7
HamD Infliximab (NR): 21.9 ± 7.6
Not reported
Post-treatment
GDS Trad. Walking: 17.3
Placebo: 23.2 ± 5.5
Placebo (NR): 23.3 ± 3.6” HamD Celecoxib: 25.5 ± 4.1
Infliximab (R): 23.6 ± 11.2 Placebo (R): 23.9 ± 4.1
HamD Infliximab (NR): 24.1 ± 3.6
Fluoxetine + EPA: 30.86 ± 5.41
Fluoxetine: 29 ± 6.93 EPA: 29.29 ± 4.84
Baseline
Depression
D.J. Allison, et al.
Physiology & Behavior 207 (2019) 104–112
Physiology & Behavior 207 (2019) 104–112
None/37 participants on stable antidepressant dose
Placebo CRP: 5.4 ± 8.2 mg/l TNF-α: not reported
Abbreviations: HAMD: Hamilton Depression Rating Scale; BDI: Beck Depression Inventory; GDS: Geriatric Depression Scale; MADRS: Montgomery-Asberg Depression Rating Scale; IDSeC: Inventory of Depressive Symptomatology: Clinician; QIDS-SR16: Quick Inventory of Depressive Symptomatology; ACE: angiotensin-converting-enzyme; ARB: angiotensin 2 receptor blocker; EPA: eicosapentaenoic acid; CRP: c-reactive protein; IL: interleukin; TNF: tumour necrosis factor; MIF: macrophage migration inhibitory factor; TGF: transforming growth factor; sCD14: soluble cluster of differentiation 14.
Placebo: 23.6
Post scores not reported (no time x group interaction) baseline, 2 weeks, 6 weeks) Placebo 42.5 ± 8.2/ 44.3 ± 9.4
TNF-α: not reported
Treatment/ Comedication/ Comorbidity % Male & Age (yrs) (Treatment/ Control) Randomized (Total/ Treatment/ Control) Source
No. of patients
Patient population Study
Table 1 (continued)
Analyzed (Total/ Treatment/ Control)
Characteristics
Baseline
Inflammatory mediators
Post-treatment
Baseline
Depression
Post-treatment
D.J. Allison, et al.
following a 6-week treatment with reboxetine combined with the antiinflammatory drug celecoxib (in comparison to reboxetine alone) [8]. The effectiveness of this intervention may be driven by a successful reduction in elevated baseline levels of inflammation; however, inflammation was not assessed. The frequent omission of baseline inflammation in the literature is apparent in the systematic review by Köhler et al. 2014 [9], which examined the effect of anti-inflammatory treatments on MDD whereby, of the 14 trials included, only 2 reported baseline scores of inflammation [9]. This may be common practice due to the assumption that depression is typically regarded as a chronic inflammatory condition. However, it is critical to acknowledge that not all cases of depression are associated with heightened levels of inflammation [2,10,11]. Failure to assess baseline inflammation or the use of study populations which do not demonstrate elevated concentrations of inflammation may contribute to the limited efficacy of many previous studies which utilize anti-inflammatory treatment strategies. The use of such populations may be confounding for several reasons. First, participants with healthy concentrations of inflammatory mediators are simply not likely to be negatively influenced by inflammation and therefore would not benefit from an anti-inflammatory intervention. Second, there is evidence to suggest that when at healthy concentrations, proinflammatory cytokines may play critical roles related to neural plasticity, neurogenesis, and synaptic strengthening which would suggest a beneficial role [12,13]. Third, a study by WarnerSchmidt et al. 2011 provided evidence that anti-inflammatory drugs may actually attenuate the benefits of serotonergic antidepressants [14]. It may therefore be possible that the implementation of unnecessary anti-inflammatory strategies in depressed individuals, who do not demonstrate elevated concentrations of inflammatory mediators is ineffective and/or detrimental to symptom relief. The fact that not all cases of MDD are associated with high inflammation make it inappropriate to expect anti-inflammatory interventions to be an effective treatment for the entirety of the depressed population. Instead, it would seem more likely that there may be a subset of the population with pathological concentrations of inflammatory mediators who demonstrate a form of inflammation-induced depression. Based on this assumption, it may be possible that past studies which have failed to assess baseline inflammation, and previous systematic reviews which utilized studies which did not assess baseline inflammation, may have understated the efficacy of anti-inflammatory treatment strategies for MDD. Therefore, the purpose of this systematic review was to evaluate the efficacy of anti-inflammatory strategies for the treatment of MDD in individuals with elevated baseline concentrations of inflammatory mediators. The identification and utilization of a better characterized sample may help to elucidate the true potential of such strategies, and help better understand the current state of the science. 2. Materials and methods The current systematic review included randomized clinical trials which investigated anti-inflammatory interventions for the treatment of depression in individuals with clinical depression and elevated concentrations of inflammatory mediators at baseline. This review was registered with the international prospective register of systematic reviews PROSPERO network (registration number CRD42017059049). 2.1. Eligibility criteria Only randomized, controlled clinical trials were included in the current review. Participants had a clinical diagnosis of depression or self-report questionnaire scores suggesting depression. As this review was interested in evaluating the efficacy of anti-inflammatory treatments for depression, only studies which included baseline assessments of inflammation, allowing for the assessment of elevated concentrations 107
Physiology & Behavior 207 (2019) 104–112
Screening
Idenficaon
D.J. Allison, et al.
Records iden!fied through database searching: N = 1618
Records screened a#er duplicates removed:
Records Excluded: N = 1399
N = 1451
Full text ar!cles assessed for eligibility
Full text ar!cles excluded Depression not assessed / scores suggest lack of depression (n=6) Baseline inflamma!on not assessed (n=9) Abstract/conference publica!on (n=17) Inappropriate study design (n=11)
Eligibility
N = 51
Studies included in qualita!ve analysis
Included
N=8
Studies included in quan!ta!ve analysis N=0
Fig. 1. PRISMA (preferred reporting items for systematic reviews and meta-analyses) flow diagram of study selection for systematic review.
Table 2 Risk of bias, as assessed on the Cochrane risk of bias tool. Source
Abbasi et al. [17] Gharekhani et al. [22] Inamdar et al. [19] Jazayeri et al. [23] Mehta et al. [20] Musil et al. [18] Prakhinkit et al. [24] Raison et al. [21]
Allocation Sequence generation
Concealment
Low Low Low Low Low Low Low Low
Low High Low Low Low High Low High
Blinding of participants and personnel
Blinded outcome assessment
Incomplete outcome data
Selective reporting
Other bias
Low Low Low Low Low Low High Low
Low High Low Low Low Low High Low
Low High High High High High High Low
Low Low Low Low Low Low Low Low
Low Low High Low Low High Low Low
of inflammatory mediators, were included. Studies utilizing pharmaceutical drugs in addition to anti-inflammatory interventions would only be included if the pharmaceuticals were administered at a constant dose throughout the intervention. We identified anti-inflammatory treatment interventions as NSAIDs, anti-cytokine treatments, or lifestyle interventions (diet and exercise). No limitations or restrictions were placed on language or date of publication.
sciences librarian for the following databases: OVID MEDLINE, OVID EMBASE, and OVID PsychINFO. Independent searches were run for each database (see Table 1 for an example of a search strategy). All trials published prior to August 22, 2018, were searched for using terms related to depression, various immune mediators, and anti-inflammatory strategies. Two investigators (D.A. and B.S.) independently reviewed titles and abstracts to remove any obvious irrelevant studies using Covidence. All titles/abstracts deemed appropriate by both reviewers were automatically advanced to the full-text screening stage by Covidence. Any conflicts between reviewers regarding inclusion/exclusion of an article were discussed until a consensus was reached. The
2.2. Search methods for identification of trials The search strategy was optimized with the assistance of a health 108
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
[16]. CRP scores greater than 3.0 mg/L suggest unhealthy levels of CRP and a high risk for future heart attack and stroke, and was used as a cutoff for high inflammation in the current study.
Table 3 Example of search strategy – optimized for OVID EMBASE. 1. 2. 3. 4. 5. 6. 7.
exp. Depression/ Depress*.mp. 1 or 2 exp. Cytokines/ cytokines.mp. 4 or 5 (Chemokine or Interferon or Interleukin or Lymphokine or Monokine or Leukocytes or Tumour Necrosis Factor).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 8. 6 or 7 9. exp. Anti-Inflammatory Agents/ 10. anti-inflammatory agents.mp. 11. 9 or 10 12. 3 and 8 and 11 13. (selective serotonin reuptake inhibitor or SSRI).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 14. (TCA or Trycyclic Antidepressant).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 15. 13 and 14 16. 12 not 15 17. exp. controlled clinical trial/ 18. exp. randomized controlled trial/ 19. exp. crossover procedure/ 20. exp. double-blind procedure/ 21. exp. single-blind procedure/ 22. random*.mp. 23. factorial*.mp. 24. (crossover* or cross over* or cross-over*).mp. 25. placebo*.mp. 26. (double* adj blind*).mp. 27. (singl* adj blind*).mp. 28. assign*.mp. 29. allocat*.mp. 30. 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 31. exp. human/ 32. 16 and 30 and 31
2.5. Assessment of bias The bias risks of the randomized clinical trials included were assessed (D.A. and B.S.) based on the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions and methodological studies. We extracted data regarding quality for 7 domains. Sequence generation was considered adequate if authors described a random component. Allocation concealment was adequate if it was justified that neither participants nor investigators were aware of participant assignment. Blinding of participants and personnel was adequate if knowledge of the allocated intervention was prevented. Blinding of outcome assessors was adequate if the trial was characterized as double-blind, based on our own assessment of the trial. Incomplete outcome data were considered to be addressed appropriately if attrition and exclusions were reported with rationale and appropriately addressed. Selective reporting was considered adequate if all pre-specified primary outcomes were reported. Overall, trials were considered to have a low risk of bias if all the above domains were followed. Any trial with uncertain risk of bias or a high risk of bias in one or more domain were considered high risk. 3. Results 3.1. Trial flow
same two investigators also independently reviewed the remaining fulltext reports against the inclusion criteria via a similar process.
