had a psychiatric disorder than the rest of the patients” (p. 291). A recent study found that onset of insulin restriction in women with T1D was associated with fear of weight gain and problems with the self-management regimen (Goebel-Fabbri 2011). Problematic eating behavior specific to the diabetes care regimen appears to be part of a constellation of pervasive noncompliance associated with higher psychiatric morbidity or poorer adjustment to illness (Pollock 1995, Wilfley 2000, Pollock-BarZiv 2005, Goebel-Fabbri 2011).
Two studies demonstrated an association between psychiatric morbidity and DEB in patients with T2D independent of weight status. In one study, overweight and obese patients had more diagnosable ED, and patients with ED had significantly more anxiety disorders and trended towards being more depressed. In the second, DEB was also strongly associated with psychopathology such as depression, low self-esteem, and general psychopathology but not weight (Herpertz 1998b, Papelbaum 2005). Given the known comorbidity between emotional disorders (depression in particular) (Anderson 2001) and diabetes, and between emotional disorders and DEB in the healthy population (Telch 1998, Stice 1999, Stice 2000, Stice 2001, Stice 2002), DEB could be part of a constellation of poor psychological adjustment and/or poor adjustment to illness, which is comorbid with overweight and T2D (Herpertz 1998a).
Behaviors considered triggers for and pathognomonic of DEB are embedded in the diabetes treatment regimen (Bantle 2006). Lack of success with MNT can leave patients feeling out of control of both eating behavior and glycemia (Surgenor 2002). Feeling out of control of eating behavior, preoccupation with food, and calorie restriction are DSM-IV-TR diagnostic criteria for bulimia, binge eating disorder, and eating disorder not otherwise specified (American Psychiatric Association 2000). Primary criteria for binge eating disorder include subjective self-evaluation of repeatedly eating amounts of food in a short period of time that are “definitely larger than most individuals would eat under similar circumstances.” Making this subjective determination (when an amount of food is large or excessive) for an individual with diabetes could be attributable to failure to adhere to MNT prescription, especially in the context of treatment of hypoglycemia. Other possibilities exist for misattribution of adherent behavior as DEB (Polonsky 1999). As caloric restraint is prescribed as part of treatment, inaccuracies in judgment regarding appropriateness of food intake can occur in the context of carbohydrate counting, falling blood glucose level, misjudgment of the causes of symptoms (Johnson 2000, Hay 2003, Davis 2004), or excessive nutrition intake related to exercise.
Hormonal evidence for dysregulation of hunger and satiety in patients with diabetes suggests difficulty controlling food intake and consequent blood glucose levels. Further, nonphysiologic dosing of insulin impacts appetite regulation (Young-Hyman 2010). Hormonal dysregulation (including loss of endogenous insulin and amylin secretion) (Koda 1992, Kruger 1999), dysregulation of incretin production, which contributes to metabolism in the gut (Dupre 2005, Higgins 2007), complications of the disease such as gastroparesis (Parkman 2004), and fluctuations in blood glucose level, particularly hypoglycemia (ADA 2002), may predispose vulnerable patients to adoption of maladaptive weight management strategies (such as insulin manipulation) to control hunger and associated weight gain.
Measurement of ED and DEB
Most studies to date have used measurement tools standardized in the general population to establish the presence of ED and DEB in patients with diabetes. Questionnaires include but are not limited to the Eating Attitudes Test (EAT) 40 (Garner 1979) and EAT-26 (Garner 1982), Eating Disorder Inventory (EDI-3) (Garner 2004), the Bulimia Test - Revised (BULIT-R) (Thelen 1991), and the Eating Disorder Examination (EDE) (Cooper 1989), which is conducted in interview format. Evaluation tools include items about attitudes and behaviors that are embedded in the diabetes treatment regimen. For example BULIT-R items (“Do you feel you have control over the amount of food you consume?” and “I eat a lot of food when I’m not even hungry.”) could refer to the diabetes care regimen (the former by prescription of dietary restraint and the latter by a prescribed meal plan), carbohydrate to insulin ratio, and/or treatment for low blood glucose. Diabetes care providers identified more than twenty questions on the EDI-3 that could be answered in the context of treatment and endorsed independent of weight concerns (Young-Hyman 2010). When questionnaires standardized in healthy populations are used, scores may be elevated in patients with both T1D and T2D, due to the overlap in items which are diagnostic of disordered eating attitudes and DEB and prescribed as part of diabetes treatment (Daneman 1998). When questionnaires or interview techniques standardized in the nondiabetic population are used, it is recommended that questions be modified to address intent of behaviors, including insulin manipulation (Criego 2009). Some studies have expanded the EDE and SCID interview format to include such questions (Peveler 2005).
