Subtyping female adolescent psychiatric inpatients with features of eating disorders along dietary restraint and negative affect dimensions

https://doi.org/10.1016/S0005-7967(03)00073-1Get rights and content

Abstract

Cluster-analytic studies of eating disorders in adult patients have yielded two subtypes (pure dietary and mixed dietary-negative affect). This study aimed to replicate the subtyping in female adolescent psychiatric inpatients with features of eating disorders. Cluster analyses of 137 patients with eating-disordered features revealed a dietary-negative affect subtype (43%) and a pure dietary subtype (57%). The dietary-negative affect subtype was characterized by greater likelihood of binge eating, greater eating-related psychopathology, and greater body image dissatisfaction. The two subtypes did not differ significantly in scores reflective of clinical syndromes (other than the significantly higher depressive affect in the negative affect subtype), but the dietary-negative affect subtype was characterized by greater personality disturbance and higher reported concerns in clinical areas, including suicidality and childhood abuse. The cluster analysis produced different results from an alternative approach to subtyping by vomiting. These findings provide further support for the reliability and validity of this subtyping scheme for eating pathology. Clinically, the findings suggest that the combination of dieting and negative affect signals a more disturbed variant of eating-disorder related psychopathology in female adolescent psychiatric inpatients.

Introduction

Risk factor models and studies of the development and maintenance eating disorders (Fairburn et al., 2003, Fairburn, Welch, Doll, Davies and O’Connor, 1997, Stice, 2002) have emphasized the potential roles of dietary restraint (Polivy & Herman, 1993) and affect regulation (Heatherton & Baumeister, 1991). Dietary restraint models posit that excessive dieting or dietary restraint increases the likelihood of binge eating (Lowe, 1993). Affect regulation models posit that emotional disturbances coupled with deficits in coping increase the likelihood of binge eating (Grilo and Shiffman, 1994, Leon, Fulkerson, Perry and Early-Zald, 1995). The dual-pathway model (Stice, 1994, Stice, 2001) posits that problems with either—or both—dietary restraint or affective regulation may trigger binge eating.

Research has found that heightened dietary restraint and negative affect predict the onset of disordered eating (Stice, 2001) and that binge eating frequently occurs during negative mood states (Agras and Telch, 1998, Davis, Freeman and Garner, 1988, Grilo, Shiffman and Carter-Campbell, 1994). The literature is not unequivocal (Stice, 2002). In one study (Fairburn et al., 2003), negative affect emerged as a weak predictor and dietary restraint was not significantly predictive of the maintenance or persistence of bulimia nervosa (BN). In contrast, a different study (Stice & Agras, 1998) found that dietary restraint—but not negative affect—significantly predicted bulimic symptomatology (subthreshold for diagnosis).

Stice and colleagues (Stice and Agras, 1999, Stice, Agras, Halmi, Mitchell and Wilson, 2001) posited that some eating-disordered individuals conform better to the dietary restraint model whereas others conform better to the negative affect model. Stice and Agras (1999) ‘subtyped’ 265 females with BN using measures of dietary restraint and negative (depressive) affect. Cluster analyses revealed a ‘pure dietary’ subtype (62% of patients) and a ‘dietary-depressive’ subtype (38% of patients). While the two subtypes were similar in their frequency of binge eating and purging, the dietary-depressive subtype was characterized by significantly higher levels of other features of eating disorders (weight- and shape-concerns) and higher levels of psychosocial maladjustment. In a second study, Stice et al. (2001) subtyped 159 females with binge eating disorder (BED) using the same approach. Cluster analyses revealed a ‘pure dietary’ subtype (64% of cases) and a ‘dietary-negative affect’ subtype (36% of cases). Unlike the Stice and Agras findings for BN (Stice & Agras, 1999), the dietary-negative affect subtype was characterized by a higher frequency of binge eating; consistent with the findings for BN, however, the BED dietary-negative affect subtype had significantly levels of the features of eating disorders and associated psychiatric and social maladjustment. Two recent studies provided further support for the subtyping findings in adults with BN (Grilo, Masheb, & Berman, 2001) and with BED (Grilo, Masheb, & Wilson, 2001a).

