Use of Latent Profile Analysis to Identify Eating Disorder Phenotypes in an Adult Australian Twin Cohort

School of Psychology, Flinders University, PO Box 2100, Adelaide, SA 5001, Australia.
Archives of General Psychiatry (Impact Factor: 14.48). 01/2007; 63(12):1377-84. DOI: 10.1001/archpsyc.63.12.1377
Source: PubMed


The relationships among the different eating disorders that exist in the community are poorly understood, especially for residual disorders in which bingeing or purging occurs in the absence of other behaviors.
To examine a community sample for the number of mutually exclusive weight and eating profiles.
Data regarding lifetime eating disorder symptoms and weight range were submitted to a latent profile analysis. Profiles were compared regarding personality, current eating and weight, retrospectively reported life events, and lifetime depressive psychopathology.
Longitudinal study among female twins from the Australian Twin Registry in whom eating was assessed by a telephone interview.
A community sample of 1002 twins (individuals) who had participated in earlier waves of data collection.
Number and clinical character of latent profiles.
The best fit was a 5-profile solution with women who were (1) of normal weight with few lifetime eating disorders (4.3%), (2) overweight (10.6% had a lifetime eating disorder), (3) underweight and generally had no eating disorders except for 5.3% who had restricting anorexia nervosa, (4) of low to normal weight (89.0% had a lifetime eating disorder), and (5) obese (37.0% had a lifetime eating disorder). Each profile contained more than 1 type of lifetime eating disorder except for the third profile. Women in the first and third profiles had the best functioning, with women in the fourth and fifth profiles having similarly poorer functioning. The women in the fourth group had a symptom profile distinctive from the other 4 groups in terms of severity; they were also more likely to have had lifetime major depression and suicidality.
Lifetime weight ranges and the severity of eating disorder symptoms affected clustering more than the type of eating disorder symptom.

Full-text preview

Available from:
  • Source
    • "To avoid local maxima, starting values for optimization were fixed at 500 random starts and 50 iterations at initial stage (Geiser, 2010). The class solution was evaluated by its empirical fit regarding parsimony and quality of classification: A minimization of the adjusted Bayesian information criteria (aBIC) and of the cross classification probabilities and a maximization of entropy and the log-likelihood value were desired (Gollwitzer, 2008;Nylund, Asparouhov, & Muthén, 2007;Wade, Crosby, & Martin, 2006). An average latent class probability of at least .80 "
    [Show abstract] [Hide abstract]
    ABSTRACT: This study’s purpose is to describe European adolescents’ alcohol use patterns by grouping adolescents regarding their current alcohol use by cluster analysis (CA). Discriminant and latent profile analyses (LPA) evaluate and validate the solution that will be described further by ANOVAs. From 25 European countries, 57,771 students (49.4% male, 13.87 years) are grouped using hierarchical and k-means clustering. Alcohol use is measured by frequency of drinking occasions during the previous month and number of beverages consumed on the last drinking occasion. CA suggests four drinking patterns: mild (73.6%), episodic (20.0%), frequent (3.8%), and heavy episodic use (2.5%). Discriminant analysis attests a classification reliability of 94%, and confirmatory LPA replicates the cluster solution with a satisfying model fit. Three of the found patterns fulfill criteria for heavy drinking and underline the importance of individualized indicated prevention by promoting responsible use.
    Full-text · Article · Jun 2015 · Journal of drug issues
  • Source
    • "As binge-eating and purging occur in many people with AN, the syndrome is further subdivided into " AN-restricting (AN-R) " and " AN-binge-eating/purging " (AN- BP) types, depending upon the presence or absence of binge-eating and purging behaviors. Individuals who binge and purge (whether they have AN-BP or BN) are noted to be more prone to behavioral disinhibition and affective instability than are those who solely restrict food intake (Rosval et al., 2006; Steiger et al., 2013; Wu et al., 2013), and various findings validate the concept that " restricting-only " and " binge-eating/purging " individuals, regardless of weight status, fall into separate classes (Keel et al., 2004; Williamson et al., 2005; Wade et al., 2006; Wildes and Marcus, 2013). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Eating disorder (ED) variants characterized by "binge-eating/purging" symptoms differ from "restricting-only" variants along diverse clinical dimensions, but few studies have compared people with these different eating-disorder phenotypes on measures of neurocognitive function and brain activation. We tested the performances of 19 women with "restricting-only" eating syndromes and 27 with "binge-eating/purging" variants on a modified n-back task, and used functional magnetic resonance imaging (fMRI) to examine task-induced brain activations in frontal regions of interest. When compared with "binge-eating/purging" participants, "restricting-only" participants showed superior performance. Furthermore, in an intermediate-demand condition, "binge-eating/purging" participants showed significantly less event-related activation than did "restricting-only" participants in a right posterior prefrontal region spanning Brodmann areas 6-8-a region that has been linked to planning of motor responses, working memory for sequential information, and management of uncertainty. Our findings suggest that working memory is poorer in eating-disordered individuals with binge-eating/purging behaviors than in those who solely restrict food intake, and that observed performance differences coincide with interpretable group-based activation differences in a frontal region thought to subserve planning and decision making. Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.
    Full-text · Article · Feb 2015 · Psychiatry Research: Neuroimaging
  • Source
    • "LPA is a so-called “person centered” approach, which means that observations are clustered on subject basis, unlike factor analysis, in which observations are clustered on item basis. LPA has been frequently used, also in the field of eating behavior, e.g. to identify eating disorder phenotypes in a twin cohort study in Australia [38]. We determined the number of latent profiles based on the minimization of Bayesian information criteria (BIC) [39] and Akaike information criteria (AIC) indices and a non-significant Lo-Mendell-Rubin Likelihood Ratio Test (LMR-LRT) [40] to test model fit. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Definitions and assessment methods of fussy/picky eating are heterogeneous and remain unclear.We aimed to identify an eating behavior profile reflecting fussy/picky eating in children and to describe characteristics of fussy eaters. Eating behavior was assessed with the Child Eating Behavior Questionnaire (CEBQ) in 4914 4-year olds in a population-based birth cohort study. Latent Profile Analysis (LPA) was used to identify eating behavior profiles based on CEBQ subscales. We found a "fussy" eating behavior profile (5.6% of children) characterized by high food fussiness, slowness in eating, and satiety responsiveness in combination with low enjoyment of food and food responsiveness. Fussy eaters were more often from families with low household income than non-fussy eaters (42% vs. 31.8% respectively; Chi2(1) = 9.97, p < .01). When they were 14 months old, fussy eaters had a lower intake of vegetables (t [3008] = 2.42, p < .05) and fish (t [169.77] = 2.40, p < .05) but higher intake of savory snacks (t [153.69] = -2.03, p < .05) and sweets (t [3008] = -2.30, p < .05) compared to non-fussy eaters. Also, fussy eaters were more likely to be underweight at 4 years of age (19.3%) than non-fussy eaters (12.3%; Chi2(1) = 7.71, p < .01). A distinct fussy eating behavior profile was identified by LPA, which was related to family and child characteristics, food intake, and BMI. This behavior profile might be used in future research and the development of interventions.
    Full-text · Article · Feb 2014 · International Journal of Behavioral Nutrition and Physical Activity
Show more