Results of the searches are presented in Fig. 1. Following the removal of duplicates, the title/abstract of 1452 unique papers were screened for eligibility. A total of 1401 papers were excluded, while 51 papers were included in the full-text screen for eligibility. Following the full-text screen, an additional 43 studies were excluded. Reasons for exclusion included the failure to assess baseline depression or the inclusion of participants with baseline scores suggesting a lack of clinical depression (n = 6), failure to assess baseline inflammation (n = 9), non-peer reviewed resources (n = 17), and inappropriate study design (n = 11).
2.3. Data extraction
3.2. Search results and study characteristics
Data were extracted independently by each investigator (D.A. and B.S.) using the same pre-structured form. The reviewers were not blinded to the study results, authors, or institutions. In addition to bibliographic information, data extraction included quality assessment, description of the participants, description of the intervention and control groups, and outcome measures related to depression and concentrations of inflammatory mediators. All data was cross-verified for accuracy by both reviewers.
Details from the 8 randomized controlled trials which met inclusion criteria are summarized in Table 2. The 8 included trials comprised a total of 437 participants. Of the 8 included trials, 5 (involving 282 participants) evaluated the efficacy of anti-inflammatory pharmaceuticals, including the NSAID, Celecoxib [17,18], the P38 mitogen-activated protein kinase inhibitor, Losmapimod [19] and the anti-tumour necrosis factor-α biologic, Infliximab [20,21]. Two studies (87 participants) evaluated the efficacy of the anti-inflammatory supplement omega-3 fatty acids [22,23]. One study (40 participants) examined the efficacy of meditative walking [24]. Due to the limited number of articles which showed both an elevated baseline concentration of inflammatory mediators and a successful reduction in inflammatory mediators following intervention, a quantitative analysis was not conducted. Instead, the studies were analyzed narratively.
2.4. Outcome measures Primary outcome measures included (1) the change in depressive symptoms measured on a continuous scale from baseline to the end of the intervention, (2) the change in inflammatory mediators, measured on a continuous scale, from baseline to the end of the intervention, and (3) the relationship between changes in scores of depression and changes in inflammatory mediators from baseline to the end of the intervention. Healthy cytokine concentrations for IL-1β, IL-2, IL-6, TNF-α, IFN-y, IL-10, and IL-4 were based on serum cytokine profiles obtained from healthy young and elderly individuals [15]. Healthy concentrations of C-reactive protein (CRP) were based on guidelines for low, moderate, and high risk for a cardiovascular event, endorsed by the centers for disease control and prevention and the American heart association
3.3. Pharmaceutical interventions Among the 5 studies which utilized anti-inflammatory pharmaceuticals, 2 failed to significantly reduce concentrations of inflammatory mediators in the treatment group [18,19], and 2 failed to report values of post inflammatory mediators [20,21]. The study by Abbasi et al. 2012 [17] was the only trial to achieve significant reductions in inflammation as shown by the reduction in the proinflammatory cytokine 109
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
IL-6 following the intervention with Celecoxib (200 mg, 2× daily). It should be noted, however, that IL-6 was the only inflammatory mediator assessed in the trial and baseline concentrations were within a healthy range when compared to the normative values from the study by Kim et al. [15]. Additionally, although a significant group x time interaction was observed, whereby the treatment group experienced a significantly greater reduction in IL-6, the placebo group also experienced a significant reduction in IL-6. In relation to changes in scores of depression, only the study by Abassi et al. 2012 showed a significant group x time interaction whereby the treatment group had a significantly greater improvement in Ham-D scores of depression. Post-hoc analysis revealed a significant reduction in the placebo group. Significant correlations between the change in IL-6 and the change in Ham-D scores of depression were also shown in both groups.
mediators make it difficult to accurately assess the efficacy of anti-inflammatory treatments for depression. Despite challenges with numerous confounding variables among human studies, prior research in animal models has produced promising results. A study by Krügel et al. 2013 [26] demonstrated the efficacy of the TNF-α inhibitor etanercept at reducing depressive symptoms in male Wistar rats [26]. Further, in a recent study, Lee et al. [27], assessed the efficacy of p-Coumaric Acid administration (a dietary phenolic acid with anti-inflammatory properties) following LPS-induced depression in rats [27]. It was shown to protect against depressive symptoms as shown by significant differences in behavioural, electrophysiological, and brain biochemical measures between treatment and control groups. The effectiveness of such anti-inflammatory interventions in animal models may relate to the ability to induce, and thereby ensure, baseline elevations in inflammation and depressive symptoms in animal models. The need to utilize human populations who demonstrate elevated concentrations of inflammatory mediators is further emphasized in the study by Raison et al. [5]. This study showed no significant difference between groups in the change of Ham-D scores of depression, or CRP concentrations. However, an exploratory analysis which focused on a subgroup of participants with CRP concentrations above 5 mg/L showed a 3.1-point greater reduction in Ham-D scores in the treatment group compared to placebo. This sub-analysis, involving a relatively small sample size (n = 22), did not reach statistical significance, however, a 3.1-point reduction would be considered clinically meaningful [25]. This may suggest, anti-inflammatory treatment strategies would produce treatment effects of greater magnitude if implemented in populations with elevated concentrations of inflammatory mediators. Findings like this, in conjunction with evidence suggesting that healthy concentrations of proinflammatory cytokines have various beneficial roles related to neural plasticity, neurogenesis, and synaptic strengthening [12,13], further emphasize the need to utilize participants which demonstrate chronic inflammation when assessing the efficacy of antiinflammatory strategies. The concurrent use of antidepressant pharmaceuticals may also influence the findings of studies which aim to evaluate the efficacy of anti-inflammatory treatments. In the current review, 5 of the 8 selected studies included participants who used some form of antidepressant pharmaceutical related to monoamine neurotransmission. Studies have shown that traditional antidepressant pharmaceuticals such as selective serotonin reuptake inhibitors (SSRIs), or selective norepinephrine reuptake inhibitors (SNRIs) may induce a considerable anti-inflammatory effect [28,29]. This may have contributed to the significant reduction in inflammation shown in the placebo group in the study by Abassi et al. 2012. In this study, the use of the non-steroidal anti-inflammatory, selective COX-2 inhibitor, Celecoxib, resulted in a significant reduction in IL-6 concentrations as well as a related significant reduction in HamD scores of depression. Interestingly, however, the placebo group also achieved a significant reduction in IL-6 concentrations and a related significant reduction in Ham-D scores of depression. Such results may relate to the continued use of the SSRI sertraline by the placebo group. Although this study still achieved a significant treatment effect, the magnitude of the effect may have been dampened due to the use of a control group which experienced a significant reduction in inflammation, potentially due to Sertraline usage. Among the 8 studies included in this review, only the study by Prakhinkit et al. [24] used both chronically inflamed, and non-medicated participants. In this study, which utilized meditative walking therapy as an anti-inflammatory treatment, a significant group by time interaction was shown for both CRP and scores of depression, as shown by the geriatric depression scale. Post-hoc analysis revealed significant reductions in CRP and scores of depression in the treatment group, and no change in either variable in the control group. Although correlations were not assessed, results from this study may highlight the importance of utilizing a patient population with elevated concentrations of
3.4. Supplement interventions Among the 2 studies which utilized omega-3 supplements [22,23], both demonstrated at least one proinflammatory mediator to be elevated above healthy concentrations at baseline. Neither study, however, demonstrated significant group x time effects for any inflammatory mediator. Improvements in scores of depression in the treatment group were observed only in the study by Gharekhani et al. 2014 [22]; however, no relation between changes in inflammatory mediators and changes in depression were apparent. 3.5. Meditative walking intervention The study which utilized meditative walking [24] demonstrated a significant elevation in baseline concentration of CRP in all groups. A significant group x time interaction was also demonstrated following the intervention consisting of meditative walking for 20 min, 3 times per week. Post hoc analysis showed a significant reduction in CRP concentrations in the treatment group and no change in the sedentary control group. CRP concentrations did however, remain well above normative healthy values post intervention [25]. A significant group x time interaction was also shown for scores of depression as assessed using the Geriatric Depression Scale. Post hoc analysis showed a significant reduction in scores of depression in the treatment group and no change in the control group. Analysis of the potential relationship between the change in depression and change in CRP concentrations was, however, not reported. 3.6. Bias of included trials Assessment of bias for the included trials is shown in Table 2. Six of the eight trials were associated with a high risk of bias. Each of these 7 trials were judged to be at high risk for bias related to incomplete outcome data. Certain trials were also deemed to be high risk for factors related to allocation concealment, blinded outcome assessment, and other biases related to for-profit bias (Table 3). 4. Discussion Following full-text review, a total of 22 original studies with an appropriate study design were identified. Among these studies, failure to assess baseline inflammation and the use of participants who did not demonstrate scores suggesting clinical depression were common, resulting in the exclusion of 14 randomized controlled trials. Among the trials included in the current review, failure to significantly reduce concentrations of inflammatory mediators to an extent greater than that of the control group was apparent in 6 of the 8 studies [18–23]. The common use of participants who do not show elevated concentrations of inflammatory mediators as well as the use of insufficient interventions which failed to effectively reduce concentrations of inflammatory 110
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
inflammatory mediators and who are free from potentially confounding anti-depressant pharmaceuticals when assessing the efficacy of antiinflammatory interventions for the treatment of depression.
