The association between a low-carbohydrate diet (LCD) score and the risk of diabetes mellitus (DM) is contradictory. This study is a systemic review of cohort studies that have focused on the association between the LCD score and DM. We searched PubMed/Medline, Scopus, Embase, ISI Web of Science, and Google Scholar for papers published through January 2017 with no language restrictions. Cohort studies that reported relative risks (RRs) with 95% confidence intervals (CI) for DM were included. Finally, 4 studies were considered for our meta-analysis. The total number of participants ranged from 479 to 85 059. Among 4 cohort studies, 8 081 cases with DM were observed over follow-up durations ranging from 3.6 to 20 years. A marginal significant association was observed between the highest LCD score and the risk of DM (RR=1.17; 95% CI: 0.9, 1.51). Moreover, the RRs for studies with energy adjustments showed a significant association (RR: 1.32; 95% CI: 1.17, 1.49; I2: 0%). Based on our findings, study qualities score of less or equal to 7 had a significant influence on the pooled effect size (RR=1.31, 95%CI: 1.15, 1.49; I2: 0%), whereas the overall RR in the studies with quality score more than 7 was 1.09 (95% CI: 0.73, 1.63). In conclusion, we have found that the highest LCD score was marginally associated with the risk of DM. However, more prospective cohort studies are needed to clarify the effects of the LCD score on the risk of DM.
© Georg Thieme Verlag KG Stuttgart · New York.