Empirically Derived Dietary Patterns Using Robust Profile Clustering in the Hispanic Community Health Study/Study of Latinos

Abstract

Background Latent class models (LCMs) have been used in exploring dietary behaviors over a wide set of foods and beverages in a given population, but are prone to overgeneralize these habits in the presence of variation by subpopulations. Objectives This study aimed to highlight unique dietary consumption differences by both study site and ethnic background of Hispanic/Latino populations in the United States, that otherwise might be missed in a traditional LCM of the overall population. This was achieved using a new model-based clustering method, referred to as robust profile clustering (RPC). Methods A total of 11,320 individuals aged 18–74 y from the Hispanic Community Health Study/Study of Latinos (2008–2011) with complete diet data were classified into 9 subpopulations, defined by study site (Bronx, Chicago, Miami, San Diego) and ethnic background. At baseline, dietary intake was ascertained using a food propensity questionnaire. Dietary patterns were derived from 132 food groups using the RPC method to identify patterns of the general Hispanic/Latino population and those specific to an identified subpopulation. Dietary patterns derived from the RPC were compared to those identified from an LCM. Results The LCM identified 48 shared consumption behaviors of foods and beverages across the entire cohort, whereas significant consumption differences in subpopulations were identified in the RPC model for these same foods. Several foods were common within study site (e.g., chicken, orange juice, milk), ethnic background (e.g., papayas, plantain, coffee), or both (e.g., rice, tomatoes, seafood). Post hoc testing revealed an improved model fit in the RPC model [Deviance Information Criterion DICRPC = 2.3 × 104, DICLCM = 9.5 × 106]. Conclusions Dietary pattern behaviors of Hispanics/Latinos in the United States tend to align by ethnic background for some foods and by location for other foods. Consideration of both factors is imperative to better understand their contributions to population health and developing targeted nutrition intervention studies.

Shared and ethnic background site-specific dietary patterns in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL)

“…Within HCHS/SOL, two previous papers derived a posteriori DPs using food items/groups at either ethnic background-specific (23) or EBS-specific (24) levels. Maldonado et al (23) described one fully and 4 partially reproducible DPs derived using FA on 34 food groups from 24-hour recalls stratifying by 6 ethnic background-specific categories.…”

“…This cohort provides the unique opportunity to define both shared and subpopulation-specific DPs by the cross-section of ethnic background and site (EBS), in support of future culturally tailored interventions. Two papers within HCHS/SOL have identified shared and subpopulation-specific a posteriori DPs using approaches conceptually similar to the MSFA (23, 24). Stephenson et al (24) used robust profile clustering to cluster participants and food items (from the food propensity questionnaire) based on consumption behaviors shared amongst all participants and those specific to 9 EBS subpopulations (24).…”

“…Two papers within HCHS/SOL have identified shared and subpopulation-specific a posteriori DPs using approaches conceptually similar to the MSFA (23, 24). Stephenson et al (24) used robust profile clustering to cluster participants and food items (from the food propensity questionnaire) based on consumption behaviors shared amongst all participants and those specific to 9 EBS subpopulations (24). Compared to robust profile clustering, MSFA derives DPs analogous to those from FA or PCA, providing a more straightforward interpretation for nutrition researchers.…”

Identifying Dietary Consumption Patterns from Survey Data: A Bayesian Nonparametric Latent Class Model

“…Model generation for complex survey data has fallen under two main approaches: (1) generate a pseudo-like population from the observed study data via a combination of bootstrapping and resampling techniques (Savitsky and Toth, 2016;Rao and Thomas, 1988;Skinner and Wakefield, 2017); (2) generate model parameter estimates first and correct for them using the sampling weights post hoc for population-based estimates and inference (Vermunt, 2002;Vermunt and Magidson, 2007;Stephenson et al, 2020b;Mattei et al, 2016). Patterson et al (2002) and Vermunt and Magidson (2007) have implemented surveyweighted approaches to latent class models, generated under a frequentist framework.…”

“…Bayesian nonparametric mixture models have been applied in nutritional settings but has not required sampling weights into the parameter estimation of the model (Fahey et al, 2007; De Vito et al, 2019; Stephenson et al, 2020a). Stephenson et al (2020b) applied a Bayesian nonparametric mixture model to diet survey data, but the weights were applied after parameters were estimated from the sampling algorithm. Kunihama et al (2016) et al used a Dirichlet process mixture model to introduce a sampling algorithm that can incorporate survey weights directly into the estimation of a Bayesian nonparametric mixture model.…”