CRP and IL-8 Predict Antidepressant Non-Response — A Meta-Analysis With Zero Heterogeneity

The inflammatory hypothesis of depression proposes that a subgroup of MDD patients is maintained by immune dysregulation rather than classic monoamine deficiency — and that this subgroup responds poorly to standard antidepressants. The research base for this hypothesis has been growing for over a decade, but remains fragmented across small studies with inconsistent findings.

This meta-analysis adds a meaningful contribution: synthesizing the CRP evidence across seven studies (N=1,493) and finding a statistically significant association (p=0.0006) with no heterogeneity whatsoever (I²=0%). In the typically heterogeneous field of biomarker psychiatry, zero heterogeneity is unusual and suggests the finding is robust to study-level variation in design, population, and measurement protocol.

What the Numbers Mean

The effect size for CRP (ES=-0.18) is small by conventional benchmarks. This is not a case where CRP alone predicts treatment outcome with diagnostic precision. It is a case where CRP is a consistent, statistically reliable signal — one that adds information to treatment planning even when its individual-level predictive value is modest.

The clinical value of a small biomarker effect depends on the decision context. For a binary antidepressant choice (start SSRI vs. explore anti-inflammatory augmentation), a consistent CRP signal is actionable even at ES=0.18. The question is not "will this patient fail an SSRI?" but "is this patient more likely to fail an SSRI than the average MDD patient, and does that shift the risk/benefit calculus toward alternatives?"

IL-8 adds specificity: it is a chemokine produced by monocytes, lymphocytes, and microglial cells that drives neutrophil recruitment and is implicated in neuroinflammatory processes. Its predictive role alongside CRP suggests the relevant biological process is not general inflammation (which would elevate a broader cytokine panel) but a more specific pathway involving innate immune activation and microglial signaling.

The Null Findings Are Also Informative

The fact that IL-1β, IL-2, IL-4, IL-10, TNF-α, and GM-CSF — all widely studied in inflammation-depression research — did not show consistent meta-analytic associations with treatment response is important. The inflammatory biomarker literature often presents a confusing picture where almost every cytokine has been reported to predict something. This meta-analysis narrows the signal: among the commonly measured panel, CRP and IL-8 are the consistent predictors, not inflammation in general.

This specificity has practical implications for clinicians considering inflammatory workup in TRD: a full cytokine panel is expensive, requires specialized laboratories, and adds clinical complexity. If CRP — a routine, cheap, widely available point-of-care test — captures the treatment-relevant signal, it is the appropriate first-step measurement.

Practical Translation

For non-prescribing clinicians, the pathway from this research to practice runs through collaborative care:

1. Case formulation: patients with predominantly anhedonic, fatigue-dominant, psychomotor-slowed depression — consistent with the "sickness behavior" phenotype associated with inflammatory subtype — are candidates for CRP measurement in the workup
2. Advocacy at referral: when referring a TRD patient for psychiatric consultation, flagging the inflammatory pattern in your referral letter and noting that CRP has not been measured increases the probability that the receiving clinician will include it in their assessment
3. Lifestyle integration: exercise is the most consistently anti-inflammatory behavioral intervention. In patients where the inflammatory hypothesis may apply, framing exercise as biologically targeted — not just a mood-adjacent adjunct — is clinically accurate and may improve adherence