TOPLINE

In the absence of obesity, low-to-moderate alcohol consumption is not strongly or consistently associated with an increased risk for type 2 diabetes, whereas obesity remains a strong predictor regardless of levels of alcohol intake, according to results from a 15-year longitudinal study of adults in Korea.

Although analyses of additive, multiplicative, and time-varying effects show some contradictory model-dependent findings, obesity and high alcohol consumption, independently and jointly, remain the factors most consistently associated with the risk for type 2 diabetes.

METHODOLOGY

• Researchers conducted a longitudinal study of 7817 adults aged 40 years or older from the Ansan and Ansung cohorts of the Korean Genome and Epidemiology Study who did not have diabetes at baseline.
• Participants underwent follow-up assessments every 2 years from 2001-2002 through 2017-2018, including evaluations of incident type 2 diabetes, BMI, and alcohol consumption.
• Alcohol consumption was classified as none (0 g/d), low (< 10 g/d), moderate (10-29.9 g/d), or high (≥ 30 g/d).
• Researchers examined baseline and time-varying obesity and alcohol exposure using Cox proportional hazards, additive hazards, and interaction models.

TAKEAWAY

• Obesity and high alcohol consumption, assessed at baseline or over time, were each independently associated with an increased risk for incident diabetes, with stronger associations in time-varying exposures.
• Participants with both obesity and high alcohol consumption had the greatest risk for diabetes (hazard ratio, 2.12).
• In adjusted Cox proportional hazards models, low-to-moderate alcohol intake was not associated with incident diabetes independently of obesity. However, a significant association emerged in an additive hazards model, indicating that findings varied by statistical approach.
• Obesity was consistently associated with a higher risk for incident diabetes across all levels of alcohol consumption.

IN PRACTICE

“This study provides insights into the dynamic impact of modifiable risk factors on diabetes and their public health implications by incorporating both baseline and time-varying exposures and evaluating additive interactions using multiplicative models,” the authors wrote.

SOURCE

This study was conducted by first author Junghyun Yoon, of the Department of Preventive Medicine, Hanyang University College of Medicine in Seoul, Republic of Korea. It was published in Diabetologia.

LIMITATIONS

Alcohol intake and other possible confounders were self-reported. Participants were recruited primarily from two Korean communities, and population-specific genetic and demographic characteristics may limit the generalizability of the findings.

DISCLOSURES

This study was supported by a National Research Foundation of Korea grant funded by the Korean government (Ministry of Science and ICT). The authors reported no relevant disclosures.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.