[3]
4.1. Limitations and future directions Although the assessment of baseline inflammation is important when evaluating the efficacy of anti-inflammatory treatments for depression, high heterogeneity between studies, and a lack of well-established normative data make it difficult to set definitive cutoffs for healthy vs pathological levels of inflammatory biomarkers. CRP is perhaps the most commonly reported indicator of inflammation, and has the most well-established normative data. CRP may however, not be the most influential inflammatory mediator in terms of its influence on symptoms of depression. Proinflammatory cytokines such as IFNy, IL-6, IL-1B, and TNF-α have been frequently shown to be associated with depression and influence various related mechanisms [1,7,30]. However, these cytokines also have less established normative values with relatively few studies assessing healthy concentrations. The high inter-study variability of concentrations of these mediators further necessitates the need for further studies. This makes establishing a set cut point for what constitutes high inflammation difficult and is a limitation to the current review, and other studies which define pathological levels of inflammation based on such inflammatory markers. More studies will be required in the future to help strengthen normative data. Future studies will also be required to further elucidate which inflammatory biomarkers are the most impactful in terms of depression, as well as what concentrations may be considered pathological. It may also be important for future studies to attempt to use medication free participants to avoid unintended anti-inflammatory effects. Although recruitment of an unmedicated depressed population may prove difficult, it may allow for a more accurate assessment of the efficacy of antiinflammatory interventions.
[4] [5] [6] [7] [8]
[9]
[10] [11]
[12]
[13]
5. Conclusion
[14]
Numerous studies have been conducted which relate to the efficacy of anti-inflammatory interventions for the treatment of MDD. This review highlights that many studies either fail to assess baseline concentrations of inflammation, or utilize participants who do not demonstrate elevated baseline concentrations of inflammatory mediators. Further, of the 8 studies included in the current review, whereby baseline inflammation was assessed, only 2 achieved reductions in inflammatory mediators to extent greater than that of the control group following intervention. There is a need for future studies which evaluate the efficacy of effective anti-inflammatory strategies for the treatment of MDD, which utilize participants who demonstrate chronic inflammation, and are free of comedications.
[15]
[16]
[17]
Acknowledgements
[18]
DJA was supported by a Michael G. DeGroote Fellowship Award. BS is supported by the Queen Elizabeth Science and Technology Award. BWT is supported by the Canada Research Chair in Child Health & Exercise Medicine.
[19]
Declaration of interest None.
[20]
References [21]
[1] A.H. Miller, C.L. Raison, The role of inflammation in depression: from evolutionary imperative to modern treatment target, Nat. Rev. Immunol. 16 (2015) 22–34, https://doi.org/10.1038/nri.2015.5. [2] Y. Dowlati, N. Herrmann, W. Swardfager, H. Liu, L. Sham, E.K. Reim, K.L. Lanctôt,
[22]
111
A meta-analysis of cytokines in major depression, Biol. Psychiatry 67 (2010) 446–457, https://doi.org/10.1016/j.biopsych.2009.09.033. S. Mostafavi, A. Battle, X. Zhu, J.B. Potash, M.M. Weissman, J. Shi, K. Beckman, C. Haudenschild, C. McCormick, R. Mei, M.J. Gameroff, H. Gindes, P. Adams, F.S. Goes, F.M. Mondimore, D.F. MacKinnon, L. Notes, B. Schweizer, D. Furman, S.B. Montgomery, A.E. Urban, D. Koller, D.F. Levinson, Type I interferon signaling genes in recurrent major depression: increased expression detected by whole-blood RNA sequencing, Mol. Psychiatry 19 (2014) 1267–1274, https://doi.org/10.1177/ 1948550614552729. M.B. Howren, D.M. Lamkin, J. Suls, Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis, Psychosom. Med. 71 (2009) 171–186, https://doi.org/10.1097/PSY.0b013e3181907c1b. C.L. Raison, J.C. Felger, A.H. Miller, Inflammation and treatment resistance in major depression: a perfect storm, Neuropsychopharmacology 36 (2013) S39, https://doi.org/10.1016/j.bbi.2011.07.214. C.L. Raison, L. Capuron, A.H. Miller, Cytokines sing the blues: inflammation and the pathogenesis of depression, 27 (2006) 24–31, https://doi.org/10.1016/j.it.2005. 11.006. A.H. Miller, V. Maletic, C.L. Raison, Inflammation and its discontents: the role of cytokines in the pathophysiology of major depression, Biol. Psychiatry 65 (2009) 732–741, https://doi.org/10.1016/j.biopsych.2008.11.029. N. Müller, M.J. Schwarz, S. Dehning, A. Douhe, A. Cerovecki, B. Goldstein-Müller, I. Spellmann, G. Hetzel, K. Maino, N. Kleindienst, H.J. Möller, V. Arolt, M. Riedel, 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 (2006) 680–684, https://doi.org/10.1038/sj.mp. 4001805. O. Köhler, M.E. Benros, M. Nordentoft, M.E. Farkouh, R.L. Iyengar, O. Mors, J. Krogh, Effect of anti-inflammatory treatment on depression, depressive symptoms, and adverse effects, JAMA Psychiatry 71 (2014) 1381, https://doi.org/10. 1001/jamapsychiatry.2014.1611. T.-L. Huang, F.-C. Lin, High-sensitivity C-reactive protein levels in patients with major depressive disorder and bipolar mania, Prog. Neuro-Psychopharmacol. Biol. Psychiatry 31 (2007) 370–372, https://doi.org/10.1016/j.pnpbp.2006.09.010. R. Haapakoski, J. Mathieu, K.P. Ebmeier, H. Alenius, M. Kivimäki, Cumulative meta-analysis of interleukins 6 and 1β, tumour necrosis factor α and C-reactive protein in patients with major depressive disorder, Brain Behav. Immun. 49 (2015) 206–215, https://doi.org/10.1016/j.bbi.2015.06.001. M.H. Rapaport, A.A. Nierenberg, P.J. Schettler, A. Cardoos, R. Walker, D. Mischoulon, Inflammation as a predictive biomarker for response to Omega-3 fatty acids in major depressive disorder: a proof of concept study, Mol. Psychiatry 21 (2016) 71–79, https://doi.org/10.1038/mp.2015.22. R. Yirmiya, I. Goshen, Immune modulation of learning, memory, neural plasticity and neurogenesis, Brain Behav. Immun. 25 (2011) 181–213, https://doi.org/10. 1016/j.bbi.2010.10.015. J.L. Warner-Schmidt, K.E. Vanover, Antidepressant effects of selective serotonin reuptake inhibitors (SSRIs) are attenuated by antiinflammatory drugs in mice and humans, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 9262–9267, https://doi.org/10. 1073/pnas.1104836108/-/DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas. 1104836108. H.O.S. Kim, H.O.S. Kim, J.C. Youn, E.C. Shin, S. Park, Serum cytokine profiles in healthy young and elderly population assessed using multiplexed bead-based immunoassays, J. Transl. Med. 9 (2011) 113, https://doi.org/10.1186/1479-5876-9113. T.A. Pearson, G.A. Mensah, R.W. Alexander, J.L. Anderson, R.O. Cannon, M. Criqui, Y.Y. Fadl, S.P. Fortmann, Y. Hong, G.L. Myers, N. Rifai, S.C. Smith, K. Taubert, R.P. Tracy, F. Vinicor, Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the centers for disease control and prevention and the American Heart Association, Circulation 107 (2003) 499–511, https://doi.org/10.1161/01.CIR. 0000052939.59093.45. S.H. Abbasi, F. Hosseini, A. Modabbernia, M. Ashrafi, S. Akhondzadeh, Effect of celecoxib add-on treatment on symptoms and serum IL-6 concentrations in patients with major depressive disorder: randomized double-blind placebo-controlled study, J. Affect. Disord. 141 (2012) 308–314, https://doi.org/10.1016/j.jad.2012.03.033. R. Musil, M.J. Schwarz, M. Riedel, S. Dehning, A. Cerovecki, I. Spellmann, V. Arolt, N. Muller, Elevated macrophage migration inhibitory factor and decreased transforming growth factor-beta levels in major depression – no influence of celecoxib treatment, J. Affect. Disord. 134 (2011) 217–225, https://doi.org/10.1016/j.jad. 2011.05.047. A. Inamdar, E. Merlo-Pich, M. Gee, C. Makumi, P. Mistry, J. Robertson, E. Steinberg, S. Zamuner, S. Learned, R. Alexander, E. Ratti, Evaluation of antidepressant properties of the p38 MAP kinase inhibitor losmapimod (GW856553) in major depressive disorder: results from two randomised, placebo-controlled, double-blind, multicentre studies using a Bayesian approach, J. Psychopharmacol. 28 (2014) 570–581 https://doi.org/10.1177/0269881114529377. D. Mehta, C.L. Raison, B.J. Woolwine, E. Haroon, E.B. Binder, A.H. Miller, J.C. Felger, Transcriptional signatures related to glucose and lipid metabolism predict treatment response to the tumor necrosis factor antagonist infliximab in patients with treatment-resistant depression, Brain Behav. Immun. 31 (2013) 205–215, https://doi.org/10.1016/j.bbi.2013.04.004. C.L. Raison, R.E. Rutherford, B.J. Woolwine, C. Shuo, P. Schettler, D.F. Drake, E. Haroon, A.H. Miller, A randomized controlled trial of the tumor necrosis factor antagonist infliximab for treatment-resistant depression, JAMA Psychiatry 70 (2013) 31, https://doi.org/10.1001/2013.jamapsychiatry.4. A. Gharekhani, M.R. Khatami, S. Dashti-Khavidaki, E. Razeghi, A.A. Noorbala,
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
[23]
[24]
[25] [26]
