Vreijling · The British journal of psychiatry : the journal of mental science 2024 · Pooled individual participant data meta-analysis of clinical trials · n=2551

Features of immunometabolic depression as predictors of antidepressant treatment outcomes: pooled analysis of four clinical trials.

Cited 32 times in the scientific literature.

Level 1 - systematic review of randomized trials

Pooled individual participant data meta-analysis of four clinical trials

PubMed 38130122 · doi:10.1192/bjp.2023.148 · record verified 2026-08-26

What was done

Analyzed individual-level data from 2,551 participants with depression across four clinical trials (iSPOT-D, n = 967; CO-MED, n = 665; GENDEP, n = 773; EMBARC, n = 146). Evaluated baseline atypical energy-related symptoms (AES), body mass index (BMI), and C-reactive protein (CRP, available in three trials), individually and as an aggregated immunometabolic depression (IMD) index, as predictors of antidepressant treatment outcomes. Mixed models evaluated the primary outcome (change in depressive symptom severity) and logistic regressions evaluated secondary outcomes (response and remission), with trial-level estimates combined via random-effects meta-analyses.

What was found

AES severity and BMI did not predict change in depressive symptoms. Higher baseline CRP predicted smaller depressive symptom reductions (n = 376, pooled beta = 0.06, 95% CI 0.0001 to 0.12, P = 0.049, I² = 3.61%). The aggregated IMD index similarly predicted smaller symptom reductions (n = 372, pooled beta = 0.12, SE = 0.12, 95% CI 0.01 to 0.22, P = 0.031, I² = 23.91%). In the subgroup receiving selective serotonin reuptake inhibitors, effect sizes were larger for CRP (pooled beta = 0.16) and the IMD index (pooled beta = 0.20). Baseline IMD features, evaluated separately or combined, did not predict categorical response or remission.

Why it matters

Although immunometabolic markers correlate statistically with slightly attenuated symptom reduction during standard antidepressant therapy, the small effect sizes and lack of association with response or remission indicate limited current utility for clinical treatment stratification.

Limits

CRP was measured in only three trials, reducing the sample size for primary biomarker analyses from 2,551 to under 380 participants. The effect sizes were very small, with the lower confidence bound for CRP bordering zero (0.0001). Categorical clinical outcomes (response and remission) were not predicted, and whether patients with IMD features benefit from alternative targeted treatments was not tested.

Cited by