AIMS: This review focuses on outlining the findings of studies that have been conducted to display the glycemic effect of Catha edulis, while trying to balance it with findings of the association of its chewing with the development of type 2 diabetes mellitus (DM).

MATERIALS AND METHODS: The search strategy adopted was based on a comprehensive research in Medline, PubMed, Web of Science, JSTOR, Scopus and Cochrane for articles, proceeding abstracts and theses to identify complete reports written in the English language about the glycemic effect of Catha edulis in humans and animals from 1976 to 2016. In addition, bibliographies were also reviewed to find additional reports not otherwise published. Thirty seven records were identified of which, 25 eligible studies were included in the meta-analysis using blood glucose as an outcome measurement. Studies were divided into four subgroups according to the experimental model, namely; non-diabetic animals, diabetic animals, non-diabetic humans and diabetic humans. The pooled mean difference (MD) of blood glucose between experimental and control were calculated using random effects model of the weighted mean difference of blood glucose with 95% confidence interval (CI). Heterogeneity between studies was tested using I(2) statistic and a value of P<0.05 was considered to indicate statistical significance.

RESULTS: The scientific reports in the literature prevailed that the glycemic effect of Catha edulis were greatly conflicting with the majority of studies indicating that Catha edulis has a mild hypoglycemic effect. However, the meta-analysis indicted that the overall result showed an insignificant reduction in blood glucose (MD=-9.70, 95% CI: -22.17 to 2.76, P=0.13, with high heterogeneity between subgroups, I(2)=88.2%, P<0.0001). In addition, pooled mean difference of blood glucose of non-diabetic animals, diabetic animals and non-diabetic humans showed an insignificant reduction in blood glucose (MD=-18.55, 95% CI: -39.55 to 2.50, P<0.08, MD=-52.13%, 95% CI: -108.24 to 3.99, P=0.07 and MD=-2.71%, 95% CI: -19.19 to -13.77, P=0.75) respectively. Conversely, a significant elevation in the pooled mean difference of blood glucose in diabetic humans was indicated (MD=67.18, 95% CI: 36.93-97.43, P<0.0001). The conflict shown in the glycemic effect of Catha edulis is thought to be cultivar-related, while demographic and epidemiological reports suggested that chewing Catha edulis might be a predisposing factor contributing to the development of type 2 DM.

METHODOLOGY: Studies were identified by searching the SCOPUS, SPORTDiscus, PubMed, Web of Science, and CNKI databases up to May 13, 2024, using the following inclusion criteria: (a) healthy population; (b) comparison of LL-BFR vs HLR training; (c) pre- and post-training assessment of muscle strength (dynamic, isometric, and isokinetic), muscle power, jump, or speed performance; (d) PEDro scale score ≥4. The methodological quality of the included studies was assessed using the PEDro tool and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach, with meta-analyses conducted using the R program.

RESULTS: A total of 41 studies, involving 853 subjects, were included in the meta-analysis. Based on the PEDro scores and GRADE assessment, the overall quality of the included studies was assessed as moderate. LL-BFR training showed a slightly smaller effect on maximal strength compared to HLR training (ES = -0.19, 95% CI [-0.31 to -0.06], p < 0.01). There were no significant differences between LL-BFR and HLR training for muscle power (ES = -0.04, 95% CI [-0.33 to 0.24], p > 0.05), jump performance (ES = -0.08, 95% CI [-0.30 to 0.15], p > 0.05), and speed (ES = -0.28, 95% CI [-0.71 to 0.15], p > 0.05). Additionally, individual characteristics (i.e., age, gender, and training status) and training parameters (i.e., training duration, frequency, cuff pressure, and cuff width) did not significantly moderate the training effect.