LAUSR

AIM This study aimed to determine the efficacy of personalized lifestyle interventions on the reversion of a prediabetic state to normoglycemia compared with regular blood glucose testing alone in prediabetes. METHODS A randomized, multicenter trial was conducted in prediabetes aged 30-70 with fasting blood glucose level of 5.6-6.9mmol/L (100-125mg/dL) and/or HbA1c level of 39-46mmol/mol (5.7-6.4%) recruited from health checkups at 16 health-promotion centers in Korea. The 799 recruited individuals were randomized to either the personalized lifestyle intervention group (LIG) or the control group (CG) by a computer generated random number list prepared by an independent statistician. The CG was provided with fasting blood glucose and HbA1c tests alone every 3months during the first year. The LIG was provided not only blood glucose test but five sessions of personalized lifestyle counseling by nutritionists every 3months during the first year aimed at improving the diet, alcohol and exercise behaviors. Data from lifestyle assesments and laboratory measurements were analyzed at 1-year after baseline. The primary outcome was the reversion rate from prediabetes to normoglycemia. Additional outcome include the effect of the lifestyle intervention program on lifestyle changes in the LIG to support primary outcome. RESULTS The 799 participants randomly allocated to the LIG (n=398) or the CG (n=401). For the analyses of outcomes, 629 participants (313 men and 316 women; mean age, 53.7±9.4years; mean body mass index (BMI), 24.7kg/m2) were included: 325 in the LIG; 304 in the CG. Diet (7.03, 95% CI=4.56-10.86, P<0.001), alcohol (2.24, 95% CI=1.48-3.41, P<0.001), and exercise behaviors (1.85, 95% CI=1.31-2.63, P<0.001) were improved relative to baseline by the personalized lifestyle intervention in the LIG after adjusting age, sex, and family history of diabetes. In terms of main outcome, the cumulative incidence of reversion from prediabetes to normoglycemia at the first year was 37.9% (123/325) [95% CI=32.6-43.1%] in the LIG and 29.6% (90/304) (95% CI=24.5-34.7%) in the CG. After adjustment for age, sex, family history of diabetes, BMI, blood pressure, and lipids, the hazard ratio for reverting to normoglycemia remained significantly higher in the LIG (1.40, 95% CI=1.06-1.83, P=0.017) than in the CG. CONCLUSION Personalized lifestyle intervention could be more effective compared with regular blood glucose testing alone in the reversion of a prediabetic state to normoglycemia in Korean prediabetics. This finding suggests that diabetes prevention care would be benefited by incorporating personalized lifestyle counseling. This study was registered at cris.nih.go.kr (KCT0001580).