S.S. Hashemi-Nazari, M.A. Mansournia, The effect of omega-3 fatty acids on depressive symptoms and inflammatory markers in maintenance hemodialysis patients: a randomized, placebo-controlled clinical trial, Eur. J. Clin. Pharmacol. 70 (2014) 655–665, https://doi.org/10.1007/s00228-014-1666-1. S. Jazayeri, S.A. Keshavarz, M. Tehrani-Doost, M. Djalali, M. Hosseini, H. Amini, M. Chamari, A. Djazayery, Effects of eicosapentaenoic acid and fluoxetine on plasma cortisol, serum interleukin-1beta and interleukin-6 concentrations in patients with major depressive disorder, Psychiatry Res. 178 (2010) 112–115, https:// doi.org/10.1016/j.psychres.2009.04.013. S. Prakhinkit, S. Suppapitiporn, H. Tanaka, D. Suksom, Effects of Buddhism walking meditation on depression, functional fitness, and endothelium-dependent vasodilation in depressed elderly, J. Altern. Complement. Med. 20 (2014) 411–416, https://doi.org/10.1089/acm.2013.0205. National Collaborating Centre for Mental Health, Depression, Management of Depression in Primary and Secondary Care (National Clinical Practice Guideline 23), Engl. Natl. Inst. Clin. Excell, London, 2004. U. Krügel, J. Fischer, S. Radicke, U. Sack, H. Himmerich, Antidepressant effects of
[27]
[28]
[29]
[30]
112
TNF-α blockade in an animal model of depression, J. Psychiatr. Res. 47 (2013) 611–616, https://doi.org/10.1016/j.jpsychires.2013.01.007. S. Lee, H.-B. Kim, E.-S. Hwang, E. Kim, S.-S. Kim, T.-D. Jeon, M. Song, J.-S. Lee, M.C. Chung, S. Maeng, J.-H. Park, Antidepressant-like effects of p-Coumaric acid on LPS-induced depressive and inflammatory changes in rats, Exp. Neurobiol. 27 (2018) 189, https://doi.org/10.5607/en.2018.27.3.189. R.J. Tynan, J. Weidenhofer, M. Hinwood, M.J. Cairns, T.A. Day, F.R. Walker, A comparative examination of the anti-inflammatory effects of SSRI and SNRI antidepressants on LPS stimulated microglia, Brain Behav. Immun. 26 (2012) 469–479, https://doi.org/10.1016/j.bbi.2011.12.011. L. Sutcigil, C. Oktenli, U. Musabak, A. Bozkurt, A. Cansever, O. Uzun, S.Y. Sanisoglu, Z. Yesilova, N. Ozmenler, A. Ozsahin, A. Sengul, Pro- and anti-inflammatory cytokine balance in major depression: effect of sertraline therapy, Clin. Dev. Immunol. 2007 (2007), https://doi.org/10.1155/2007/76396. B.M. Campbell, E. Charych, A.W. Lee, T. Moller, Kynurenines in CNS disease: regulation by inflammatory cytokines, Front. Neurosci. 8 (2014) 1–22, https://doi.org/ 10.3389/fnins.2014.00012.
Contents lists available at ScienceDirect
Physiology & Behavior journal homepage: www.elsevier.com/locate/physbeh
Review
The efficacy of anti-inflammatory treatment interventions on depression in individuals with major depressive disorder and high levels of inflammation: A systematic review of randomized clinical trials
T
David J. Allison , Bhanu Sharma, Brian W. Timmons ⁎
Child Health & Exercise Medicine Program, McMaster University, Hamilton, ON, Canada
ARTICLE INFO
ABSTRACT
Keywords: Major depressive disorder Depression Inflammation Cytokine Anti-inflammatory
Background: Studies evaluating the efficacy of anti-inflammatory treatment strategies for major depressive disorder (MDD) commonly include participants who do not demonstrate elevated concentrations of baseline inflammatory mediators, or simply fail to assess baseline inflammation. This may result in an underestimation of the efficacy of such treatment strategies. Sampling and methods: This systematic review included randomized controlled trials related to the use of antiinflammatory treatment strategies in individuals who demonstrate elevated concentrations of inflammatory mediators, identified by searching OVID MEDLINE, OVID EMBASE, and OVID PsychINFO. Results: Fifty-one randomized controlled trials were identified via a title and abstract screen, of which 43 were excluded following full-text screening. Of the 8 included trials (n = 437), 5 utilized anti-inflammatory pharmaceuticals, 2 utilized omega-3 fatty acids, and 1 utilized exercise and meditation. Among these studies, only the trial related to exercise and meditation demonstrated both an elevation in baseline inflammation, and a significant reduction in inflammation following intervention. Issues related to insufficient interventions and/or lack of reporting of inflammatory mediators at one or more time-points were common. Conclusion: Among the growing number of studies which examine the potential antidepressant benefit of antiinflammatory treatment strategies, few have studied populations which demonstrate elevated baseline concentrations. Further, studies commonly fail to induce significant changes in inflammation following intervention. Together, this may explain the low efficacy frequently reported with such interventions. Future studies which utilize both stronger interventions (sufficient to reduce levels of inflammation), and participants with elevated baseline concentrations, may produce a more substantial influence on symptoms of (MDD).
1. Introduction Numerous studies have demonstrated that major depressive disorder (MDD) is associated with an activation of the inflammatory response system and elevated concentrations of proinflammatory mediators such as interleukin (IL)-6, IL-1β, and tumour necrosis factor (TNF) α [1–4]. Elevations in proinflammatory mediators have been shown to be capable of influencing symptoms of MDD via several mechanisms related to monoamine neurotransmission, glutamate neurotransmission, and hippocampal functioning/neurogenesis [5,6]. The ability of inflammatory mediators to influence brain chemistry and functioning may contribute to the ineffectiveness of traditional antidepressant medications (which target monoamine pathways) in certain subsets of depressed patients [7]. Depressed patients with increased concentrations of inflammatory mediators have been shown to have a ⁎
higher likelihood of demonstrating treatment resistance [7]. Additionally, improvements in depression following treatment with traditional anti-depressant pharmaceuticals have been shown to be associated with a reduction in inflammation [7]. Growing evidence supporting an inflammatory etiology of MDD has led to numerous studies which have evaluated the efficacy of anti-inflammatory treatment strategies. Despite evidence pertaining to the various mechanisms by which proinflammatory mediators are likely to contribute to symptoms of MDD, previous studies often demonstrate mixed results related to efficacy. It is important to note however, that past studies have frequently failed to assess baseline levels of inflammation and/or performed anti-inflammatory interventions on depressed individuals who do not demonstrate signs of chronic inflammation. For example, a study by Müller et al. 2006 [8], demonstrated significantly greater improvements in depression
Corresponding author at: Department of Kinesiology, Faculty of Science, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4L8, Canada. E-mail address: [email protected] (D.J. Allison).
https://doi.org/10.1016/j.physbeh.2019.05.006 Received 10 September 2018; Received in revised form 25 April 2019; Accepted 8 May 2019 Available online 09 May 2019 0031-9384/ © 2019 Elsevier Inc. All rights reserved.
105
60/20/20/20
Jazayeri et al. [23]
54/27/27
Gharekhan et al. (2014)
128/64/64
40/20/20
Abbasi et al. [17]
Inamdar et al. [19]
Randomized (Total/ Treatment/ Control)
No. of patients
Patient population
Source
Study
42/14/14/14
101/51/50
45/25/20
37/19/18
Analyzed (Total/ Treatment/ Control)
29%/29%/33%
Not stated
56.8 ± 13.09/ 57.2 ± 15.19
52%/60%
65%/70% 35.1 ± 8.0/ 34.2 ± 6.9
% Male & Age (yrs) (Treatment/ Control)
Characteristics
Fluoxetine (20 mg/daily/
None/None
Placebo (7.5 mg/daily)
Placebo (paraffin oil (1800 mg daily/ 4 months) ACE inhibitors; ARB; Statins Hemodialysis patients Losmapimod (7.5 mg daily)
Sertraline + Celecoxib (400 mg daily/ 6 weeks) Sertraline + Placebo (400 mg daily/6 weeks) Sertraline (constant dose) No Comorbidities Omega-3 (1800 mg daily/ 4 months)
Treatment/ Comedication/ Comorbidity
Fluoxetine: IL-6: 2.12 ± 1.44
IL-10:1.71 ± 1.07 pg/ml TNF-RII:1682 ± 1090.02 pg/ml
Fluoxetine: IL-6: 1.83 ± 1.51
IL-6: 6.07 ± 7.65 pg/ml TNF-alpha: 18.01 ± 73.88 pg/ml IL-10:1.78 ± 1.33 pg/ml TNFRII:1535.06 ± 1025.18 pg/ml
IL-6: 7.10 ± 11.59 pg/ml TNF-alpha: 11.67 ± 36.19 pg/ml
Placebo IL-6 2.56 ± 0.64 pg/ml Omega-3 CRP: 8.0 ± 6.78 mg/l IL-6: 136.85 ± 183.56 ng/l IL-10: 132.06 ± 94.81 pg/ml TNFα: 116.57 ± 172.32 ng/l Placebo CRP: 12.69 ± 14.02 mg/l IL-6: 106.48 ± 136.51 ng/l IL-10: 107.24 ± 63.01 pg/ml TNFα: 77.43 ± 59.82 ng/l
Placebo IL-6 2.78 ± 0.72 pg/ml Omega-3 CRP: 9.24 ± 8.79 mg/l IL-6: 144.38 ± 167.66 ng/l IL-10: 123.94 ± 71.91 pg/ml TNFα: 122.09 ± 171.69 ng/l Placebo CRP: 7.73 ± 4.89 mg/l IL-6: 103.54 ± 145.18 ng/l IL-10: 111.76 ± 67.39 pg/ml TNFα: 80.93 ± 64.89 ng/l
Losmapimod IL-6: 5.21 ± 9.71 pg/ml TNF-alpha: 5.46 ± 2.02 pg/ml IL-10: 2.03 ± 1.36 pg/ml TNF-RII: 1496.63 ± 987.15 pg/ml Placebo
2.16 ± 0.60 pg/ml
2.79 ± 0.76 pg/ml
Losmapimod IL-6: 5.77 ± 9.01 pg/ml TNF-alpha: 5.98 ± 2.58 pg/ml IL-10: 4.10 ± 10.94 pg/ml TNF-RII: 1841.07 ± 1292.82 pg/ml Placebo
Celecoxib IL-6
Post-treatment
Celecoxib IL-6
Baseline
Inflammatory mediators
Table 1 Study characteristics of identified clinical trials investigating anti-inflammatory interventions for depression in individuals with elevated inflammation.
