Evaluation and interpretation of latent class modelling strategies to characterise dietary trajectories across early life: a longitudinal study from the Southampton Women’s Survey

Abstract

There is increasing interest in modelling longitudinal dietary data and classifying individuals into subgroups (latent classes) who follow similar trajectories over time. These trajectories could identify population groups and timepoints amenable to dietary interventions. This paper aimed to provide a comparison and overview of two latent class methods: group-based trajectory modelling (GBTM) and growth mixture modelling (GMM). Data from 2963 mother-child dyads from the longitudinal Southampton Women’s Survey were analysed. Continuous diet quality indices (DQIs) were derived using principal component analysis from interviewer-administered food frequency questionnaires collected in mothers pre-pregnancy, at 11- and 34-weeks’ gestation, and in offspring at 6 and 12 months and 3, 6-7 and 8-9 years. A forward modelling approach from 1-6 classes was used to identify the optimal number of DQI latent classes. Models were assessed using the Akaike and Bayesian Information Criteria, probability of class assignment, ratio of the odds of correct classification, group membership, and entropy. Both methods suggested that five classes were optimal; with a strong correlation (Spearman’s=0.98) between class assignment for the two methods. The dietary trajectories were categorised as stable with horizontal lines and were defined as poor (GMM=4%, GBTM=5%), poor-medium (23%, 23%), medium (39%, 39%), medium-better (27%, 28%) and best (7%, 6%). Both GBTM and GMM are suitable for identifying dietary trajectories. GBTM is recommended as it is computationally less intensive, but results could be confirmed using GMM. The stability of the diet quality trajectories from pre-pregnancy underlines the importance of promotion of dietary improvements from preconception onwards.

Birth outcomes and early growth patterns associated with age at adiposity rebound: the Ma’anshan birth cohort (MABC) study

“…In this study, the GBTM was used to fit trajectories to children’s repeated measures of physical indicators. Existing studies have demonstrated that both GBTM and growth mixture modelling are suitable for modeling trajectories [ 56 ], and although both methods have different approaches to longitudinal noble modeling, studies have observed strong agreement on the optimal number of trajectories su and have recommended the use of GBTM because it is less computationally intensive [ 57 ]. The strengths of GBTM for fitting somatic growth indicators have been reported and adopted in a number of studies, primarily because it allows researchers to identify and characterize potential subgroups or categories within populations based on different trajectories [ 22 ].…”

Diet Quality Trajectories over Adulthood in a Biracial Urban Sample from the Healthy Aging in Neighborhoods of Diversity across the Life Span Study

“…There is also evidence that dietary patterns during selected life stages can influence eating practices later in life [ 10 , 11 , 12 , 13 ]. For instance, food preferences early in life set the foundation for food choices and eating habits in later childhood and adulthood [ 14 , 15 , 16 ]. Findings from the Saskatchewan Pediatric Bone Mineral Accrual Study suggest that healthy dietary habits established during childhood and adolescence moderately continue into adulthood [ 17 ].…”

“…The group-based trajectory modeling (GBTM) method has been used to explore dietary pattern trajectories since 2014 [ 13 , 14 , 18 , 20 ]. This method uses longitudinal data to account for between-individual variation and describe the continuity of different behaviors of groups of individuals through time [ 21 ].…”

“…One study examining dietary pattern trajectories in Chinese adults found that the individuals in the trajectory with the greatest adherence to a meat dietary pattern were at greatest risk of overweight and obesity [ 22 ]. There are only a few studies using latent class methods to examine changes or stability in diet quality trajectories [ 13 , 14 ]. Mertens et al reported that diet quality, measured by the Healthy Eating Index (HEI)-2010, changed in Flemish adults over 10 years [ 23 ].…”