Diurnal Cortisol Covaries with Observer-Rated but Not Self-Reported Professional Relationship Quality: A Four-Wave Within-Person Study.

Stress impairs prefrontal-dependent behavioral control and shifts processing toward automatic defaults, but whether this extends to the quality of real-world professional relationships, where skilled interpersonal engagement is itself a predictor of client outcomes, is unknown. In a four-wave longitudinal study of 46 community supervision officers, we tested whether diurnal cortisol predicted relationship quality assessed by independent observers from 2,126 video-recorded sessions (732 supervisees) and by officers themselves via survey (432 supervisees). Two-level models with latent mean centering identified two within-person effects implicating different cortisol indices and different relational dimensions: higher total cortisol output (log AUCg) predicted lower observer-rated Caring/Fairness, and flatter diurnal slope predicted lower observer-rated Trust. These effects were independent, robust across sensitivity specifications, and specific, with neither index predicting the other's target. The slope effect reflected the trajectory of diurnal decline rather than evening or waking level alone. No cortisol index predicted officer-reported relationship quality, consistent with both the controlled-to-automatic framework and the trait-like variance structure of officer self-report (63-73% between-person). The findings suggest that HPA dysregulation covaries with what independent observers see in professional relational behavior but not with what the professional perceives. That distinct cortisol indices reached distinct relational dimensions is consistent with separable biological pathways rather than a general stress-to-relationship process.