Placebo 20.33 ± 7.56
Placebo 21 ± 4.72
HamD
HAMD-17 = 24.5 QIDS-SR16 = 18.0”
BDI = 12.6 IDS-C = 45.0
Placebo
(continued on next page)
BDI = 5.94 HAMD-17 = 11.94 IDS-C = 22.93 IDS-SR = 25.35 QIDS-SR16 = 8.74” HamD
IDS-SR = 29.34 QIDS-SR16 = 10.31 Placebo
Change Scores Losmapimod BDI = 7.05 HAMD-17 = 14.90 IDS-C = 28.44
13.44 ± 5.66
23.52 ± 7.49
Losmapimod BDI = 12.9 IDS-C = 45.3 HAMD-17 = 24.49 QIDS-SR16 = 17.8
Placebo 11.25 ± 1.25 BDI Omega-3
8.8 ± 2.1
HamD Celecoxib
Post-treatment
Placebo 21.3 ± 1.9 BDI Omega-3
22.2 ± 1.7
HamD Celecoxib
Baseline
Depression
D.J. Allison, et al.
Physiology & Behavior 207 (2019) 104–112
57/14/13/15/ 15
Mehta et al. [20]
106
40/20/20
45/15/15/15
60/30/30
Musil et al. [18]
Prakhinkit et al. [24]
Raison et al. [21]
*treatment & controls groups divided into responders and non- responders
Randomized (Total/ Treatment/ Control)
No. of patients
Patient population
Source
Study
Table 1 (continued)
60/30/30
40/13/14/13
32/18/14/20
53/12/12/15/14
Analyzed (Total/ Treatment/ Control) 8 weeks) + EPA placebo EPA (1000 mg/daily/ 8 weeks) + fluoxetine placebo EPA (1000 mg/daily/ 8 weeks) + Fluoxetine (20 mg/ daily/8 weeks) None/None Infliximab (5 mg/kg at baseline, 2 weeks, 6 weeks) Placebo None/Constant antidepressant regimen (types not specified)
37 ± 8.49/ 34 ± 8.46/ 33.86 ± 10.85
33%/33%
74.8 ± 1.7/ 74.0 ± 1.9/ 81.0 ± 1.7
0%
Placebo
44.8 ± 12.2/ 44.3 ± 11.1/ 40.0 ± 10.4
Traditional Walking (20 min: 3×/week) Sedentary None/None Infliximab (5 mg/kg at
None/ Benzodiazepine & Reboxetine at variable doses Meditative Walking (20 min: 3×/week)
Celecoxib (400 mg/daily/ 6 weeks)
7.6/ 8.0/ 10.0/ 8.6
61%/36%/75%
33.3%/ 33.3% 38.9 ± 46.0 ± 42.1 ± 46.6 ±
35.7%/38.5%/
Treatment/ Comedication/ Comorbidity
% Male & Age (yrs) (Treatment/ Control)
Characteristics
Not reported (no time x group interaction)
IL-6: 1.2 ± 0.4 (pg/ml)
IL-6: 0.8 ± 0.2 (pg/ml) Infliximab: CRP: 6.3 ± 8.9 mg/l
Traditional Walking CRP: 6.2 ± 2.2 (mg/l) IL-6: 1.1 ± 0.5 (pg/ml) Walking + Meditation CRP: 4.7 ± 1.5 (mg/l) IL-6: 0.7 ± 0.2 (pg/ml) Sedentary CRP: 4.2 ± 1.9 (mg/l)
Celecoxib: MIF: 49.2 ± 30.0 TGF-β: 9968 ± 2311 sCD14: 1441 ± 187 Placebo: MIF: 23.11 ± 8.3 TGF- β: 9968 ± 2311 sCD14: 1104 ± 772
CRP not assessed post Post Cytokine Transcripts not provided
IL-1B: 4.73 ± 3.37 EPA: IL-6: 2.38 ± 2.21 IL-1B: 5.31 ± 5.22 Fluoxetine+EPA: IL-6: 4.56 ± 7.28 IL-1B: 5.03 ± 4.01
Post-treatment
Traditional Walking CRP: 8.1 ± 2.4 (mg/l) IL-6: 0.7 ± 0.1 (pg/ml) Walking + Meditation CRP: 6.3 ± 1.8 (mg/l) IL-6: 0.9 ± 0.2 (pg/ml) Sedentary CRP: 3.8 ± 1.9 (mg/l)
CRP: 7.4 mg/l Placebo (R): CRP: 7.8 mg/l Placebo (NR): CRP: 3.1 mg/l Celecoxib: MIF: 32.7 ± 16.2 TGF-β: 9484 ± 3081 sCD14: 1474 ± 285 Placebo: MIF: 30.8 ± 11.7 TGF- β:8554 ± 2266 sCD14: 1279 ± 608
Infliximab (R):
Infliximab (NR): CRP: 6.0 mg/l
IL-1B: 4.21 ± 2.97 EPA: IL-6: 2.11 ± 1.44 IL-1B: 6.12 ± 3.56 Fluoxetine+EPA: IL-6: 4.21 ± 8.37 IL-1B: 2.55 ± 2.84
Baseline
Inflammatory mediators
(continued on next page)
Sed. Control: 18.6 HamD
Walking+Medit. 8.6*
Walking+Medit. 16.8
Sed. Control: 17.9 HamD Inflaximab: 24.1
GDS Trad. Walking: 15.5
Placebo: 16.9 ± 8.6
Infliximab (R): 12.7 ± 6.8 Placebo (R): 20.2 ± 6.7 Placebo (NR): 8.1 ± 3.8” HamD Celexcoxib: 11.8 ± 7.7
HamD Infliximab (NR): 21.9 ± 7.6
Not reported
Post-treatment
GDS Trad. Walking: 17.3
Placebo: 23.2 ± 5.5
Placebo (NR): 23.3 ± 3.6” HamD Celecoxib: 25.5 ± 4.1
Infliximab (R): 23.6 ± 11.2 Placebo (R): 23.9 ± 4.1
HamD Infliximab (NR): 24.1 ± 3.6
Fluoxetine + EPA: 30.86 ± 5.41
Fluoxetine: 29 ± 6.93 EPA: 29.29 ± 4.84
Baseline
Depression
D.J. Allison, et al.
Physiology & Behavior 207 (2019) 104–112
Physiology & Behavior 207 (2019) 104–112
None/37 participants on stable antidepressant dose
Placebo CRP: 5.4 ± 8.2 mg/l TNF-α: not reported
Abbreviations: HAMD: Hamilton Depression Rating Scale; BDI: Beck Depression Inventory; GDS: Geriatric Depression Scale; MADRS: Montgomery-Asberg Depression Rating Scale; IDSeC: Inventory of Depressive Symptomatology: Clinician; QIDS-SR16: Quick Inventory of Depressive Symptomatology; ACE: angiotensin-converting-enzyme; ARB: angiotensin 2 receptor blocker; EPA: eicosapentaenoic acid; CRP: c-reactive protein; IL: interleukin; TNF: tumour necrosis factor; MIF: macrophage migration inhibitory factor; TGF: transforming growth factor; sCD14: soluble cluster of differentiation 14.
Placebo: 23.6
Post scores not reported (no time x group interaction) baseline, 2 weeks, 6 weeks) Placebo 42.5 ± 8.2/ 44.3 ± 9.4
TNF-α: not reported
Treatment/ Comedication/ Comorbidity % Male & Age (yrs) (Treatment/ Control) Randomized (Total/ Treatment/ Control) Source
No. of patients
Patient population Study
Table 1 (continued)
Analyzed (Total/ Treatment/ Control)
Characteristics
Baseline
Inflammatory mediators
Post-treatment
Baseline
Depression
Post-treatment
D.J. Allison, et al.
