Background: Testosterone affects insulin resistance, but the effect of testosterone treatment on type 2 diabetes (T2D) remains controversial. We aimed to investigate the association between circulating total testosterone (TT) and glycaemic variability using continuous glucose monitoring (CGM) in patients with T2D.
Methods:A total of 248 men with T2D were enrolled in the study. Clinical characteristics and plasma for glycated haemoglobin (HbA1c) and C-peptide assessment were collected. TT was measured using a chemiluminescent immunometric assay.All patients were subjected to a 3-day CGM before making adjustments for hypoglycaemic therapy.Results: TT positively correlated with the standard deviation of mean blood glucose (SDBG) (P < 0.05), especially in older patients. Linear regression analysis showed that SDBG was associated with HbA1c (β = 0.354, P < 0.001) and TT (β = 0.164, P = 0.008) after adjusting for age, duration of diabetes, body mass index, fasting/postprandial C-peptide, and use of different hypoglycaemic drugs. The cut-off value of TT for predicting glycaemic variability was 14.76 mmol/L according to receiver operating characteristic (ROC) analysis. SDBG, the coefficient of variation, the incremental area under the curve of glucose (AUC) > 10 mmol/L, and AUC night were increased in the group with TT > 14.76 nmol/L (P < 0.01 for all variables). Body mass index and fasting/postprandial C-peptide were lower in the group with TT > 14.76 nmol/L than in the group with TT ≤ 14.76 nmol/L (P < 0.05).Conclusions: Circulating TT levels should be assessed in patients with T2D in addition to HbA1c for predicting glycaemic variability. More frequent blood glucose monitoring or CGM is suggested for patients with T2D and high testosterone levels.Clinical trials registration: NCT03519529, ClinicalTrials.gov KEYWORDS glycaemic variability, testosterone, type 2 diabetes Bo Ding and Rui Sun contributed equally to this article.