Revisiting ‘intensive’ blood glucose control: A causal directed acyclic graph-guided systematic review of randomized controlled trials

Chi Jung Huang, Wei Ting Wang, Shih Hsien Sung, Chen Huan Chen, Gregory Y.H. Lip, Hao Min Cheng*, Chern En Chiang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Scopus citations


Aim: To clarify the importance of HbA1c reduction and antidiabetic drug use in preventing major adverse cardiovascular events (MACE) for patients with type 2 diabetes (T2D). Materials and Methods: We conducted an updated systematic review of contemporary large randomized controlled trials assessing the relative efficacy and safety of antidiabetic drugs with less hypoglycaemia risk in adult T2D patients. Mixed-effects meta-regression was performed to examine the associations of HbA1c reduction with subsequent risk of macrovascular and microvascular events. We evaluated the potential mediating role of HbA1c reduction in the relationship between antidiabetic drugs and MACE. Results: Eighteen placebo-controlled trials comprising 155 610 participants were included. The effects of treatment differed among antidiabetic drug classes for most adverse outcomes with high heterogeneity (I2: 63.7%-95.8%). Mean HbA1c reduction was lowest with dipeptidyl peptidase-4 inhibitors (0.30%), followed by sodium-glucose co-transporter-2 inhibitors (0.46%), and was highest with glucagon-like peptide-1 receptor agonists (0.58%) and thiazolidinediones (0.60%). Lower relative risks of MACE were significantly associated with larger reductions in achieved HbA1c (β −0.3182; 95% CI: −0.5366 to −0.0998; P =.0043), even after adjusting for drug classes. When considering HbA1c lowering as a mediator to be controlled, beneficial effects owing to specific antidiabetic treatment for MACE were not observed (χ2 = 1.4494; P =.6940). The proportion mediated by HbA1c reduction was 50.0%-63.5% for these antidiabetic agents. Conclusions: The main benefits of antidiabetic agents might result from the reduction in blood sugar levels and are generally independent of drugs used. Risk reduction in MACE was proportional to the magnitude of HbA1c decrease conferred by antidiabetic agents with less hypoglycaemic hazard.

Original languageEnglish
Pages (from-to)2341-2352
Number of pages12
JournalDiabetes, Obesity and Metabolism
Issue number12
StatePublished - Dec 2022


  • antidiabetic agents
  • cardiovascular events
  • dipeptidyl peptidase-4 inhibitor
  • glucagon-like peptide-1 receptor agonist
  • HbA1c
  • meta-analysis
  • pioglitazone
  • risk factors
  • sodium-glucose co-transporter-2 inhibitor
  • type 2 diabetes


Dive into the research topics of 'Revisiting ‘intensive’ blood glucose control: A causal directed acyclic graph-guided systematic review of randomized controlled trials'. Together they form a unique fingerprint.

Cite this