following a 6-week treatment with reboxetine combined with the antiinflammatory drug celecoxib (in comparison to reboxetine alone) [8]. The effectiveness of this intervention may be driven by a successful reduction in elevated baseline levels of inflammation; however, inflammation was not assessed. The frequent omission of baseline inflammation in the literature is apparent in the systematic review by Köhler et al. 2014 [9], which examined the effect of anti-inflammatory treatments on MDD whereby, of the 14 trials included, only 2 reported baseline scores of inflammation [9]. This may be common practice due to the assumption that depression is typically regarded as a chronic inflammatory condition. However, it is critical to acknowledge that not all cases of depression are associated with heightened levels of inflammation [2,10,11]. Failure to assess baseline inflammation or the use of study populations which do not demonstrate elevated concentrations of inflammation may contribute to the limited efficacy of many previous studies which utilize anti-inflammatory treatment strategies. The use of such populations may be confounding for several reasons. First, participants with healthy concentrations of inflammatory mediators are simply not likely to be negatively influenced by inflammation and therefore would not benefit from an anti-inflammatory intervention. Second, there is evidence to suggest that when at healthy concentrations, proinflammatory cytokines may play critical roles related to neural plasticity, neurogenesis, and synaptic strengthening which would suggest a beneficial role [12,13]. Third, a study by WarnerSchmidt et al. 2011 provided evidence that anti-inflammatory drugs may actually attenuate the benefits of serotonergic antidepressants [14]. It may therefore be possible that the implementation of unnecessary anti-inflammatory strategies in depressed individuals, who do not demonstrate elevated concentrations of inflammatory mediators is ineffective and/or detrimental to symptom relief. The fact that not all cases of MDD are associated with high inflammation make it inappropriate to expect anti-inflammatory interventions to be an effective treatment for the entirety of the depressed population. Instead, it would seem more likely that there may be a subset of the population with pathological concentrations of inflammatory mediators who demonstrate a form of inflammation-induced depression. Based on this assumption, it may be possible that past studies which have failed to assess baseline inflammation, and previous systematic reviews which utilized studies which did not assess baseline inflammation, may have understated the efficacy of anti-inflammatory treatment strategies for MDD. Therefore, the purpose of this systematic review was to evaluate the efficacy of anti-inflammatory strategies for the treatment of MDD in individuals with elevated baseline concentrations of inflammatory mediators. The identification and utilization of a better characterized sample may help to elucidate the true potential of such strategies, and help better understand the current state of the science. 2. Materials and methods The current systematic review included randomized clinical trials which investigated anti-inflammatory interventions for the treatment of depression in individuals with clinical depression and elevated concentrations of inflammatory mediators at baseline. This review was registered with the international prospective register of systematic reviews PROSPERO network (registration number CRD42017059049). 2.1. Eligibility criteria Only randomized, controlled clinical trials were included in the current review. Participants had a clinical diagnosis of depression or self-report questionnaire scores suggesting depression. As this review was interested in evaluating the efficacy of anti-inflammatory treatments for depression, only studies which included baseline assessments of inflammation, allowing for the assessment of elevated concentrations 107
Physiology & Behavior 207 (2019) 104–112
Screening
Idenficaon
D.J. Allison, et al.
Records iden!fied through database searching: N = 1618
Records screened a#er duplicates removed:
Records Excluded: N = 1399
N = 1451
Full text ar!cles assessed for eligibility
Full text ar!cles excluded Depression not assessed / scores suggest lack of depression (n=6) Baseline inflamma!on not assessed (n=9) Abstract/conference publica!on (n=17) Inappropriate study design (n=11)
Eligibility
N = 51
Studies included in qualita!ve analysis
Included
N=8
Studies included in quan!ta!ve analysis N=0
Fig. 1. PRISMA (preferred reporting items for systematic reviews and meta-analyses) flow diagram of study selection for systematic review.
Table 2 Risk of bias, as assessed on the Cochrane risk of bias tool. Source
Abbasi et al. [17] Gharekhani et al. [22] Inamdar et al. [19] Jazayeri et al. [23] Mehta et al. [20] Musil et al. [18] Prakhinkit et al. [24] Raison et al. [21]
Allocation Sequence generation
Concealment
Low Low Low Low Low Low Low Low
Low High Low Low Low High Low High
Blinding of participants and personnel
Blinded outcome assessment
Incomplete outcome data
Selective reporting
Other bias
Low Low Low Low Low Low High Low
Low High Low Low Low Low High Low
Low High High High High High High Low
Low Low Low Low Low Low Low Low
Low Low High Low Low High Low Low
of inflammatory mediators, were included. Studies utilizing pharmaceutical drugs in addition to anti-inflammatory interventions would only be included if the pharmaceuticals were administered at a constant dose throughout the intervention. We identified anti-inflammatory treatment interventions as NSAIDs, anti-cytokine treatments, or lifestyle interventions (diet and exercise). No limitations or restrictions were placed on language or date of publication.
sciences librarian for the following databases: OVID MEDLINE, OVID EMBASE, and OVID PsychINFO. Independent searches were run for each database (see Table 1 for an example of a search strategy). All trials published prior to August 22, 2018, were searched for using terms related to depression, various immune mediators, and anti-inflammatory strategies. Two investigators (D.A. and B.S.) independently reviewed titles and abstracts to remove any obvious irrelevant studies using Covidence. All titles/abstracts deemed appropriate by both reviewers were automatically advanced to the full-text screening stage by Covidence. Any conflicts between reviewers regarding inclusion/exclusion of an article were discussed until a consensus was reached. The
2.2. Search methods for identification of trials The search strategy was optimized with the assistance of a health 108
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
[16]. CRP scores greater than 3.0 mg/L suggest unhealthy levels of CRP and a high risk for future heart attack and stroke, and was used as a cutoff for high inflammation in the current study.
Table 3 Example of search strategy – optimized for OVID EMBASE. 1. 2. 3. 4. 5. 6. 7.
exp. Depression/ Depress*.mp. 1 or 2 exp. Cytokines/ cytokines.mp. 4 or 5 (Chemokine or Interferon or Interleukin or Lymphokine or Monokine or Leukocytes or Tumour Necrosis Factor).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 8. 6 or 7 9. exp. Anti-Inflammatory Agents/ 10. anti-inflammatory agents.mp. 11. 9 or 10 12. 3 and 8 and 11 13. (selective serotonin reuptake inhibitor or SSRI).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 14. (TCA or Trycyclic Antidepressant).mp. [mp = title, abstract, heading word, drug trade name, original title, device manufacturer, drug manufacturer, device trade name, keyword, floating subheading] 15. 13 and 14 16. 12 not 15 17. exp. controlled clinical trial/ 18. exp. randomized controlled trial/ 19. exp. crossover procedure/ 20. exp. double-blind procedure/ 21. exp. single-blind procedure/ 22. random*.mp. 23. factorial*.mp. 24. (crossover* or cross over* or cross-over*).mp. 25. placebo*.mp. 26. (double* adj blind*).mp. 27. (singl* adj blind*).mp. 28. assign*.mp. 29. allocat*.mp. 30. 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 or 27 or 28 or 29 31. exp. human/ 32. 16 and 30 and 31
2.5. Assessment of bias The bias risks of the randomized clinical trials included were assessed (D.A. and B.S.) based on the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions and methodological studies. We extracted data regarding quality for 7 domains. Sequence generation was considered adequate if authors described a random component. Allocation concealment was adequate if it was justified that neither participants nor investigators were aware of participant assignment. Blinding of participants and personnel was adequate if knowledge of the allocated intervention was prevented. Blinding of outcome assessors was adequate if the trial was characterized as double-blind, based on our own assessment of the trial. Incomplete outcome data were considered to be addressed appropriately if attrition and exclusions were reported with rationale and appropriately addressed. Selective reporting was considered adequate if all pre-specified primary outcomes were reported. Overall, trials were considered to have a low risk of bias if all the above domains were followed. Any trial with uncertain risk of bias or a high risk of bias in one or more domain were considered high risk. 3. Results 3.1. Trial flow
same two investigators also independently reviewed the remaining fulltext reports against the inclusion criteria via a similar process.
Results of the searches are presented in Fig. 1. Following the removal of duplicates, the title/abstract of 1452 unique papers were screened for eligibility. A total of 1401 papers were excluded, while 51 papers were included in the full-text screen for eligibility. Following the full-text screen, an additional 43 studies were excluded. Reasons for exclusion included the failure to assess baseline depression or the inclusion of participants with baseline scores suggesting a lack of clinical depression (n = 6), failure to assess baseline inflammation (n = 9), non-peer reviewed resources (n = 17), and inappropriate study design (n = 11).
2.3. Data extraction
3.2. Search results and study characteristics
Data were extracted independently by each investigator (D.A. and B.S.) using the same pre-structured form. The reviewers were not blinded to the study results, authors, or institutions. In addition to bibliographic information, data extraction included quality assessment, description of the participants, description of the intervention and control groups, and outcome measures related to depression and concentrations of inflammatory mediators. All data was cross-verified for accuracy by both reviewers.
Details from the 8 randomized controlled trials which met inclusion criteria are summarized in Table 2. The 8 included trials comprised a total of 437 participants. Of the 8 included trials, 5 (involving 282 participants) evaluated the efficacy of anti-inflammatory pharmaceuticals, including the NSAID, Celecoxib [17,18], the P38 mitogen-activated protein kinase inhibitor, Losmapimod [19] and the anti-tumour necrosis factor-α biologic, Infliximab [20,21]. Two studies (87 participants) evaluated the efficacy of the anti-inflammatory supplement omega-3 fatty acids [22,23]. One study (40 participants) examined the efficacy of meditative walking [24]. Due to the limited number of articles which showed both an elevated baseline concentration of inflammatory mediators and a successful reduction in inflammatory mediators following intervention, a quantitative analysis was not conducted. Instead, the studies were analyzed narratively.
