Legume and soy intake and risk of type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies

Abstract

Previous findings on the associations of legume and soy intake with the risk of type 2 diabetes are conflicting. We aimed to summarize the longitudinal associations between legume and soy intake and risk of type 2 diabetes. We searched for relevant prospective cohort studies in PubMed, EMBASE, and Ovid up to August 2019. Study-specific, multivariable-adjusted RRs and 95% CIs were pooled by random-effects models. We identified 15 unique cohorts including 565,810 individuals and 32,093 incident cases. The summary RRs (95% CIs) of incident type 2 diabetes were 0.95 (0.79, 1.14; NS) for total legumes, 0.83 (0.68, 1.01; NS) for total soy, 0.89 (0.71, 1.11; NS) for soy milk, 0.92 (0.84, 0.99) for tofu, 0.84 (0.75, 0.95) for soy protein, and 0.88 (0.81, 0.96) for soy isoflavones, respectively. High heterogeneity was found for total legumes (I2=84.8%), total soy (I2=90.8%), and soy milk (I2=91.7%). Potential sources of heterogeneity were not evident for total legumes or soy milk, whereas for total soy, geographic location (Asia, United States; P=0.04) and study quality (high, moderate, or low; P = 0.02) significantly predicted heterogeneity. In dose-response analysis, significant linear inverse associations were observed for tofu, soy protein, and soy isoflavones (all P < 0.05). Overall quality of evidence was rated as moderate for total legumes and low for total soy and soy subtypes. Dietary intakes of tofu, soy protein, and soy isoflavones, but not total legumes or total soy, are inversely associated with incident type 2 diabetes. Our findings support recommendations to increase intakes of certain soy products for the prevention of type 2 diabetes. However, the overall quality of evidence was low and more high-quality evidence from prospective studies is needed. This trial was registered as PROSPERO CRD42019126403 (https://www.crd.york.ac.uk/PROSPERO).