Depression and Inflammation: What the Research Can Establish
Author: Angela Bogdanova · Published: October 8, 2026 · Editorial Policy
Research cluster: Depression & Depressive Disorders · Written by Artificial Sapiens
Inflammation and depression are connected in a reproducible, scientifically important, and still incompletely understood way. Studies repeatedly find that some people with depressive symptoms or major depressive disorder have higher concentrations of inflammatory markers than comparison groups. Experimental and genetic research identifies biological pathways through which immune activity may affect mood, motivation, energy, and behavior. Clinical trials are now examining whether targeting inflammation can help selected patients. The evidence supports these research directions, while leaving open important questions about causation, individual diagnosis, and treatment selection. Osimo et al. (2019); Miller and Raison (2016).
The practical conclusion is precise: inflammation can be one component of depression-related biology, but an elevated blood marker does not establish depression, and a depression diagnosis does not establish a state of systemic inflammation. Treatment decisions still depend on a comprehensive clinical evaluation, current guidelines, medical history, and individual circumstances. See the Hub's clinical depression overview and the clinical assessment pathway. NIMH; NICE NG222.
I. What “Inflammation” Means in Depression Research
Inflammation is a coordinated immune response that helps the body respond to infection, injury, and other disturbances. It involves immune cells, signaling molecules, blood vessels, tissue responses, and regulatory mechanisms. Acute inflammation can be protective and short-lived. In studies of depression, researchers more often investigate chronic or low-grade systemic inflammatory activity, commonly assessed through blood biomarkers rather than obvious redness, swelling, or fever. These processes interact with sleep, metabolism, stress regulation, and multiple organ systems. Miller and Raison (2016).
C-reactive protein (CRP) is a protein made mainly by the liver whose concentration rises with certain inflammatory processes. High-sensitivity CRP (hs-CRP) is an assay capable of measuring lower concentrations of the same protein; it is not a depression-specific test. Interleukin-6 (IL-6) is a cytokine involved in multiple immune and metabolic pathways. Tumor necrosis factor alpha (TNF-α), interleukin-1 beta (IL-1β), and fibrinogen are other commonly studied markers. These measures represent partly different aspects of physiology, and a single blood result cannot reveal the complete immune state of the brain. Osimo et al. (2019); Szabo et al. (2026).
The phrase “inflammatory depression” appears in research discussion as a proposed biological characterization of some depressive presentations. It is not an independent diagnosis in standard clinical classification systems, and there is no universally established laboratory threshold that confirms such a condition in an individual. Major depressive disorder is diagnosed using clinical criteria: the pattern and duration of symptoms, impairment, history, and relevant differential diagnoses. Read more about what a depression diagnosis involves. NIMH.
II. What Population Studies Actually Find
1. The association is measurable, but people vary substantially
A foundational systematic review by Osimo and colleagues pooled studies involving 13,541 participants with depression and 155,728 controls. Using CRP above 3 mg/L as its research definition of low-grade inflammation, the review estimated a prevalence of 27% among people with depression (95% confidence interval, 21%–34%). The pooled odds of exceeding that threshold were higher in depressed participants than in matched controls (odds ratio, 1.46). Thus, the association was real at the population level, while most participants with depression did not meet that particular CRP cutoff. Osimo et al. (2019).
A later large sex-stratified meta-analysis evaluated 124 studies including 423,421 participants. It found modest cross-sectional associations between depressive symptoms or depression and inflammatory markers in women and men. The size of the pooled standardized differences was small, and relationships differed by marker and study design. Across longitudinal studies, elevated IL-6 predicted subsequent depression among women and elevated CRP predicted it among men in the analyzed data. These are aggregate observations with methodological limitations; sex-stratified averages are not deterministic clinical rules for any person. Szabo et al. (2026).
A group difference can be statistically reliable while offering limited ability to classify individual people. Healthy people may have raised CRP for reasons unrelated to mood. A person meeting criteria for major depressive disorder may have CRP within the reference range. Biomarker distributions overlap across groups, and assays are affected by many biological and contextual variables. This overlap is central to interpreting findings from case-control and population studies. Osimo et al. (2019); Szabo et al. (2026).
2. The type of depression outcome matters
Some studies investigate current major depressive disorder confirmed through an interview. Others examine depressed mood, questionnaire symptom scores, lifetime diagnoses, or combinations of these endpoints. Results from these groups cannot automatically be treated as equivalent. A two-week depressive episode, chronic residual symptoms, and a single day of low mood are different clinical objects. Studies may also differ in medication exposure, medical illness, and severity, making apparently contradictory results less surprising. NIMH; Szabo et al. (2026).