2.4. Outcome measures Primary outcome measures included (1) the change in depressive symptoms measured on a continuous scale from baseline to the end of the intervention, (2) the change in inflammatory mediators, measured on a continuous scale, from baseline to the end of the intervention, and (3) the relationship between changes in scores of depression and changes in inflammatory mediators from baseline to the end of the intervention. Healthy cytokine concentrations for IL-1β, IL-2, IL-6, TNF-α, IFN-y, IL-10, and IL-4 were based on serum cytokine profiles obtained from healthy young and elderly individuals [15]. Healthy concentrations of C-reactive protein (CRP) were based on guidelines for low, moderate, and high risk for a cardiovascular event, endorsed by the centers for disease control and prevention and the American heart association
3.3. Pharmaceutical interventions Among the 5 studies which utilized anti-inflammatory pharmaceuticals, 2 failed to significantly reduce concentrations of inflammatory mediators in the treatment group [18,19], and 2 failed to report values of post inflammatory mediators [20,21]. The study by Abbasi et al. 2012 [17] was the only trial to achieve significant reductions in inflammation as shown by the reduction in the proinflammatory cytokine 109
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
IL-6 following the intervention with Celecoxib (200 mg, 2× daily). It should be noted, however, that IL-6 was the only inflammatory mediator assessed in the trial and baseline concentrations were within a healthy range when compared to the normative values from the study by Kim et al. [15]. Additionally, although a significant group x time interaction was observed, whereby the treatment group experienced a significantly greater reduction in IL-6, the placebo group also experienced a significant reduction in IL-6. In relation to changes in scores of depression, only the study by Abassi et al. 2012 showed a significant group x time interaction whereby the treatment group had a significantly greater improvement in Ham-D scores of depression. Post-hoc analysis revealed a significant reduction in the placebo group. Significant correlations between the change in IL-6 and the change in Ham-D scores of depression were also shown in both groups.
mediators make it difficult to accurately assess the efficacy of anti-inflammatory treatments for depression. Despite challenges with numerous confounding variables among human studies, prior research in animal models has produced promising results. A study by Krügel et al. 2013 [26] demonstrated the efficacy of the TNF-α inhibitor etanercept at reducing depressive symptoms in male Wistar rats [26]. Further, in a recent study, Lee et al. [27], assessed the efficacy of p-Coumaric Acid administration (a dietary phenolic acid with anti-inflammatory properties) following LPS-induced depression in rats [27]. It was shown to protect against depressive symptoms as shown by significant differences in behavioural, electrophysiological, and brain biochemical measures between treatment and control groups. The effectiveness of such anti-inflammatory interventions in animal models may relate to the ability to induce, and thereby ensure, baseline elevations in inflammation and depressive symptoms in animal models. The need to utilize human populations who demonstrate elevated concentrations of inflammatory mediators is further emphasized in the study by Raison et al. [5]. This study showed no significant difference between groups in the change of Ham-D scores of depression, or CRP concentrations. However, an exploratory analysis which focused on a subgroup of participants with CRP concentrations above 5 mg/L showed a 3.1-point greater reduction in Ham-D scores in the treatment group compared to placebo. This sub-analysis, involving a relatively small sample size (n = 22), did not reach statistical significance, however, a 3.1-point reduction would be considered clinically meaningful [25]. This may suggest, anti-inflammatory treatment strategies would produce treatment effects of greater magnitude if implemented in populations with elevated concentrations of inflammatory mediators. Findings like this, in conjunction with evidence suggesting that healthy concentrations of proinflammatory cytokines have various beneficial roles related to neural plasticity, neurogenesis, and synaptic strengthening [12,13], further emphasize the need to utilize participants which demonstrate chronic inflammation when assessing the efficacy of antiinflammatory strategies. The concurrent use of antidepressant pharmaceuticals may also influence the findings of studies which aim to evaluate the efficacy of anti-inflammatory treatments. In the current review, 5 of the 8 selected studies included participants who used some form of antidepressant pharmaceutical related to monoamine neurotransmission. Studies have shown that traditional antidepressant pharmaceuticals such as selective serotonin reuptake inhibitors (SSRIs), or selective norepinephrine reuptake inhibitors (SNRIs) may induce a considerable anti-inflammatory effect [28,29]. This may have contributed to the significant reduction in inflammation shown in the placebo group in the study by Abassi et al. 2012. In this study, the use of the non-steroidal anti-inflammatory, selective COX-2 inhibitor, Celecoxib, resulted in a significant reduction in IL-6 concentrations as well as a related significant reduction in HamD scores of depression. Interestingly, however, the placebo group also achieved a significant reduction in IL-6 concentrations and a related significant reduction in Ham-D scores of depression. Such results may relate to the continued use of the SSRI sertraline by the placebo group. Although this study still achieved a significant treatment effect, the magnitude of the effect may have been dampened due to the use of a control group which experienced a significant reduction in inflammation, potentially due to Sertraline usage. Among the 8 studies included in this review, only the study by Prakhinkit et al. [24] used both chronically inflamed, and non-medicated participants. In this study, which utilized meditative walking therapy as an anti-inflammatory treatment, a significant group by time interaction was shown for both CRP and scores of depression, as shown by the geriatric depression scale. Post-hoc analysis revealed significant reductions in CRP and scores of depression in the treatment group, and no change in either variable in the control group. Although correlations were not assessed, results from this study may highlight the importance of utilizing a patient population with elevated concentrations of
3.4. Supplement interventions Among the 2 studies which utilized omega-3 supplements [22,23], both demonstrated at least one proinflammatory mediator to be elevated above healthy concentrations at baseline. Neither study, however, demonstrated significant group x time effects for any inflammatory mediator. Improvements in scores of depression in the treatment group were observed only in the study by Gharekhani et al. 2014 [22]; however, no relation between changes in inflammatory mediators and changes in depression were apparent. 3.5. Meditative walking intervention The study which utilized meditative walking [24] demonstrated a significant elevation in baseline concentration of CRP in all groups. A significant group x time interaction was also demonstrated following the intervention consisting of meditative walking for 20 min, 3 times per week. Post hoc analysis showed a significant reduction in CRP concentrations in the treatment group and no change in the sedentary control group. CRP concentrations did however, remain well above normative healthy values post intervention [25]. A significant group x time interaction was also shown for scores of depression as assessed using the Geriatric Depression Scale. Post hoc analysis showed a significant reduction in scores of depression in the treatment group and no change in the control group. Analysis of the potential relationship between the change in depression and change in CRP concentrations was, however, not reported. 3.6. Bias of included trials Assessment of bias for the included trials is shown in Table 2. Six of the eight trials were associated with a high risk of bias. Each of these 7 trials were judged to be at high risk for bias related to incomplete outcome data. Certain trials were also deemed to be high risk for factors related to allocation concealment, blinded outcome assessment, and other biases related to for-profit bias (Table 3). 4. Discussion Following full-text review, a total of 22 original studies with an appropriate study design were identified. Among these studies, failure to assess baseline inflammation and the use of participants who did not demonstrate scores suggesting clinical depression were common, resulting in the exclusion of 14 randomized controlled trials. Among the trials included in the current review, failure to significantly reduce concentrations of inflammatory mediators to an extent greater than that of the control group was apparent in 6 of the 8 studies [18–23]. The common use of participants who do not show elevated concentrations of inflammatory mediators as well as the use of insufficient interventions which failed to effectively reduce concentrations of inflammatory 110
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
inflammatory mediators and who are free from potentially confounding anti-depressant pharmaceuticals when assessing the efficacy of antiinflammatory interventions for the treatment of depression.
[3]
4.1. Limitations and future directions Although the assessment of baseline inflammation is important when evaluating the efficacy of anti-inflammatory treatments for depression, high heterogeneity between studies, and a lack of well-established normative data make it difficult to set definitive cutoffs for healthy vs pathological levels of inflammatory biomarkers. CRP is perhaps the most commonly reported indicator of inflammation, and has the most well-established normative data. CRP may however, not be the most influential inflammatory mediator in terms of its influence on symptoms of depression. Proinflammatory cytokines such as IFNy, IL-6, IL-1B, and TNF-α have been frequently shown to be associated with depression and influence various related mechanisms [1,7,30]. However, these cytokines also have less established normative values with relatively few studies assessing healthy concentrations. The high inter-study variability of concentrations of these mediators further necessitates the need for further studies. This makes establishing a set cut point for what constitutes high inflammation difficult and is a limitation to the current review, and other studies which define pathological levels of inflammation based on such inflammatory markers. More studies will be required in the future to help strengthen normative data. Future studies will also be required to further elucidate which inflammatory biomarkers are the most impactful in terms of depression, as well as what concentrations may be considered pathological. It may also be important for future studies to attempt to use medication free participants to avoid unintended anti-inflammatory effects. Although recruitment of an unmedicated depressed population may prove difficult, it may allow for a more accurate assessment of the efficacy of antiinflammatory interventions.
[4] [5] [6] [7] [8]
[9]
[10] [11]
[12]
[13]
5. Conclusion
[14]
Numerous studies have been conducted which relate to the efficacy of anti-inflammatory interventions for the treatment of MDD. This review highlights that many studies either fail to assess baseline concentrations of inflammation, or utilize participants who do not demonstrate elevated baseline concentrations of inflammatory mediators. Further, of the 8 studies included in the current review, whereby baseline inflammation was assessed, only 2 achieved reductions in inflammatory mediators to extent greater than that of the control group following intervention. There is a need for future studies which evaluate the efficacy of effective anti-inflammatory strategies for the treatment of MDD, which utilize participants who demonstrate chronic inflammation, and are free of comedications.
[15]
[16]
[17]
Acknowledgements
[18]
DJA was supported by a Michael G. DeGroote Fellowship Award. BS is supported by the Queen Elizabeth Science and Technology Award. BWT is supported by the Canada Research Chair in Child Health & Exercise Medicine.
[19]
Declaration of interest None.