Two questionnaire exceptions were found: the Diabetes Eating Problems Survey (DEPS), created by Antisdel, Laffel, and Anderson (2001) and refined by Markowitz et al. (2010), and the AHEAD (Assessing Health and Eating among Adolescents with Diabetes) survey (Neumark-Sztainer 2002a). Both include questions regarding the adjustment of insulin specifically for the purposes of weight reduction, and both couch questions in terms of diabetes care and issues related to glycemic control and weight gain due to treatment (Antisdel 2001, Neumark-Sztainer 2002a). However, neither questionnaire has been validated in a clinical population with an independent diagnosis of ED. Findings from a study validating a questionnaire that assesses hunger and satiety in the context of diabetic care, the Diabetes Treatment and Satiety Scale (DTSS-20), suggest that patients with T1D routinely experience contradictory clinical situations (regarding blood glucose levels, usual MNT, and insulin dosing) during which they feel full, hungry, and/or out of control of food intake (Young-Hyman 2011a). It is speculated that the lack of appropriate hunger and satiety cues is related to hormonal dysregulation. (See section regarding physiologic dysregulation of appetite.)
To establish diagnosis of ED or document subclinical DEB, thorough evaluation in the diabetes population should include assessment of adjustment to illness, overall psychological status, weight and shape concerns, specific questions regarding maladaptive use of the insulin or medication regimen to lose weight, and reliability of proprioceptive cues regarding hunger and satiety in the context of blood glucose levels (Young-Hyman 2010).
Limitations of Current Research Findings
Gaps in research regarding the association of DEB and diabetes include: 1) assessment of DEB in the context of adherence to medical regimen and adjustment to illness; 2) understanding the contribution of insulin and medication dosing to feelings of hunger and satiety; 3) dietary prescriptions/medical nutrition therapy as potential sources of information/attitudes leading to feelings of loss of control over food intake; 4) need for appropriate comparison groups such as healthy-weight matched individuals seeking to prevent weight gain or weight loss, minority comparison groups, and other chronic disease groups with conditions affecting weight/metabolism; 5) incomplete psychological characterization of samples; 6) discrimination of preexisting/evolving psychopathology associated with noncompliance; and 7) the need to use diabetes-specific assessment tools.
Prior studies of the prevalence of DEB in individuals with diabetes have not systematically addressed issues of regimen intensity, responsibility taking for regimen decisions, expectations of health care providers for glycemic and weight outcomes, and eating cognitions associated with medical care. The contributions of diabetes treatment knowledge or regimen adherence to the prevalence or endorsement of DEB are not established. Very few studies were found to have monitored patients from the time of diagnosis to establish the relationships or chronology between psychological symptomatology (depression in particular), DEB, weight gain, or regimen adjustment to prevent weight gain. Studies in adult women have not taken into account use of hormones for birth control or hormone replacement therapy (HRT), which can cause excess weight gain and increase appetite (Abrams 1992, Gallo 2004, Gallo 2006). Last, no studies to date have simultaneously assessed physiologic markers of hormones and incretins contributing to dysregulation of hunger and satiety and symptoms of DEB in the context of diabetes care.
Recommendations for Screening and Care
DEB is accepted as a serious and potentially life-threatening comorbidity of diabetes, despite controversy about diagnosis and prevalence. No published randomized controlled