Effect of L-carnitine supplementation on cardiac-metabolic risk factors in hemodialysis patients: a systematic review and meta-analysis of clinical trials

Several studies have investigated the potential beneficial effects of L-carnitine in hemodialysis (HD) patients. Therefore, we conducted a systematic review and meta-analysis of clinical trials that evaluated the impact of L-carnitine supplementation on cardiometabolic risk factors, including weight, body mass index (BMI), triglycerides (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), systolic blood pressure (SBP), diastolic blood pressure (DBP), albumin, and hemoglobin in HD patients. A systematic search was conducted in Web of Science, Scopus, PubMed, and Cochrane databases without language or time restrictions up to January 2026 to identify relevant clinical trials. The pooled weighted mean differences (WMDs) and 95% CIs were computed using the random-effects model. Sixteen studies comprising a total of 699 patients were included in the meta-analysis. The findings showed that L-carnitine supplementation had no significant effect on weight ( P = 0.890), BMI ( P = 0.733), TG ( P = 0.898), LDL-C ( P = 0.465), TC ( P = 0.180), HDL-C ( P = 0.988), albumin ( P = 0.820), or hemoglobin ( P = 0.322) in HD patients. In exploratory subgroup analyses, L-carnitine supplementation was associated with significant reductions in TG (dose ≥ 1000 mg/day), TC (age < 50 years), SBP (dose < 1000 mg/day), and DBP (duration > 12 weeks), as well as an increase in HDL-C (dose < 1000 mg/day and duration ≤ 12 weeks). All subgroup findings are hypothesis-generating and require confirmation. Taken together, the present study found no significant overall effects of oral L-carnitine supplementation on cardiometabolic risk factors in HD patients. Exploratory subgroup analyses proposed potential effects on TG, HDL-C, TC SBP, and DBP in specific subgroups; however, these results are hypothesis-generating due to the risk of Type I error, small subgroup sample sizes, and substantial heterogeneity. These findings should be interpreted with caution and confirmed in future well-designed studies. CRD42023480114.