Traditional and Non-Traditional Clustering Techniques for Identifying Chrononutrition Patterns in University Students

🎖️ Top 10% JournalJan 28, 2026Nutrients

Using Different Grouping Methods to Find Daily Eating Patterns in University Students

AI simplified

Abstract

Five distinct meal timing patterns were identified among 388 Mexican university students, with food intake quality varying significantly between early and late eaters.

  • Clusters included Early, Early-Intermediate, Late-Intermediate, Late, and Late with early breakfast.
  • Chronotype was associated with meal timing patterns, with more morning types in the early clusters.
  • Food intake quality was healthier among early eaters compared to late eaters.
  • Moderate concordance was found between clustering methods, with the highest agreement between traditional and non-traditional techniques.
  • The findings support the need for comparing multiple clustering methods in chrononutrition research.

AI simplified

Key numbers

459
Sample Size
Total participants included in the analysis.
0.376
Concordance Level
Mean Adjusted Rand Index across clustering methods.
41.5%
Percentage of Healthy Eaters
Healthy food intake proportion in K-means identified Early pattern.

Full Text

What this is

  • This study examines chrononutrition patterns among university students in Mexico, focusing on meal timing.
  • It compares four clustering techniques: K-means, Hierarchical, Gaussian Mixture Models, and Spectral clustering.
  • The research identifies five distinct meal timing patterns and assesses the concordance between clustering methods.
  • Findings emphasize the importance of methodological transparency in chrononutrition research.

Essence

  • Five meal timing patterns were identified among Mexican university students using various clustering techniques. While traditional and non-traditional methods yielded similar core structures, they did not produce identical patterns, highlighting the need for methodological transparency.

Key takeaways

  • Five distinct meal timing patterns were identified: Early, Early-Intermediate, Late-Intermediate, Late, and Late with early breakfast. These patterns reflect varying meal times and eating behaviors among university students.
  • Food intake quality differed significantly by meal timing pattern, with early eaters showing healthier intake compared to late eaters. This suggests that meal timing may influence dietary choices.
  • Concordance between clustering methods was moderate (mean Adjusted Rand Index = 0.376), indicating that while core patterns were similar, the choice of clustering algorithm affected the specific boundaries and classifications.

Caveats

  • The cross-sectional design limits causal inferences about meal timing and food intake quality. Longitudinal studies are needed to assess these relationships over time.
  • Self-reported meal timings may introduce recall bias, potentially affecting the accuracy of the data. Objective measures would enhance reliability.
  • The sample was predominantly female (72.8%), which may limit the generalizability of findings to male populations and precludes sex-stratified analyses.

AI simplified