BackgroundPreoperative carbohydrate loading is an important element of the enhanced recovery after surgery (ERAS) paradigm in adult patients undergoing elective surgery. However, preoperative carbohydrate loading remains controversial in terms of improvement in postoperative outcomes and safety. We conducted a Bayesian network meta-analysis to evaluate the effects and safety of different doses of preoperative carbohydrates administrated in adult patients after elective surgery.MethodsMEDLINE (PubMed), Web of Science, EMBASE, EBSCO, the Cochrane Central Register of Controlled Trials, and China National Knowledge Infrastructure (CNKI) were searched to identify eligible trials until 16 September 2022. Outcomes included postoperative insulin resistance, residual gastric volume (RGV) during the surgery, insulin sensitivity, fasting plasma glucose (FPG), fasting serum insulin (Fin) level, the serum levels of C-reactive protein (CRP), postoperative scores of pain, patients’ satisfaction, thirst, hunger, anxiety, nausea and vomit, fatigue, and weakness within the first 24 h after surgery and the occurrences of postoperative infection. The effect sizes were estimated using posterior mean difference (continuous variables) or odds ratios (dichotomous variables) and 95 credible intervals (CrIs) with the change from baseline in a Bayesian network meta-analysis with random effect.ResultsFifty-eight articles (N = 4936 patients) fulfilled the eligibility criteria and were included in the meta-analysis. Both preoperative oral low-dose carbohydrate loading (MD: –3.25, 95% CrI: –5.27 to –1.24) and oral high-dose carbohydrate loading (MD: –2.57, 95% CrI: –4.33 to –0.78) were associated with postoperative insulin resistance compared to placebo/water. When trials at high risk of bias were excluded, association with insulin resistance was found for oral low-dose carbohydrate loading compared with placebo/water (MD: –1.29, 95%CrI: –2.26 to –0.27) and overnight fasting (MD: –1.17, 95%CrI: –1.88 to –0.43). So, there was large uncertainty for all estimates vs. control groups. In terms of safety, oral low-dose carbohydrate administration was associated with the occurrences of postoperative infection compared with fasting by 0.42 (95%Crl: 0.20–0.81). In the other outcomes, there was no significant difference between the carbohydrate and control groups.ConclusionAlthough preoperative carbohydrate loading was associated with postoperative insulin resistance and the occurrences of postoperative infection, there is no evidence that preoperative carbohydrate administration alleviates patients’ discomfort.Systematic review registration[https://www.crd.york.ac.uk/PROSPERO/], identifier [CRD42022312944].