This study systematically assessed the effectiveness of lifestyle interventions on glycemic indicators among adults (⩾18years) without IGT or diabetes. Randomized controlled trials using physical activity (PA), diet (D), or their combined strategies (PA+D) with follow-up ⩾12months were systematically searched from multiple electronic-databases between inception and May 4, 2016. Outcome measures included fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), fasting insulin (FI), homeostasis model assessment-estimated insulin resistance (HOMA-IR), and bodyweight. Included studies were divided into low-range (FPG <5.5mmol/L or HbA1c <5.5%) and high-range (FPG ⩾5.5mmol/L or HbA1c ⩾5.5%) groups according to baseline glycemic levels. Seventy-nine studies met inclusion criteria. Random-effect models demonstrated that compared with usual care, lifestyle interventions achieved significant reductions in FPG (-0.14mmol/L [95%CI, -0.19, -0.10]), HbA1c (-0.06% [-0.09, -0.03]), FI (%change: -15.18% [-20.01, -10.35]), HOMA-IR (%change: -22.82% [-29.14, -16.51]), and bodyweight (%change: -3.99% [-4.69, -3.29]). The same effect sizes in FPG reduction (0.07) appeared among both low-range and high-range groups. Similar effects were observed among all groups regardless of lengths of follow-up. D and PA+D interventions had larger effects on glucose reduction than PA alone. Lifestyle interventions significantly improved FPG, HbA1c, FI, HOMA-IR, and bodyweight among adults without IGT or diabetes, and might reduce progression of hyperglycemia to type 2 diabetes mellitus.
Keywords: Glucose regulation; Lifestyle intervention; Meta-analysis; Systematic review.
Published by Elsevier B.V.