Related Experiment Video
Updated: Apr 19, 2026
Measuring Attention and Visual Processing Speed by Model-based Analysis of Temporal-order Judgments
Published on: January 23, 2017
Identifying temporal eating patterns: a comparison of latent class analysis and dynamic time warping-based cluster
Beshada R Jima1, Rebecca M Leech1, David W Dunstan2
1Institute for Physical Activity and Nutrition, School of Exercise and Nutrition Sciences, Deakin University, Burwood, Victoria, Australia.
Background:
Temporal eating patterns (TEPs) are associated with diet quality and obesity, although inconsistencies exist because of variations in methods and input variables. Direct comparisons of analytical approaches for deriving TEPs are rare.
Objectives:
The aim of this study was to compare latent class analysis (LCA) and modified dynamic time warping (MDTW)-based cluster analysis for deriving TEPs and examine their associations with diet quality and obesity.
Methods:
This cross-sectional study included 672 adults (18-65 y) in Victoria, Australia, from the "EveryDayLife" survey (2017-2020). Participants completed a 1- to 7-d food diary via the "FoodNow" app. LCA used hourly presence/absence of eating occasions (EOs), whereas MDTW-based clustering used hourly energy intake (EI) input variables. Methods were compared using patterns visualization, membership overlap, kappa statistics, adjusted R2, and AUC.
Results:
Both methods identified 3 distinct TEPs. Class 1/Cluster 1 had peaks during 07:00-09:00, 12:00, and 18:00-19:00 h ("conventional" pattern). Class 2/Cluster 2 had later peaks in EOs or EI (after 13:00 h). Class 3/Cluster 3 showed modest, evenly spaced EOs or EI concentrated earlier in the day. Membership overlap between similar TEPs was 56.2%-73.1%, with fair agreement (κ = 0.38, P < 0.001). Class 1/Cluster 1 showed higher diet quality than Class 2/Cluster 2, respectively, whereas no significant associations were observed with BMI. LCA explained slightly more variance in diet quality (6% compared with 4%) compared with MDTW-based clustering, with a similar proportion observed for BMI (∼13%). The AUCs for discriminating high diet quality (LCA: 0.635 compared with MDTW: 0.616; P = 0.565) and obesity (LCA: 0.758 compared with MDTW: 0.756; P = 0.934) were not significantly different between the 2 methods.
Conclusions:
LCA and MDTW-based clustering identified comparable yet noninterchangeable TEPs, suggesting both approaches are suitable for deriving TEPs.
More Related Videos
07:12Protocol for Data Collection and Analysis Applied to Automated Facial Expression Analysis Technology and Temporal Analysis for Sensory Evaluation
Published on: August 26, 2016
11:52Temporal Ordering of Dynamic Expression Data from Detailed Spatial Expression Maps
Published on: February 9, 2017
Related Concept Videos
00:54Time-Series Graph
01:05Noncompartmental Analysis: Mean Residence Time
After the administration of a drug through intravenous bolus injection, the drug molecules are distributed throughout the body and remain there for varying periods. The MRT represents the average time these drug molecules stay in the...