[20]
References [21]
[1] A.H. Miller, C.L. Raison, The role of inflammation in depression: from evolutionary imperative to modern treatment target, Nat. Rev. Immunol. 16 (2015) 22–34, https://doi.org/10.1038/nri.2015.5. [2] Y. Dowlati, N. Herrmann, W. Swardfager, H. Liu, L. Sham, E.K. Reim, K.L. Lanctôt,
[22]
111
A meta-analysis of cytokines in major depression, Biol. Psychiatry 67 (2010) 446–457, https://doi.org/10.1016/j.biopsych.2009.09.033. S. Mostafavi, A. Battle, X. Zhu, J.B. Potash, M.M. Weissman, J. Shi, K. Beckman, C. Haudenschild, C. McCormick, R. Mei, M.J. Gameroff, H. Gindes, P. Adams, F.S. Goes, F.M. Mondimore, D.F. MacKinnon, L. Notes, B. Schweizer, D. Furman, S.B. Montgomery, A.E. Urban, D. Koller, D.F. Levinson, Type I interferon signaling genes in recurrent major depression: increased expression detected by whole-blood RNA sequencing, Mol. Psychiatry 19 (2014) 1267–1274, https://doi.org/10.1177/ 1948550614552729. M.B. Howren, D.M. Lamkin, J. Suls, Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis, Psychosom. Med. 71 (2009) 171–186, https://doi.org/10.1097/PSY.0b013e3181907c1b. C.L. Raison, J.C. Felger, A.H. Miller, Inflammation and treatment resistance in major depression: a perfect storm, Neuropsychopharmacology 36 (2013) S39, https://doi.org/10.1016/j.bbi.2011.07.214. C.L. Raison, L. Capuron, A.H. Miller, Cytokines sing the blues: inflammation and the pathogenesis of depression, 27 (2006) 24–31, https://doi.org/10.1016/j.it.2005. 11.006. A.H. Miller, V. Maletic, C.L. Raison, Inflammation and its discontents: the role of cytokines in the pathophysiology of major depression, Biol. Psychiatry 65 (2009) 732–741, https://doi.org/10.1016/j.biopsych.2008.11.029. N. Müller, M.J. Schwarz, S. Dehning, A. Douhe, A. Cerovecki, B. Goldstein-Müller, I. Spellmann, G. Hetzel, K. Maino, N. Kleindienst, H.J. Möller, V. Arolt, M. Riedel, 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 (2006) 680–684, https://doi.org/10.1038/sj.mp. 4001805. O. Köhler, M.E. Benros, M. Nordentoft, M.E. Farkouh, R.L. Iyengar, O. Mors, J. Krogh, Effect of anti-inflammatory treatment on depression, depressive symptoms, and adverse effects, JAMA Psychiatry 71 (2014) 1381, https://doi.org/10. 1001/jamapsychiatry.2014.1611. T.-L. Huang, F.-C. Lin, High-sensitivity C-reactive protein levels in patients with major depressive disorder and bipolar mania, Prog. Neuro-Psychopharmacol. Biol. Psychiatry 31 (2007) 370–372, https://doi.org/10.1016/j.pnpbp.2006.09.010. R. Haapakoski, J. Mathieu, K.P. Ebmeier, H. Alenius, M. Kivimäki, Cumulative meta-analysis of interleukins 6 and 1β, tumour necrosis factor α and C-reactive protein in patients with major depressive disorder, Brain Behav. Immun. 49 (2015) 206–215, https://doi.org/10.1016/j.bbi.2015.06.001. M.H. Rapaport, A.A. Nierenberg, P.J. Schettler, A. Cardoos, R. Walker, D. Mischoulon, Inflammation as a predictive biomarker for response to Omega-3 fatty acids in major depressive disorder: a proof of concept study, Mol. Psychiatry 21 (2016) 71–79, https://doi.org/10.1038/mp.2015.22. R. Yirmiya, I. Goshen, Immune modulation of learning, memory, neural plasticity and neurogenesis, Brain Behav. Immun. 25 (2011) 181–213, https://doi.org/10. 1016/j.bbi.2010.10.015. J.L. Warner-Schmidt, K.E. Vanover, Antidepressant effects of selective serotonin reuptake inhibitors (SSRIs) are attenuated by antiinflammatory drugs in mice and humans, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 9262–9267, https://doi.org/10. 1073/pnas.1104836108/-/DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas. 1104836108. H.O.S. Kim, H.O.S. Kim, J.C. Youn, E.C. Shin, S. Park, Serum cytokine profiles in healthy young and elderly population assessed using multiplexed bead-based immunoassays, J. Transl. Med. 9 (2011) 113, https://doi.org/10.1186/1479-5876-9113. T.A. Pearson, G.A. Mensah, R.W. Alexander, J.L. Anderson, R.O. Cannon, M. Criqui, Y.Y. Fadl, S.P. Fortmann, Y. Hong, G.L. Myers, N. Rifai, S.C. Smith, K. Taubert, R.P. Tracy, F. Vinicor, Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the centers for disease control and prevention and the American Heart Association, Circulation 107 (2003) 499–511, https://doi.org/10.1161/01.CIR. 0000052939.59093.45. S.H. Abbasi, F. Hosseini, A. Modabbernia, M. Ashrafi, S. Akhondzadeh, Effect of celecoxib add-on treatment on symptoms and serum IL-6 concentrations in patients with major depressive disorder: randomized double-blind placebo-controlled study, J. Affect. Disord. 141 (2012) 308–314, https://doi.org/10.1016/j.jad.2012.03.033. R. Musil, M.J. Schwarz, M. Riedel, S. Dehning, A. Cerovecki, I. Spellmann, V. Arolt, N. Muller, Elevated macrophage migration inhibitory factor and decreased transforming growth factor-beta levels in major depression – no influence of celecoxib treatment, J. Affect. Disord. 134 (2011) 217–225, https://doi.org/10.1016/j.jad. 2011.05.047. A. Inamdar, E. Merlo-Pich, M. Gee, C. Makumi, P. Mistry, J. Robertson, E. Steinberg, S. Zamuner, S. Learned, R. Alexander, E. Ratti, Evaluation of antidepressant properties of the p38 MAP kinase inhibitor losmapimod (GW856553) in major depressive disorder: results from two randomised, placebo-controlled, double-blind, multicentre studies using a Bayesian approach, J. Psychopharmacol. 28 (2014) 570–581 https://doi.org/10.1177/0269881114529377. D. Mehta, C.L. Raison, B.J. Woolwine, E. Haroon, E.B. Binder, A.H. Miller, J.C. Felger, Transcriptional signatures related to glucose and lipid metabolism predict treatment response to the tumor necrosis factor antagonist infliximab in patients with treatment-resistant depression, Brain Behav. Immun. 31 (2013) 205–215, https://doi.org/10.1016/j.bbi.2013.04.004. C.L. Raison, R.E. Rutherford, B.J. Woolwine, C. Shuo, P. Schettler, D.F. Drake, E. Haroon, A.H. Miller, A randomized controlled trial of the tumor necrosis factor antagonist infliximab for treatment-resistant depression, JAMA Psychiatry 70 (2013) 31, https://doi.org/10.1001/2013.jamapsychiatry.4. A. Gharekhani, M.R. Khatami, S. Dashti-Khavidaki, E. Razeghi, A.A. Noorbala,
Physiology & Behavior 207 (2019) 104–112
D.J. Allison, et al.
[23]
[24]
[25] [26]
S.S. Hashemi-Nazari, M.A. Mansournia, The effect of omega-3 fatty acids on depressive symptoms and inflammatory markers in maintenance hemodialysis patients: a randomized, placebo-controlled clinical trial, Eur. J. Clin. Pharmacol. 70 (2014) 655–665, https://doi.org/10.1007/s00228-014-1666-1. S. Jazayeri, S.A. Keshavarz, M. Tehrani-Doost, M. Djalali, M. Hosseini, H. Amini, M. Chamari, A. Djazayery, Effects of eicosapentaenoic acid and fluoxetine on plasma cortisol, serum interleukin-1beta and interleukin-6 concentrations in patients with major depressive disorder, Psychiatry Res. 178 (2010) 112–115, https:// doi.org/10.1016/j.psychres.2009.04.013. S. Prakhinkit, S. Suppapitiporn, H. Tanaka, D. Suksom, Effects of Buddhism walking meditation on depression, functional fitness, and endothelium-dependent vasodilation in depressed elderly, J. Altern. Complement. Med. 20 (2014) 411–416, https://doi.org/10.1089/acm.2013.0205. National Collaborating Centre for Mental Health, Depression, Management of Depression in Primary and Secondary Care (National Clinical Practice Guideline 23), Engl. Natl. Inst. Clin. Excell, London, 2004. U. Krügel, J. Fischer, S. Radicke, U. Sack, H. Himmerich, Antidepressant effects of
[27]
[28]
[29]
[30]
112
TNF-α blockade in an animal model of depression, J. Psychiatr. Res. 47 (2013) 611–616, https://doi.org/10.1016/j.jpsychires.2013.01.007. S. Lee, H.-B. Kim, E.-S. Hwang, E. Kim, S.-S. Kim, T.-D. Jeon, M. Song, J.-S. Lee, M.C. Chung, S. Maeng, J.-H. Park, Antidepressant-like effects of p-Coumaric acid on LPS-induced depressive and inflammatory changes in rats, Exp. Neurobiol. 27 (2018) 189, https://doi.org/10.5607/en.2018.27.3.189. R.J. Tynan, J. Weidenhofer, M. Hinwood, M.J. Cairns, T.A. Day, F.R. Walker, A comparative examination of the anti-inflammatory effects of SSRI and SNRI antidepressants on LPS stimulated microglia, Brain Behav. Immun. 26 (2012) 469–479, https://doi.org/10.1016/j.bbi.2011.12.011. L. Sutcigil, C. Oktenli, U. Musabak, A. Bozkurt, A. Cansever, O. Uzun, S.Y. Sanisoglu, Z. Yesilova, N. Ozmenler, A. Ozsahin, A. Sengul, Pro- and anti-inflammatory cytokine balance in major depression: effect of sertraline therapy, Clin. Dev. Immunol. 2007 (2007), https://doi.org/10.1155/2007/76396. B.M. Campbell, E. Charych, A.W. Lee, T. Moller, Kynurenines in CNS disease: regulation by inflammatory cytokines, Front. Neurosci. 8 (2014) 1–22, https://doi.org/ 10.3389/fnins.2014.00012.