The potential association between baseline cortisol levels and type 2 diabetes in Tunisian Women? a case-control study and meta-analysis across blood, saliva, and urine
摘要
Purpose: Several studies have suggested that cortisol could be involved in type 2 diabetes (T2D) across different populations. However, this association has not been investigated in the Tunisian population. Our study assessed whether there is a link between T2D and morning blood cortisol levels in diabetic women from Tunisia. Subsequently, a comprehensive meta-analysis was carried out across blood, saliva, and urine samples. Methods: Thirty-three women from southern Tunisia were enrolled in this study. Multivariate logistic regression and receiver operating characteristic (ROC) curve analysis were performed to assess the cortisol-T2D association using R software. In addition, a meta-analysis was conducted using RevMan5.4 software. Besides, a comprehensive search of electronic databases (Google Scholar, PubMed, Scopus, and Circort) was performed. Seventy-three studies reporting baseline cortisol levels in patients diagnosed with T2D and healthy controls were included. The standardized mean difference in cortisol levels was calculated using Random effects model. Results: Morning plasma cortisol levels were significantly higher in T2D cases versus controls (P < 0.001). Multivariate logistic regression, adjusted for age and BMI, confirmed this association (OR 1.021; 95% CI 1.009, 1.041; P = 0.007). ROC analysis showed moderate discriminative ability for cortisol (AUC 0.76; 95% CI [0.53, 0.98]), with an optimal cutoff of 390.3 nmol/L. The meta-analyses revealed significantly elevated blood and nocturnal salivary cortisol levels in diabetic patients compared to controls (SMD = 0.97; 95% CI [0.71, 1.22];