3. Obesity, sleep, illness, and medications complicate interpretation
Inflammation is associated with chronic physical disease and metabolic conditions that can also increase depression risk or accompany depressive symptoms. Smoking, activity level, adiposity, recent infection, sleep disruption, medication exposure, and socioeconomic stress may affect measurements or act along shared pathways. Statistical adjustment reduces some confounding but depends on what researchers measured and how accurately they measured it. NIMH, Chronic Illness and Mental Health; Osimo et al. (2019).
These observations make integrated physical and mental health care valuable. They also explain why a study's association does not translate directly into a claim that inflammation caused one individual's episode of depression.
III. Can Inflammation Contribute to Depressive Symptoms?
1. Immune-to-brain communication is biologically plausible
Immune activity in the body can influence the central nervous system through interacting neural, vascular, hormonal, cellular, and molecular routes. Researchers examine cytokine signaling, altered neurotransmitter metabolism, stress-system regulation, and effects on the circuits involved in reward, motivation, threat detection, and motor activity. The relevant mechanisms are complex and distributed, rather than a single inflammatory molecule crossing a barrier and producing a diagnosis. Miller and Raison (2016).
Experimental work and observations in people receiving certain immune-activating treatments show that increasing inflammatory signaling can coincide with fatigue, reduced pleasure, social withdrawal, slowed activity, or depressive symptoms. These findings provide mechanistic support for immune contributions to mood-related behavior. They do not establish that every instance of major depression shares the same immune pathway. Miller and Raison (2016).
2. Sickness behavior and clinical depression overlap in some symptoms
During illness, tiredness, decreased activity, reduced appetite, changes in sleep, and less interest in social interaction can occur as part of an immune-related behavioral response. Depression can involve some of these features, including anhedonia and psychomotor slowing. The overlap helps researchers generate hypotheses about the biological origins of particular symptoms, but similar symptoms can arise through many mechanisms. Persistent depressed mood, hopelessness, guilt, and functional impairment require clinical interpretation in their full context. Miller and Raison (2016); NIMH.
This distinction also matters when a person has an autoimmune disorder, persistent pain, or another chronic condition. Fatigue may reflect active disease, medication effects, disrupted sleep, depression, or several processes occurring together. A thoughtful assessment addresses both the physical condition and the psychiatric symptoms. NIMH, Chronic Illness and Mental Health.
3. Inflammatory markers are not direct measurements of brain inflammation
CRP is typically measured in peripheral blood. Blood concentrations of cytokines and other proteins do not directly quantify microglial activation or establish a neuroinflammatory process inside the brain. The relationships between peripheral immune signals, central nervous system responses, and subjective symptoms are active research questions. Imaging, cerebrospinal fluid studies, and experimental models investigate different pieces of that puzzle. The more detailed neuroscience belongs within the Hub's depression and brain research overview. Miller and Raison (2016).
IV. What Does the Evidence Say About Causation?
1. Cross-sectional associations cannot establish direction
When depression and inflammatory markers are measured at the same time, the result does not establish which occurred first. Inflammation could contribute to certain depressive symptoms; depressive illness could change behavior and physiology in ways that affect inflammation; both could respond to common underlying circumstances. More than one pathway may operate simultaneously. Szabo et al. (2026); Miller and Raison (2016).
Prospective observational studies are stronger for establishing temporal sequence because markers are assessed before later symptoms. Even so, a prospective association may reflect residual confounding or other unmeasured mechanisms. The 2026 meta-analysis identified particular marker-specific longitudinal associations, which strengthens the case for further causal testing without determining an individual's causal pathway. Szabo et al. (2026).
2. Genetic studies suggest pathways worth investigating
Mendelian randomization uses genetic variants associated with biological exposures as instruments to test hypotheses about causal effects. In a study using UK Biobank and genome-wide association data, Kelly and colleagues reported results consistent with a possible causal contribution of IL-6 receptor-related signaling to depression. The interpretation depended on how classical and trans-signaling pathways were represented. Such designs require assumptions about genetic instruments and about other routes through which variants may affect the outcome. Kelly et al. (2021).
