OBJECTIVES: To assess the effects of interventions for people with type 2 diabetes fasting during Ramadan.

SEARCH METHODS: We searched CENTRAL, MEDLINE, PsycINFO, CINAHL, WHO ICTRP and ClinicalTrials.gov (29 June 2022) without language restrictions.

SELECTION CRITERIA: Randomised controlled trials (RCTs) conducted during Ramadan that evaluated all pharmacological or behavioural interventions in Muslims with T2DM.

DATA COLLECTION AND ANALYSIS: Two authors screened and selected records, assessed risk of bias and extracted data independently. Discrepancies were resolved by a third author. For meta-analyses we used a random-effects model, with risk ratios (RRs) for dichotomous outcomes and mean differences (MDs) for continuous outcomes with their associated 95% confidence intervals (CIs). We assessed the certainty of evidence using the GRADE approach.

MAIN RESULTS: We included 17 RCTs with 5359 participants, with a four-week study duration and at least four weeks of follow-up. All studies had at least one high-risk domain in the risk of bias assessment. Four trials compared dipeptidyl-peptidase-4 (DPP-4) inhibitors with sulphonylurea. DPP-4 inhibitors may reduce hypoglycaemia compared to sulphonylureas (85/1237 versus 165/1258, RR 0.53, 95% CI 0.41 to 0.68; low-certainty evidence). Serious hypoglycaemia was similar between groups (no events were reported in two trials; 6/279 in the DPP-4 versus 4/278 in the sulphonylurea group was reported in one trial, RR 1.49, 95% CI 0.43 to 5.24; very low-certainty evidence). The evidence was very uncertain about the effects of DPP-4 inhibitors on adverse events other than hypoglycaemia (141/1207 versus 157/1219, RR 0.90, 95% CI 0.52 to 1.54) and HbA1c changes (MD -0.11%, 95% CI -0.57 to 0.36) (very low-certainty evidence for both outcomes). No deaths were reported (moderate-certainty evidence). Health-related quality of life (HRQoL) and treatment satisfaction were not evaluated. Two trials compared meglitinides with sulphonylurea. The evidence is very uncertain about the effect on hypoglycaemia (14/133 versus 21/140, RR 0.72, 95% CI 0.40 to 1.28) and HbA1c changes (MD 0.38%, 95% CI 0.35% to 0.41%) (very low-certainty evidence for both outcomes). Death, serious hypoglycaemic events, adverse events, treatment satisfaction and HRQoL were not evaluated. One trial compared sodium-glucose co-transporter-2 (SGLT-2) inhibitors with sulphonylurea. SGLT-2 may reduce hypoglycaemia compared to sulphonylurea (4/58 versus 13/52, RR 0.28, 95% CI 0.10 to 0.79; low-certainty evidence). The evidence was very uncertain for serious hypoglycaemia (one event reported in both groups, RR 0.90, 95% CI 0.06 to 13.97) and adverse events other than hypoglycaemia (20/58 versus 18/52, RR 1.00, 95% CI 0.60 to 1.67) (very low-certainty evidence for both outcomes). SGLT-2 inhibitors result in little or no difference in HbA1c (MD 0.27%, 95% CI -0.04 to 0.58; 1 trial, 110 participants; low-certainty evidence). Death, treatment satisfaction and HRQoL were not evaluated. Three trials compared glucagon-like peptide 1 (GLP-1) analogues with sulphonylurea. GLP-1 analogues may reduce hypoglycaemia compared to sulphonylurea (20/291 versus 48/305, RR 0.45, 95% CI 0.28 to 0.74; low-certainty evidence). The evidence was very uncertain for serious hypoglycaemia (0/91 versus 1/91, RR 0.33, 95% CI 0.01 to 7.99; very low-certainty evidence). The evidence suggests that GLP-1 analogues result in little to no difference in adverse events other than hypoglycaemia (78/244 versus 55/255, RR 1.50, 95% CI 0.86 to 2.61; very low-certainty evidence), treatment satisfaction (MD -0.18, 95% CI -3.18 to 2.82; very low-certainty evidence) or change in HbA1c (MD -0.04%, 95% CI -0.45% to 0.36%; 2 trials, 246 participants; low-certainty evidence). Death and HRQoL were not evaluated. Two trials compared insulin analogues with biphasic insulin. The evidence was very uncertain about the effects of insulin analogues on hypoglycaemia (47/256 versus 81/244, RR 0.43, 95% CI 0.13 to 1.40) and serious hypoglycaemia (4/131 versus 3/132, RR 1.34, 95% CI 0.31 to 5.89) (very low-certainty evidence for both outcomes). The evidence was very uncertain for the effect of insulin analogues on adverse effects other than hypoglycaemia (109/256 versus 114/244, RR 0.83, 95% CI 0.44 to 1.56; very low-certainty evidence), all-cause mortality (1/131 versus 0/132, RR 3.02, 95% CI 0.12 to 73.53; very low-certainty evidence) and HbA1c changes (MD 0.03%, 95% CI -0.17% to 0.23%; 1 trial, 245 participants; very low-certainty evidence). Treatment satisfaction and HRQoL were not evaluated. Two trials compared telemedicine with usual care. The evidence was very uncertain about the effect of telemedicine on hypoglycaemia compared with usual care (9/63 versus 23/58, RR 0.42, 95% CI 0.24 to 0.74; very low-certainty evidence), HRQoL (MD 0.06, 95% CI -0.03 to 0.15; very low-certainty evidence) and HbA1c change (MD -0.84%, 95% CI -1.51% to -0.17%; very low-certainty evidence). Death, serious hypoglycaemia, AEs other than hypoglycaemia and treatment satisfaction were not evaluated. Two trials compared Ramadan-focused patient education with usual care. The evidence was very uncertain about the effect of Ramadan-focused patient education on hypoglycaemia (49/213 versus 42/209, RR 1.17, 95% CI 0.82 to 1.66; very low-certainty evidence) and HbA1c change (MD -0.40%, 95% CI -0.73% to -0.06%; very low-certainty evidence). Death, serious hypoglycaemia, adverse events other than hypoglycaemia, treatment satisfaction and HRQoL were not evaluated. One trial compared drug dosage reduction with usual care. The evidence is very uncertain about the effect of drug dosage reduction on hypoglycaemia (19/452 versus 52/226, RR 0.18, 95% CI 0.11 to 0.30; very low-certainty evidence). No participants experienced adverse events other than hypoglycaemia during the study (very low-certainty evidence). Death, serious hypoglycaemia, treatment satisfaction, HbA1c change and HRQoL were not evaluated.