Food- and Nutrient-Based Dietary Patterns and Depression in Korean Adults: A Machine Learning Approach Using KNHANES...

Area of Science:

• Nutrition science
• Epidemiology
• Psychiatry

Background:

• Dietary patterns are inconsistently linked to depression.
• Methodological variations in identifying dietary patterns contribute to inconsistent findings.
• Data-driven approaches like machine learning can improve objectivity and reproducibility in dietary pattern identification.

Purpose of the Study:

• To identify dietary patterns using machine learning.
• To examine the association between identified dietary patterns and depression in Korean adults.

Main Methods:

• Utilized data from 21,321 Korean adults (aged 19-64) from the Korea National Health and Nutrition Examination Survey (2016-2021).
• Applied K-means clustering to identify dietary patterns based on food group and nutrient intake.
• Assessed dietary intake via 24-hour recall; depression status determined by physician diagnosis.

Main Results:

• Three distinct dietary patterns were identified using both food group and nutrient-based clustering.
• A balanced, diverse dietary pattern (food group-based) was associated with significantly lower odds of depression (OR 0.64; 95% CI, 0.47-0.88; p=0.007).
• No significant associations were found for nutrient-based clusters or a high-processed food pattern.

Conclusions:

• Adherence to balanced, diverse, whole-food-based dietary patterns is associated with reduced depression risk.
• Food group-based clustering offers more reproducible and interpretable insights than nutrient-based approaches.
• Food group-based clustering has potential utility in epidemiological research and public health strategies for depression prevention.