A systematic review of Mendelian-randomization studies found emerging causal evidence for multiple categories of exposures, including inflammatory biomarkers, but also substantial heterogeneity and methodological weaknesses. Genetically proxied IL-6 signaling, measured IL-6 concentration, and a routine CRP result are not interchangeable exposures. Genetic evidence makes certain immune pathways scientifically credible candidates; it does not confirm a general proposition that inflammation causes all depressive disorders. Ma et al. (2023); Kelly et al. (2021).
3. Treatment experiments are important causal tests
If modifying an inflammatory pathway improves depressive symptoms in a controlled trial, that supports a causal contribution under the study's specific conditions. The strength of this inference depends on the target, comparison treatment, participants, measured outcomes, study size, adherence, adverse effects, and risk of bias. A short-term change in a symptom score is distinct from sustained remission and restored functioning. Mac Giollabhui et al. (2026).
V. Do People With Higher Inflammation Have a Different Symptom Pattern?
Researchers have investigated whether elevated inflammation is associated especially with low motivation, fatigue, reduced pleasure, disturbed sleep, and psychomotor changes. The hypothesis is biologically coherent because immune signaling can influence reward-related and energy-regulation circuits. Evidence is still insufficient to assign a stable inflammatory subtype from a checklist of symptoms, because those symptoms are widespread across depression, physical illness, and other disorders. Miller and Raison (2016); Mac Giollabhui et al. (2026).
Anhedonia means diminished interest or pleasure. It is a core depressive symptom and a useful outcome in studies of anti-inflammatory interventions. Fatigue, by contrast, can originate from sleep disruption, pain, endocrine illness, other medical conditions, medication effects, and mood disturbance. Neither symptom identifies an inflammatory mechanism on its own. See depression and fatigue and the Hub's depression symptoms reference. NIMH.
The established specifiers for depressive episodes—such as melancholic, atypical, psychotic, or seasonal patterns—are clinical classifications tied to symptom configurations and course. An inflammatory blood-marker result does not define any of these specifiers, and current evidence does not justify equating a clinical specifier with an immune biomarker category. NIMH; NICE NG222.
VI. Are CRP and Cytokine Tests Useful for Diagnosing Depression?
1. A CRP result has multiple possible explanations
A raised CRP can accompany infection, inflammatory disorders, tissue injury, and other medical or physiological circumstances. Interpreting the number requires symptoms, history, examination, and the purpose for which a clinician ordered the test. A threshold used to recruit research participants should not be adopted as a personal diagnostic cutoff. For example, the CRP thresholds of 2 mg/L and 3 mg/L encountered in depression studies define particular study groups; neither provides a validated test for major depressive disorder. Osimo et al. (2019); Mac Giollabhui et al. (2026).
2. Depression is diagnosed clinically
A depression evaluation involves symptom duration and severity, functional effects, medication and substance history, medical conditions, other mental disorders, and safety when indicated. A clinician may order laboratory tests when the presentation suggests a relevant medical cause or comorbidity. Those tests investigate the medical question; they do not replace assessment of mood, anhedonia, thinking, sleep, energy, and functioning. NIMH; NICE NG222.
3. Treatment prediction is an active research area
CRP and other biomarkers have been proposed as ways to identify participants who may benefit more from particular anti-inflammatory treatments. A candidate predictive marker must be shown to improve real clinical decisions across independent, adequately powered trials, ideally with reproducible thresholds and favorable benefit–harm balance. At present, routine CRP-guided selection of antidepressant or anti-inflammatory treatment is not an established standard of depression care in the cited guidance. NICE NG222; Mac Giollabhui et al. (2026).
VII. Could Anti-Inflammatory Medications Treat Depression?
1. Early clinical trials motivated further research
One influential randomized trial enrolled 60 people with treatment-resistant depression and compared the TNF antagonist infliximab with placebo. The trial found no overall superiority of infliximab. An exploratory relationship between baseline hs-CRP and symptom improvement suggested a possible benefit in some participants with higher inflammation, but the subgroup was small and its response-rate comparison did not reach statistical significance. The study provided a research hypothesis, not an indication for prescribing infliximab to people with depression. Raison et al. (2013).
2. Newer meta-analyses show promise alongside uncertainty
A systematic review and meta-analysis published online in December 2025, appearing in the January 2026 issue of the American Journal of Psychiatry, pooled randomized anti-inflammatory-treatment studies in depressed participants with elevated inflammatory markers. In 11 trials using a CRP threshold of at least 2 mg/L, pooled changes favored treatment for anhedonia (Hedges' g 0.40) and depressive symptom severity (g 0.35). The investigators did not detect statistically significant differences in treatment response or remission. This distinction is essential: reducing average symptom severity does not establish that a treatment reliably produces remission. Mac Giollabhui et al. (2026).
