Eating Disorders and Eating Problems Among Adolescents With Type 1 Diabetes: Exploring Relationships With Temperament and Character

Eating Disorders and Eating Problems Among Adolescents With Type 1 Diabetes: Exploring Relationships With Temperament and Character Vasileia Grylli,1 Andrea Hafferl-Gattermayer,1 Gudrun Wagner,1 Edith Schober,2 and Andreas Karwautz1 1 University Clinic of Neuropsychiatry of Childhood and Adolescence and 2University Clinic of Pediatrics and Adolescent Medicine, Vienna, Austria Objective To determine temperament and character among adolescents with type 1 diabetes with and without disordered eating. Method A clinical sample of 199 adolescents from multiple centers with a mean age of 14.1 (SD, 2.5) years were screened and diagnosed for eating disorders. Assessed were temperament and character as conceptualized by Cloninger, glycemic control, and depression. Results Adolescent patients with clinical eating disorders or subthreshold eating problems had significantly higher mean scores in harm avoidance and lower mean scores in self-directedness. Harm avoidance remained significant even after controlling for depressive pathology. Discussion This study is the first to show evidence that among youths (in particular, girls) with type 1 diabetes, there is an association between low selfdirectedness, high harm avoidance, and the presence of eating, weight, and shape pathology. For these particular youths, important implications for clinical practice are outlined. Key words eating disorders; type 1 diabetes; temperament; character; adolescence. Eating disorders like anorexia nervosa (AN) and bulimia nervosa (BN), as well as their partial syndromes (eating disorders not otherwise specified [ED-NOS]), are relatively common among young women in Western cultures and constitute the third most prevalent chronic health condition among adolescent females (Rosen, 2003). About 0.2–0.7% and 1.0–2.5% of adolescent and young adult women suffer from AN and BN, respectively (de Groot & Rodin, 1994; Karwautz & Treasure, 2000). In this context, overweight adolescents predominantly seem to be vulnerable, since a higher prevalence of unhealthy weight control behaviors could be evidenced in this group (Boutelle, Neumark-Sztainer, Story, & Resnick, 2002). Adolescent girls and young women with type 1 diabetes have an increased risk for eating disorders as well as for milder, subthreshold eating problems (Affenito, 2001; Grylli, Karwautz, Hafferl-Gattermayer, & Schober, 2003; Nielsen, 2002; Striegel-Moore, Nicholson, & Tamborlane, 1992). Thus, diabetes in combination with an eating disorder places these young girls in a particularly difficult situation: First, having type 1 diabetes involves the necessary adaptation to a lifestyle that even though liberalized through medical advances still remains far from the “carefree” lifestyle of a person without this chronic disease; second, eating disorders are responsible for a number of disorder-specific disadvantages. As a result of both, patients with type 1 diabetes and an eating disorder are doubly disadvantaged (Ward, Troop, Cachia, Watkins, & Treasure, 1995). In the largest controlled study of eating disorders in adolescent girls with type 1 diabetes to date (Jones, Lawson, Daneman, Olmsted, & Rodin, 2000), the prevalence of eating disorders meeting criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), was 10%, and their subthreshold variants 14%, both being about twice as common in adolescent females All correspondence should be sent to Andreas Karwautz, Eating Disorders Unit, University Clinic of Neuropsychiatry of Childhood and Adolescence, Medical University of Vienna, Waehringer Guertel 18-20, A-1090, Vienna, Austria. E-mail: . Journal of Pediatric Psychology () pp. –,  doi:./jpepsy/jsi Journal of Pediatric Psychology vol.  no.  © Society of Pediatric Psychology ; all rights reserved.  Grylli et al. with type 1 diabetes as in peers without type 1 diabetes. Studies have also pointed to a deterioration of metabolic control in adolescent girls with type 1 diabetes combined with an eating disorder or subthreshold disordered eating behavior (Jones et al., 2000; Rydall, Rodin, Olmsted, Devenyi, & Daneman, 1997), but this topic is still controversial (e.g., Engström et al., 1999; Herpertz et al., 1998). Poor metabolic control is of crucial importance because it seems to be relevant during the onset, maintenance, and deterioration of short-term and long-term diabetes-related microvascular and macrovascular complications. Retinopathy, for example, a serious complication related to type 1 diabetes, is increased fivefold in this population (Nielsen, 2002). Recent developments in the study of general and specific risk factors that favor the manifestation of eating disorders have focused on aspects that are both extrinsic (familial and sociocultural environments, the mother– daughter relationship, or the internalization of the thin ideal as a landmark of beauty in Western cultures) and intrinsic (genetic vulnerability, differential regulation of hormonal release, or function of proteins) (e.g., Fairburn, Cooper, Doll, & Welch, 1999; Karwautz et al., 2001). In this context, personality traits like temperament and character and their putative neurobiological correlates are prominent features in the cascade of predisposing and maintaining risk factors for an eating disorder (Fassino et al., 2002; Karwautz, Rabe-Hesketh, Collier, & Treasure, 2002; Kaye & Strober, 1999; Klump et al., 2000; Woodside et al., 2002). One dimensional system describing normal personality traits is that developed by Cloninger (Cloninger, Przybeck, Svrakic, & Wetzel, 1994), according to which seven traits are modeled into two groups: (1) temperament, containing the traits of novelty seeking (response to novelty), harm avoidance (response to aversive stimuli), reward dependence (response to reward maintenance), and persistence (perseverance despite frustration); and (2) character, consisting of the traits of self-directedness (self-identification as an autonomous individual), cooperativeness (identification of the self as an integral part of the society), and self-transcendence (sense of a unity with all existing things). The first dimension, temperament, is assumed to be correlated with the activity of dopamine, serotonin, and noradrenalin in the brain, whereas the second, character, is based more on social goals and values. This dimensional approach to personality was used to deepen the knowledge about patients with eating disorders and to put forward neurobiological, therapeutic, and prognostic implications (Bulik, Sullivan, Joyce, Carter, & McIntosh, 1998; Cloninger & Svrakic, 1999; Fassino et al., 2002). In particular, temperamental traits as biologically rooted aspects of personality may aid early identification and early prevention of the development of psychiatric disorders and eating disorders. Studies have consistently shown that patients with AN are characterized by high harm avoidance, high pe (...truncated)