LAUSR

OBJECTIVE To perform a qualitative synthesis (systematic review) and quantitative analysis (meta-analysis) to summarize the evidence regarding the relationship between basal cortisol levels and metabolic syndrome (MetS) in adults. METHODS A systematic search was performed in the PubMed, Embase, and PsycINFO databases for observational studies on the association between basal cortisol levels and MetS. The quality of individual studies was assessed by the Newcastle-Ottawa score. A random effects model was used to report pooled quantitative results and the I2 statistic was used to assess heterogeneity. Egger's and Begg's tests were used to evaluate publication bias. RESULTS Twenty-six studies (19 cross-sectional and seven case-control) met the inclusion criteria for the systematic review. The majority was classified as having a low risk of bias and used established criteria for the diagnosis of MetS. Twenty-one studies provided data on basal cortisol levels as continuous values and were included in the meta-analysis; they comprised 35 analyses and 11,808 subjects. Pooled results showed no significant difference in basal cortisol levels between subjects with and without MetS (standardized mean difference [SMD] = 0.02, 95% confidence interval [CI]=-0.11 to 0.14). There was high heterogeneity between the studies when all comparisons were considered (I2 = 83.1%;p < 0.001). Paradoxically, meta-analysis of studies evaluating saliva samples showed no significantly lower basal cortisol levels among subjects with MetS (SMD=-0.18, 95% CI=-0.37 to 0.01), whereas those studies that evaluated serum samples (SMD = 0.11, 95% CI=-0.02 to 0.24) and urine samples (SMD = 0.73, 95% CI=-0.40 to 1.86) showed no significantly higher basal cortisol levels among subjects with MetS. In the subgroup and meta-regression analyses, a significant difference in basal cortisol levels was observed according to study design, population base, age, gender, cortisol level assessment method, and study quality. CONCLUSION This systematic review and meta-analysis does not reveal any association between basal cortisol levels and MetS based on results of observational studies. The results of a random-effect meta-analysis showed no significant difference in basal cortisol levels between subjects with and without MetS. The present findings should be considered in order to help future studies.