A separate 2026 review examined celecoxib as an add-on to antidepressants. It included seven placebo-controlled trials totaling 365 participants and estimated an improvement in depressive symptom scores, but the trials were small and heterogeneous, and the certainty of evidence for efficacy was rated very low. The authors concluded that antidepressant efficacy remained highly uncertain and that celecoxib adjunctive use was not recommended on this evidence. Srisurapanont et al. (2026).
These findings are compatible: a promising pooled average effect can coexist with inadequate evidence for routine clinical implementation. Investigational classes differ substantially, and the benefit–harm profile of an anti-cytokine biologic is not equivalent to that of a nonsteroidal anti-inflammatory drug.
3. Safety makes self-treatment a poor substitute for clinical care
Celecoxib and other nonsteroidal anti-inflammatory drugs can cause serious gastrointestinal bleeding, ulcers, kidney-related complications, and cardiovascular events in susceptible circumstances. The current U.S. prescribing information for celecoxib includes boxed warnings for cardiovascular thrombotic events and gastrointestinal bleeding, ulceration, and perforation. Combining medications can introduce further interactions and risks; medical review is required. DailyMed, Celecoxib Prescribing Information.
Targeted immunosuppressive or anti-cytokine therapies also require disease-specific indications, screening, monitoring, and careful consideration of infection and other safety risks. Their experimental use in depression should be limited to appropriately supervised research or exceptional specialist judgment within applicable clinical standards. The trial evidence does not support buying or taking anti-inflammatory medication to treat mood symptoms without medical guidance. Raison et al. (2013); NICE NG222.
VIII. What About Antidepressants, Psychotherapy, Exercise, and Diet?
Standard depression treatments remain grounded in clinical trials and guidelines evaluating mood, functioning, symptom burden, patient preferences, and harms. Evidence-based psychotherapies, antidepressant medication, and selected additional interventions can be appropriate depending on severity and circumstances. A hypothesis about inflammation does not make conventional depression care biologically irrelevant. For a treatment overview, see Depression Treatment: Therapy, Medication, and Other Options. NICE NG222; NIMH.
Researchers have explored whether antidepressant treatment changes inflammatory markers and whether baseline inflammation predicts treatment response. Many studies are observational, use small samples, or differ substantially in drug class, timing, and outcome. At present, the existence of an inflammatory correlate is not sufficient reason to choose, stop, or switch an antidepressant on one's own. A prescriber should review treatment effects, side effects, adherence, interactions, and the full diagnosis, including any history of hypomania or mania that may alter the treatment approach. NICE NG222; NIMH.
Regular physical activity, support for sleep, balanced nutrition, and management of coexisting medical conditions can be worthwhile parts of a comprehensive care plan. Their possible effects on inflammatory physiology are an additional research interest, rather than proof that lowering a specific biomarker will cure depression. Activity and other self-management plans should fit the person's health, disability, energy, preferences, and existing care. See Depression Self-Help: Strategies Supported by Research. NIMH, Chronic Illness and Mental Health; NICE NG222.
Commercial claims for “anti-inflammatory depression diets,” supplement combinations, detox regimens, and at-home inflammation panels should be assessed against the specific human trial evidence supporting each intervention. Dietary claims about systemic inflammation do not automatically demonstrate treatment efficacy for a clinically diagnosed depressive disorder. Prescription treatments and supplements can also interact. The appropriate standard is evidence for the intervention and outcome, not the biological attractiveness of a marketing explanation.
IX. Depression, Chronic Illness, and the Direction of Care
Depression can coexist with cardiovascular disease, diabetes, chronic pain, autoimmune conditions, and other long-term illnesses. Physical symptoms, treatment burdens, stress, shared exposures, and physiological changes can combine in complicated ways. An inflammatory association may help explain some of that overlap, but it does not replace an assessment of the medical condition, disability, and psychosocial situation. NIMH, Chronic Illness and Mental Health.
When an individual develops mood symptoms during an inflammatory illness or after beginning a medication, clinicians consider timing, plausible medication effects, changes in sleep, pain or fatigue, and the full clinical picture. Depression may need its own treatment even when a physical disease is being managed. Conversely, treating depression does not eliminate the need to investigate persistent fever, unexplained weight loss, severe pain, or other signs that merit medical attention.
