Peripheral blood CADM1 expression and multi-gene signatures in depressive disorders: a pilot case-control study from...

Depressive disorders are a leading cause of global disability, yet gene expression studies in Middle Eastern populations remain scarce. This study investigated peripheral blood expression of seven candidate genes, previously identified in animal models and early-onset depression, in adults with and without depressive disorders in United Arab Emirates (UAE).

In a case-control design, 49 adults with depressive disorders (ICD-10 major depressive disorder, dysthymia, or adjustment disorder with depressed features) and 49 non-depressed controls were assessed with the Beck Depression Inventory-II and Patient Health Questionnaire-9. Peripheral blood expression of ADCY3, DGKA, FAM46A, CADM1, KIAA1539, MARCKS, and RAPH1 was measured by quantitative real-time PCR, normalised to GAPDH. Analyses included Mann-Whitney U tests, principal component analysis, fold-change estimation, pre-specified receiver operating characteristic analysis, and severity-restricted, age-balanced, and RNA-quality sensitivity analyses.

CADM1 showed modest upregulation in the depressive disorder group (1.30-fold; Mann-Whitney p = 0.049, uncorrected) that did not survive multiple testing correction (Benjamini-Hochberg p = 0.178). Principal component analysis revealed an unadjusted PC2 difference (p = 0.006), driven by CADM1 and RAPH1, that attenuated after age adjustment (MANCOVA p = 0.156). CADM1 correlated negatively with depression severity (BDI-II ρ = -0.280, p = 0.005), and the signal strengthened monotonically across severity strata (moderate-severe BDI-II ≥ 20, p = 0.029; severe BDI-II ≥ 29, p = 0.014), arguing against diagnostic heterogeneity as the primary driver. However, propensity score matching on age abolished the CADM1 effect (p = 0.965). Exploratory subgroup analyses suggested sex-specific patterns (CADM1 in males, p = 0.039; KIAA1539 in females, p = 0.010). The CADM1 area under the receiver operating characteristic curve was 0.616, indicating no diagnostic utility. Post-hoc power for individual genes was 5-19%.

This pilot provides the first peripheral blood gene expression data for depression in a UAE population. Given the small sample, the 13-year mean age difference between groups, and unmeasured confounders, CADM1 is best framed as a hypothesis-generating candidate only; the signal was unstable after age balancing and the discrimination performance (AUC = 0.616) is below any clinically useful threshold. Validation in larger, strictly age-matched studies is required before any clinical interpretation.