Long-term Effect of Lifestyle Interventions on the Cardiovascular and All-Cause Mortality of Subjects With Prediabetes and Type 2 Diabetes: A Systematic Review and Meta-analysis

• Published in: Diabetes care Vol. 45; no. 11; pp. 2787 - 2795
• Main Authors: Zucatti, Kelly P., Teixeira, Paula P., Wayerbacher, Laura F., Piccoli, Giovana F., Correia, Poliana E., Fonseca, Natasha K.O., Moresco, Karla S., Guerra, Bruno A., Maduré, Michelle G., Farenzena, Laura P., Frankenberg, Anize D., Brietzke, Elisa, Halpern, Bruno, Franco, Oscar, Colpani, Verônica, Gerchman, Fernando
• Format: Journal Article
• Language: English
• Published: United States American Diabetes Association 01.11.2022
• Subjects: Bias; Cardiovascular diseases; Cardiovascular System; Clinical trials; Data search; Diabetes; Diabetes mellitus; Diabetes mellitus (non-insulin dependent); Diabetes Mellitus, Type 2; Geographical distribution; Geographical locations; Heterogeneity; Humans; Hyperglycemia; Life Style; Lifestyles; Meta-analysis; Mortality; Prediabetic State; Quality; Regression analysis; Risk; Statistical models; Subgroups; Systematic review
• ISSN: 0149-5992; 1935-5548; 1935-5548
• DOI: 10.2337/dc22-0642

Lifestyle interventions improve the metabolic control of individuals with hyperglycemia. We aimed to determine the effect of lifestyle interventions on cardiovascular and all-cause mortality in this population. Searches were made through MEDLINE, Cochrane CENTRAL, Embase, and Web of Science (no date/language restriction, until 15 May 2022). We included randomized clinical trials (RCTs) of subjects with prediabetes and type 2 diabetes, comparing intensive lifestyle interventions with usual care, with a minimum of 2 years of active intervention. Data from the 11 RCTs selected were extracted in duplicate. A frequentist and arm-based meta-analysis was performed with random-effects models to estimate relative risk (RR) for mortality, and heterogeneity was assessed through I2 metrics. A generalized linear mixed model (GLMM) was used to confirm the findings. Lifestyle interventions were not superior to usual care in reducing cardiovascular (RR 0.99; 95% CI 0.79-1.23) or all-cause (RR 0.93; 95% CI 0.85-1.03) mortality. Subgroup, sensitivity, and meta-regression analyses showed no influence of type of intervention, mean follow-up, age, glycemic status, geographical location, risk of bias, or weight change. All of these results were confirmed with the GLMM. Most studies had a low risk of bias according to the RoB 2.0 tool and the certainty of evidence was moderate for both outcomes. Most studies had a low risk of bias according to the RoB 2.0 tool, and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach resulted in moderate certainty of evidence for both outcomes. Differences in lifestyle programs and in usual care between the studies should be considered in the interpretation of our results. Intensive lifestyle interventions implemented so far did not show superiority to usual care in reducing cardiovascular or all-cause mortality for subjects with prediabetes and type 2 diabetes.