Objective To understand the impact of healthy lifestyle on diabetes incidence among adults with different socioeconomic status (SES) in Guizhou province and provide reference for diabetes prevention and control.

Methods A total of 9 280 permanent residents aged ≥ 18 years in Guizhou Province were recruited using multistage cluster random sampling. Face-to-face interviews with a questionnaire developed by the Chinese Center for Disease Control and Prevention, physical examinations, and laboratory tests were conducted among the residents from November 2010 to December 2012 to collect baseline data, and follow-up surveys were conducted from December 2016 to June 2021. After excluding those with a history of diabetes at baseline, unclear diabetes status, loss to follow-up, and unclear follow-up outcomes, 4 747 adults were included in the final analysis. Multivariate Cox proportional hazards regression models were used to analyze the effect of healthy lifestyle on diabetes incidence among adults with different SES.

Results The 4 747 adults were followed for 0.1 – 9.52 years, with a mean follow-up of 6.98 ± 1.44 years. During follow-up, 560 cases of diabetes were identified, with an incidence rate of 11.80% and an incidence density of 16.89/1 000 person-years. Multivariate Cox proportional hazards regression analysis showed that after adjustment for gender, age, ethnicity, marital status, place of residence, family history of diabetes, and SES score, adults with 3, 4, and ≥ 5 healthy lifestyle factors had 34. 5% (hazard ratio HR = 0.655, 95% confidence interval 95%CI: 0.517 – 0.830), 38.1% (HR = 0.619, 95%CI: 0.479 – 0.800), and 70.8% (HR = 0.292, 95%CI: 0.184 – 0.465) lower diabetes risk, respectively, compared with those with ≤ 2 healthy lifestyle factors. After adjustment for gender, age, ethnicity, marital status, place of residence, family history of diabetes, and number of healthy lifestyle factors, adults with an SES score of ≥ 12 had a 28.7% (HR = 0.713, 95%CI: 0.541 – 0.940) lower risk of diabetes compared with those with an SES score of ≤ 8. After adjustment for gender, age, ethnicity, marital status, place of residence, and family history of diabetes, among adults with an SES score of ≤ 8, those with 5 or more healthy lifestyle factors had a 60.4% (HR = 0.396, 95%CI: 0.198 – 0.788) lower risk of diabetes compared with those with 2 or fewer factors. Among adults with SES scores of 9 – 11, those with 3, 4, and ≥ 5 healthy lifestyle factors had 42.0% (HR = 0.580, 95%CI: 0.397 – 0.848), 36.6% (HR = 0.634, 95%CI: 0.424 – 0.947), and 78.0% (HR = 0.220, 95%CI: 0.094 – 0.514) lower diabetes risk, respectively. Among adults with an SES score of ≥ 12, those with 4 and 5 or more healthy lifestyle factors had 50.4% (HR = 0.496, 95%CI: 0.276 – 0.894) and 71.4% (HR = 0.286, 95%CI: 0.110 – 0.740) lower diabetes risk, respectively compared with those with 2 or fewer factors.

Conclusions Both higher SES and more healthy lifestyle factors can reduce the risk of diabetes among adults in Guizhou province. However, those with low SES need to maintain a sufficient number of healthy lifestyle factors to reduce their diabetes risk.