Since eating and body image disturbances most frequently begin during adolescence, continued investigation of subtypes should target this developmental era. In addition, it has been suggested (Grilo, Devlin, Cachelin, & Yanovski, 1997) that such research efforts consider these problems broadly rather than solely focusing on samples selected based on formal diagnoses (e.g. anorexia nervosa or BN) or being above threshold for formal diagnoses (Fairburn, Cooper and Shafran, 2003, Garfinkel, Kennedy and Kaplan, 1995). The relevance of studying adolescents with features of eating and body image disturbances is suggested by several convergent findings. Studies have found that adolescents more often report onset of a single symptom rather than multiple symptoms (Stice, Killen, Hayward, & Taylor, 1998a). Many persons who present for treatment for eating disorders are ‘partial syndrome’ or ‘eating disorder not otherwise specified’ (Bunnell, Shenker, Nussbaum, Jacobson and Cooper, 1990, Fairburn et al., 2003). Studies have found that isolated symptoms of eating disorders (at levels below threshold for diagnosis) prospectively predict the onset of binge eating and purging during late adolescence (Stice, Killen, Hayward, & Taylor, 1998b). Available studies with adults (Grilo et al., in press, Kendler, MacLean, Neale, Kessler, Heath and Eaves, 1991, Striegel-Moore et al., 2000) and with adolescents (Lewinsohn, Striegel-Moore, & Seeley, 2000) have found few meaningful differences between persons who meet full criteria versus those who have partial syndrome. Longitudinal research has found that symptoms of eating-related concerns prospectively predict ‘partial syndrome’ eating disorders (Killen, Taylor, Hayward, Haydel, Wilson, Hammer, Kraemer, Blair-Greiner and Strachowski, 1996, Leon, Fulkerson, Perry and Early-Zald, 1995), formal diagnoses of eating disorders (Herzog, Hopkins, & Burns, 1993), and onset of depression among initially non-depressed adolescents (Stice, Hayward, Cameron, Killen, & Taylor, 2000).

Thus, the present study aimed to test Stice’s (Stice and Agras, 1999, Stice, Agras, Halmi, Mitchell and Wilson, 2001) subtyping scheme in female adolescent psychiatric inpatients with features of eating disorders. The first aim was to examine the reliability of the subtyping scheme by attempting to replicate it in an additional and independent patient group. The second aim was to investigate the validity of the subtyping. Concurrent validity was examined by comparing the subtypes on a number of behavioral and psychological variables. In addition, the concurrent validity of Stice’s (Stice and Agras, 1999, Stice, Agras, Halmi, Mitchell and Wilson, 2001) subtyping scheme was compared to that of an alternative approach (by the purging/non-purging distinction).

Section snippets

Participants

Participants were 137 adolescent girls selected for the presence of eating-disorder features (see subsequently) from a nearly consecutive series of 281 female adolescent inpatients admitted to the evaluation and crisis intervention unit of a private not-for-profit teaching psychiatric hospital. These patients were hospitalized for a variety and range of serious psychiatric problems (i.e. this was not an eating-disorder unit characterized by particular recruitment and selection criteria).

Cluster analysis

Participants’ scores on the dietary restraint items, the BDI, and the RSE were analyzed with a cluster analysis. Cluster analysis (SPSS Quick Cluster algorithm; Norusis, 1994) groups cases on the basis of similarity in levels of the selected variables. Quick Cluster selects k participants (k is the number of cluster requested), with well-separated non-missing values as initial centers, and then iteratively clusters participants into one of the groups on the basis of squared Euclidean distances.

Discussion

This study provides further support for the reliability and validity of Stice’s (Stice and Agras, 1999, Stice, Agras, Halmi, Mitchell and Wilson, 2001) subtyping model of eating pathology. The subtyping scheme was extended to a study group of female adolescent psychiatric inpatients with features of eating disorders. Cluster analysis revealed a dietary-negative affect subtype (43%) and a pure dietary subtype (57%). The dietary-negative affect subtype was characterized by greater likelihood of

Acknowledgements

This research was supported by a Donaghue Medical Research Foundation Investigator Award.

References (51)

  • D.W Bunnell et al.

    Subclinical versus formal eating disorders: differentiating psychological features

    International Journal of Eating Disorders

    (1990)
  • R Davis et al.

