Obective  To explore effects of different dietary patterns on the prevalence of prediabetes in adult residents in Jiangsu province and to provide evidences for the intervention of rational dietary pattern on diabetes.

  Methods  Multistage cluster random sampling was used to select 1 865 non-diabetic adult residents in Jiangsu province from May 2014 through December 2015. Questionnaire survey, physical examination and clinical tests were conducted among the residents to collect essential data. Factor analysis was adopted to identify the dietary patterns and logistic regression model was applied to analyze the influence of the dietary patterns on prediabetes.

  Results  Totally 315 participants with prediabetes were identified and the prevalence of prediabetes was 16.9% among the participants. The results of factor analysis confirmed four dietary patterns in the participants: animal, healthy, traditional, and starchy-vegetable diet patterns and the cumulative contribution of the four dietary patterns to total variance was 54.000%. The results of multivariate logistic regression analysis indicated that healthy diet pattern was a protective factor against prediabetes after adjusting for confounding factors as gender, age, residence, smoking, alcohol drinking, whether with parenal diabetes, body mass index, hypertension, and dyslipidemia. The risk of prediabetes for the participants with a dietary pattern score in the third and the fourth interquartile range were 0.689 (odds ratio OR = 0.689, 95% confidence interval 95%CI: 0.481 – 0.986; P = 0.042) and 0.518 (OR = 0.518 , 95%CI: 0.356 – 0.756; P = 0.001) compared to the participants with the score in the first interquartile range.

  Conclusion  Healthy diet pattern can reduce the risk of prediabetes in adult residents in Jiangsu province.