# Dietary Habits and Age–Health Gradient Among Older Adults in a Region of Japan

## TL;DR

This study explores how dietary habits affect health in older Japanese adults, finding that more diverse diets are linked to better health outcomes.

## Contribution

The study introduces a data-driven classification of dietary habits using an ordinal latent block model and identifies associations with health indicators.

## Key findings

- Dietary habits were categorized into six distinct types representing a continuum from low to high dietary diversity.
- Females with highly diverse diets showed a negative association between age and LDL-C levels.
- The study used a model-based clustering approach to classify dietary habits from food frequency questionnaire data.

## Abstract

Background/Objectives: With increasing life expectancy, interest in healthy aging has grown substantially. Dietary habits are among the key factors that contribute to achieving healthy aging. This study analyzes the relationship between dietary habits and the age–health association in older adults, using the first two years of data from an ongoing annual cohort study conducted in a region of Japan. Methods: We used observational data from approximately 1200 community-dwelling males and females aged 55 to 75 at baseline, drawing on the first two years of a ten-year annual cohort study conducted from 2023 to 2032. First, dietary habits were classified using an ordinal latent block model (OLBM), a model-based clustering approach applied to food frequency questionnaire (FFQ) data. We then examined whether the age–health gradient—measured across 33 indicators—differed significantly across the derived dietary habit types, using random effects models. Results: Dietary habits in the analyzed sample were categorized into six distinct types. Parameter estimates from the model suggest that the extracted patterns represent a continuum ranging from low to high dietary diversity. Regression analyses indicated that, in females, a negative association between age and LDL-C levels was observed among those with highly diverse dietary habits. Conclusions: The data-driven classification of dietary habits based on FFQ responses highlights the potential importance of dietary diversity.