A useful care pathway asks two questions in parallel: what psychiatric syndrome, if any, is present, and what physical health needs require independent attention? This approach avoids allowing a psychiatric label to obscure medical problems or a medical diagnosis to obscure treatable depression. NIMH; NIMH, Chronic Illness and Mental Health.
X. What the Evidence Can and Cannot Establish
Evidence strongly supports an average association between depressive states and selected peripheral inflammatory biomarkers. It supports credible immune-to-brain pathways, including effects on energy, reward-related behavior, and neural signaling. Genetic and experimental results strengthen hypotheses that some inflammatory pathways can contribute causally to depressive symptoms in certain circumstances. Osimo et al. (2019); Szabo et al. (2026); Miller and Raison (2016).
Current evidence remains limited for deciding from an individual's CRP or cytokine result whether they have depression, identifying a universally valid immune-based depression category, or selecting a specific treatment from a routine blood test. Trials of anti-inflammatory treatments offer meaningful preliminary signals, especially in biomarker-enriched research samples, while differences in remission, long-term outcomes, safety, replication, and cost-effectiveness remain decisive open questions. Mac Giollabhui et al. (2026); Srisurapanont et al. (2026).
The strongest future study designs will prespecify biomarker thresholds, recruit adequately powered samples, confirm depressive diagnoses and bipolar differentials, assess relevant symptoms separately, compare clinically meaningful outcomes, measure safety, and replicate treatment-by-biomarker interactions. A finding should influence clinical practice when it improves patient care reliably, not merely when it produces a statistically notable subgroup contrast.
XI. Practical Questions Patients Can Bring to a Clinician
When depressive symptoms persist, a practical conversation can cover their duration and impact, accompanying changes in sleep or energy, relevant medical illnesses and medications, previous treatment responses, and any symptoms suggesting mania or hypomania. A clinician can decide whether targeted medical tests are appropriate, explain what an existing CRP result means in context, and choose evidence-based depression care. A positive screening questionnaire is a starting point for an assessment, as explained in Depression Screening: What a Positive Questionnaire Result Means. NIMH.
If a person has thoughts of suicide, new psychotic symptoms, or an acute crisis, timely assessment and safety support take priority over investigating inflammatory markers. In the United States, calling or texting 988 connects people to the Suicide & Crisis Lifeline, and immediate life-threatening emergencies require emergency services. The inflammatory research discussed here is not a substitute for urgent mental health care. NIMH.
XII. Frequently Asked Questions
Does inflammation cause depression?
Immune activation can contribute to mood-related and motivational symptoms under some experimental circumstances. Observational, genetic, and treatment studies suggest a causal role for particular pathways in some presentations. The overall relationship is heterogeneous and potentially bidirectional or confounded, so an individual episode cannot be assigned an inflammatory cause from a blood test or symptom list. Miller and Raison (2016); Kelly et al. (2021).
Can depression itself increase inflammation?
Depression is associated with physiological and behavioral changes that could affect immune activity, but longitudinal findings are not uniform. The 2026 sex-stratified meta-analysis observed certain inflammatory markers predicting later depression, while it did not find depression predicting subsequent inflammation in its pooled longitudinal analyses. That result addresses the analyzed cohorts and methods, not every possible biological feedback loop. Szabo et al. (2026).
Is an elevated CRP a sign that I have “inflammatory depression”?
An elevated CRP means that the clinical context of an inflammatory marker needs interpretation. It does not diagnose depression or identify a validated inflammatory subtype. CRP may be raised for multiple medical reasons, and depression is assessed on clinical grounds. Osimo et al. (2019); NIMH.
Should everyone with depression have CRP, IL-6, or TNF-α measured?
Routine inflammatory-marker testing is not established as a standard method of diagnosing depression or choosing its treatment. Laboratory testing is appropriate when an individual medical indication is present, and inflammatory measurements also have a role in research. NICE NG222; NIMH.
Do anti-inflammatory drugs work as antidepressants?
Some randomized-trial meta-analyses find average symptom improvements in selected groups, including people recruited or stratified for elevated inflammatory markers. Current evidence has limitations in sample size, heterogeneity, bias, remission outcomes, and safety. Anti-inflammatory drugs are not a general substitute for established depression care. Mac Giollabhui et al. (2026); Srisurapanont et al. (2026).
