Elevated cortisol/DHEA ratio mediates the association between symptom severity and working memory impairment in drug-naive, first-episode OCD

Abstract

Objective: Dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis has been implicated in the pathophysiology of obsessive-compulsive disorder (OCD) and its associated cognitive deficits. However, the role of the cortisol/dehydroepiandrosterone (DHEA) ratio, a functional index of the catabolic/anabolic balance, remains poorly characterized in the early stages of the illness. This study aimed to investigate the association between this neuroendocrine marker and executive function in a rigorously defined cohort of drug-naive, first-episode OCD patients.

Methods: Seventy-five drug-naive, first-episode patients with OCD and 50 age- and sex-matched healthy controls (HCs) were recruited. Plasma concentrations of cortisol and DHEA were quantified by ELISA. Executive functions were assessed using a comprehensive battery, including response inhibition (Stop-Signal Task, SST), working memory (2-back task), and cognitive flexibility (Wisconsin Card Sorting Test, WCST). Data were analyzed using group comparisons, partial correlations, and stepwise multiple regression with Benjamini-Hochberg False Discovery Rate (FDR) correction. Mediation modeling was employed to explore mechanistic pathways.

Results: Compared to HCs, the OCD group exhibited significant deficits in response inhibition and working memory, accompanied by elevated cortisol (P = 0.006), reduced DHEA (P < 0.001), and a markedly higher cortisol/DHEA ratio (P < 0.001). The elevated ratio was robustly correlated with greater symptom severity (Y-BOCS) and poorer inhibitory control (longer SSRT) after FDR correction (P_adj < 0.05). Multiple regression analysis identified the cortisol/DHEA ratio as a significant independent predictor of response inhibition deficits (P_adj = 0.002), whereas DHEA levels specifically predicted psychomotor slowing in working memory (P_adj < 0.001). Critically, mediation analysis revealed that the cortisol/DHEA ratio showed a significant indirect effect on the relationship between symptom severity and working memory accuracy (95% CI: -0.0692 to -0.0096), suggesting a "suppression" effect where neuroendocrine dysregulation serves as a distinct biological pathway for cognitive impairment. No significant hormonal associations were observed for cognitive flexibility (P > 0.05).

Conclusion: This study provides novel evidence that an elevated cortisol/DHEA ratio is not only a hallmark of drug-naive OCD but also independently predicts inhibitory dysfunction. Furthermore, this ratio appears to mediate the impact of clinical symptomatology on working memory capacity. These findings highlight the potential utility of the cortisol/DHEA ratio as a state-dependent biomarker and suggest that restoring the neuroendocrine balance could offer a therapeutic avenue for ameliorating cognitive deficits in OCD.

Keywords: cortisol/DHEA ratio; executive function; mediation analysis; obsessive-compulsive disorder; response inhibition; working memory.

Figure 1

Pearson correlation matrix heatmap of hormonal biomarkers, OCD symptom severity, and cognitive performance. Pearson correlation matrix of hormonal biomarkers, OCD symptom severity, and cognitive performance measures. Analysis revealed that cortisol levels and the cortisol/DHEA ratio were positively associated with OCD symptom severity (Y-BOCS) and stop-signal reaction time (SSRT), indicating that a higher neuroendocrine stress load correlates with greater symptom burden and poorer inhibitory control. Conversely, these markers were inversely correlated with response inhibition accuracy (SST) and working memory accuracy (2-back). DHEA levels generally demonstrated an opposing, protective profile, correlating negatively with symptom severity and positively with cognitive accuracy. Consistent with the group comparisons, no significant associations were observed between hormonal measures and Wisconsin Card Sorting Test (WCST) performance. Values represent Pearson correlation coefficients (r). To control for Type I error inflation, all P-values were adjusted using the Benjamini-Hochberg False Discovery Rate (FDR) method. **P_adj < 0.01,*P_adj < 0.05. OCD, obsessive-compulsive disorder; HC, healthy controls; Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; HAMA, Hamilton Anxiety Scale; HAMD, Hamilton Depression Scale; SST, Stop Signal Task; SSRT, Stop Signal Reaction Time; RT, reaction time on 2-back task; CC, correct count on 2-back task; WCST, Wisconsin Card Sorting Test; Re, number of error responses; Rpe, number of persistent errors.

Figure 2

Path Coefficients for the Cortisol/DHEA Ratio in the Association Between OCD Severity and Working Memory Performance. Statistical Mediation Model. Path diagram illustrating the mediating role of the Cortisol/DHEA ratio in the relationship between OCD symptom severity (Y-BOCS) and Working Memory (2-Back Correct Count). Values on paths represent unstandardized regression coefficients (B) with standard errors (SE) in parentheses. The solid lines indicate significant pathways (*P < 0.05), while the dashed line indicates a non-significant direct effect (c’). The box at the top reports the significant indirect effect (a × b) with its bias-corrected 95% bootstrap confidence interval. *P < 0.05. Y-BOCS, Yale-Brown Obsessive-Compulsive Scale; WM, Working Memory; Cortisol/DHEA, cortisol-to-dehydroepiandrosterone ratio.

Figure 3

Hypothesized conceptual model illustrating the pathophysiological link between OCD symptomatology, HPA-axis dysregulation, and executive dysfunction. The schematic proposes that the chronic distress and allostatic load associated with OCD symptom severity drive a maladaptive shift in the HPA axis, resulting in an elevated Cortisol/DHEA ratio. This neuroendocrine imbalance—characterized by excessive glucocorticoid activity relative to neuroprotective DHEA—is hypothesized to exert neurotoxic effects on glucocorticoid-sensitive brain regions, particularly within frontostriatal circuits (e.g., Dorsolateral Prefrontal Cortex). The resulting compromise in neural integrity ultimately manifests as broad executive dysfunction, encompassing specific deficits in both working memory and response inhibition.