Abstract
OBJECTIVE: The relationship between cortisol and perceived stress is poorly understood. We sought to determine whether perceived stress is associated with cortisol biomarkers in adolescents. METHODS: We examined 229 adolescents (mean age = 15.8 years). We measured perceived stress with the 14-item Perceived Stress Scale (PSS) questionnaire, serum cortisol (sCOR), salivary Cortisol Awakening Response (CAR: 30-minute post-awakening - awakening), salivary Diurnal Cortisol Slope (DCS: evening-awakening), and total daily salivary cortisol (TDC). We used multivariable linear regression to estimate baseline associations between PSS, TDC, sCOR, and FBG. We used mixed effects linear regression to estimate baseline associations between PSS and CAR and DCS. We tested twelve-week longitudinal associations between PSS and cortisol biomarkers using random effects regression. Analyses were adjusted for age, gender, and BMI. RESULTS: There were statistically significant associations between PSS and TDC (beta= -104.36 ± 34.3; p = 0.002) at baseline and between PSS and DCS at 12 weeks (beta= -0.058 ± 0.02; p = 0.01), but no association between PSS and sCOR or CAR (p > 0.26 for all) at baseline or 12-weeks. There were no associations between change in PSS and change in cortisol biomarkers longitudinally. CONCLUSION: In adolescents, perceived stress measured by the PSS was inconsistently associated with TDC and DCS, and consistently unassociated with CAR and sCOR. Studies reporting on PSS outcomes should exercise caution when making conclusions about cortisol biomarkers. There's a need for an instrument that captures a global measure of perceived stress and is sensitive to HPA functioning.