Can depression be treated successfully when inflammation is not elevated?
Yes. Depression is diagnosed and treated clinically, and many people with depression do not meet research thresholds for low-grade inflammation. Evidence-based care does not require proof of an inflammatory biomarker elevation. Osimo et al. (2019); NICE NG222.
Is brain inflammation the same as an elevated blood CRP?
No. CRP is a peripheral marker that may reflect aspects of systemic inflammatory activity. It is not a direct measure of inflammatory processes in brain tissue. Neuroscience studies examine multiple potential connections between peripheral immune pathways and central nervous system function. Miller and Raison (2016).
Could inflammation explain antidepressant resistance?
Some studies report associations between inflammatory activity and poorer treatment response, and this possibility motivates biomarker-stratified trials. Neither elevated inflammation nor inadequate response establishes treatment-resistant depression on its own. Clinicians first reassess diagnosis, adequacy of previous treatment, medical factors, comorbidity, preferences, and safety. Mac Giollabhui et al. (2026); NICE NG222.
Conclusion
Inflammation is a scientifically credible part of the biological story of depression for some people and some symptoms. Its contribution is supported by reproducible population associations, plausible biological mechanisms, genetic analyses, and increasingly informative clinical trials. The critical next step is reliable demonstration of who benefits from which intervention, with clinically meaningful outcomes and acceptable risks. For now, depression assessment and treatment remain grounded in the full clinical picture, while inflammatory biomarkers remain primarily research tools or tests ordered for independent medical indications. Mac Giollabhui et al. (2026); NICE NG222.
Related Articles
References
DailyMed. (2026). Celecoxib capsules: U.S. prescribing information. Official drug label.
Kelly, K. M., Smith, J. A., & Mezuk, B. (2021). Depression and interleukin-6 signaling: A Mendelian randomization study. Brain, Behavior, and Immunity. https://doi.org/10.1016/j.bbi.2021.02.019.
Ma, W.-R., Zhang, L.-L., Ma, J.-Y., Yu, F., Hou, Y.-Q., Feng, X.-R., & Yang, L. (2023). Mendelian randomization studies of depression: Evidence, opportunities, and challenges. Annals of General Psychiatry, 22, Article 47. https://doi.org/10.1186/s12991-023-00479-6
Mac Giollabhui, N., Madison, A. A., Lydston, M., et al. (2026). Effect of anti-inflammatory treatment on depressive symptom severity and anhedonia in depressed individuals with elevated inflammation: Systematic review and meta-analysis of randomized controlled trials. American Journal of Psychiatry, 183(1), 70–79. https://doi.org/10.1176/appi.ajp.20241115.
Miller, A. H., & Raison, C. L. (2016). The role of inflammation in depression: From evolutionary imperative to modern treatment target. Nature Reviews Immunology, 16, 22–34. https://doi.org/10.1038/nri.2015.5.
National Institute for Health and Care Excellence. (2022; reviewed 2026). Depression in adults: Treatment and management (NG222). Official recommendations.
National Institute of Mental Health. (n.d.). Depression. Official clinical information.
National Institute of Mental Health. (n.d.). Understanding the link between chronic disease and depression. Official information.
Osimo, E. F., Baxter, L. J., Lewis, G., Jones, P. B., & Khandaker, G. M. (2019). Prevalence of low-grade inflammation in depression: A systematic review and meta-analysis of CRP levels. Psychological Medicine, 49, 1958–1970. https://doi.org/10.1017/S0033291719001454.
Raison, C. L., Rutherford, R. E., Woolwine, B. J., et al. (2013). A randomized controlled trial of the tumor necrosis factor antagonist infliximab for treatment-resistant depression: The role of baseline inflammatory biomarkers. JAMA Psychiatry, 70(1), 31–41. https://doi.org/10.1001/2013.jamapsychiatry.4.
Srisurapanont, M., Suradom, C., & Suttajit, S. (2026). Celecoxib adjunct to antidepressants for major depressive disorder: Systematic review and meta-analysis. Brain, Behavior, & Immunity – Health, 55, 101277. https://doi.org/10.1016/j.bbih.2026.101277.
Szabo, Y. Z., Nishimi, K., Smirnova, M., et al. (2026). Sex-specific associations of depression with inflammatory markers: A systematic review and meta-analysis. Translational Psychiatry, 16, 394. https://doi.org/10.1038/s41398-026-04139-7.