    A naturalistic investigation of eating behavior in bulimia nervosa

    Journal of Consulting and Clinical Psychology

    (1988)
  • C.G Fairburn et al.

    Risk factors for binge eating disorder: a community-based, case-control design

    Archives of General Psychiatry

    (1998)
  • C.G Fairburn et al.

    Understanding persistence of bulimia nervosa: a five-year naturalistic study

    Journal of Consulting and Clinical Psychology

    (2003)
  • C.G Fairburn et al.

    Risk factor for bulimia nervosa: a community-based, case control study

    Archives of General Psychiatry

    (1997)
  • P.E Garfinkel et al.

    Views on classification and diagnosis of eating disorders

    Canadian Journal of Psychiatry

    (1995)
  • P.E Garfinkel et al.

    Purging and non-purging forms of bulimia nervosa in a community sample

    International Journal of Eating Disorders

    (1996)
  • D.M Garner

    Eating Disorder Inventory—2: Professional manual

    (1991)
  • C.M Grilo

    The assessment and treatment of binge eating disorder

    Journal of Practical Psychiatry and Behavioral Health

    (1998)
  • C.M Grilo et al.

    Report of the National Institutes of Health (NIH) workshop on the development of research priorities in eating disorders

    Psychopharmacology Bulletin

    (1997)
  • C.M Grilo et al.

    A comparison of adolescent inpatients with and without substance abuse using the Millon Adolescent Clinical Inventory

    Journal of Youth and Adolescence

    (1996)
  • C.M Grilo et al.

    Subtyping women with bulimia nervosa along dietary and negative affect dimensions: a replication in a treatment-seeking sample

    Eating and Weight Disorders

    (2001)
  • C.M Grilo et al.

    Subtyping binge eating disorder

    Journal of Consulting and Clinical Psychology

    (2001)
  • C.M Grilo et al.

    A comparison of different methods for assessing the features of eating disorders in patients with binge eating disorder

    Journal of Consulting and Clinical Psychology

    (2001)
  • Grilo, C. M., Sanislow, C. A., Shea, M. T., Skodal, A. E., Stout, R. L., Pagano, M. E., Yen, S., & McGlashan, T. H....
  • Cited by (35)

    • Contribution of temperament to eating disorder symptoms in emerging adulthood: Additive and interactive effects

      2015, Eating Behaviors
      Citation Excerpt :

      Martin et al. (2000) found that children with higher NE are at a greater risk for developing ED symptoms during adolescence, indicating that early temperament can have implications for ED development later in life. NE has also been associated with the degree of body dissatisfaction and rates of binge eating in a sample of women with an ED (Grilo, 2004), and may moderate the link between dietary restraint and binge eating (Stice, Akutagawa, Gaggar, & Agras, 2000). These findings highlight the possibility that NE may predispose individuals to develop ED symptoms when exposed to additional risk factors (e.g., poor self-regulation, dietary restraint).

    • Cognitive Behavioral Treatment for Recurrent Binge Eating in Adolescent Girls: A Pilot Trial

      2013, Cognitive and Behavioral Practice
      Citation Excerpt :

      Adolescents in general seem to experience more intense and labile moods (Larson, Csikszenthmihalyi, & Graef, 1980), and youth with eating disorders have been found to have very high rates of comorbid mood disorders (Lewinsohn et al., 2000). Subtyping studies suggest that 30% to 40% of people who seek treatment for clinical binge eating fit a “dietary depressive” subtype whose binge eating is typified by greater negative affect and more depressed mood (Grilo, 2004; Grilo, Masheb, & Berman, 2001; Grilo, Masheb, & Wilson, 2001; Stice, 2001; Stice & Agras, 1999). Clinical samples of adults who fit this dietary depressive subtype report more severe eating disorder behavior (i.e., binge eating) and weight and shape concerns; they also have higher rates of other Axis I and II disorders, poorer social adjustment, and poorer response to treatment (Chen & le Grange, 2007; Grilo, 2004; Grilo, Masheb, & Berman, 2001; Grilo, Masheb, & Wilson, 2001; Stice, 2001; Stice & Agras, 1999).

    View all citing articles on Scopus
    View full text