Eating disorders and type 1 diabetes

This Update investigates the link between insulin misuse and eating disorders in female patients with diabetes, and looks at the need for better screening and detection. It is by Dr Susan Hart, BSc, MNutrDiet, PhD.

Dr Susan Hart BSc, MNutrDiet, PhD.
Dr Hart is an Accredited Practising Dietitian; Program Coordinator,
Derwent House, Royal Prince Alfred Hospital Eating Disorders Day Program, NSW.

INTRODUCTION
INSULIN misuse, motivated by weight and shape concerns, is a well-documented behaviour in young women with type 1 diabetes (T1DM).

Of the few published interventions targeting this behaviour, one study delivered a six-week intervention, which improved eating disorder attitudes but did not improve glycaemic control.¹ An epidemiological study within Australia reported a lifetime prevalence of eating disorders as 15.4%, with diagnoses of anorexia nervosa (4.3%), bulimia nervosa (2.9%), binge eating disorder (2.9%) and eating disorder not otherwise specified (EDNOS, 5.3%).²

The average age of onset of eating disorders is 14–16 years. The average duration of illness is 5–7 years, with long-term impairment of quality of life in domains such as financial security, interpersonal relationships, workplace functioning and physical problems.

Many studies have clearly identified that women with T1DM have higher rates of eating disorders. In 356 women with T1DM, 10% met diagnostic thresholds for anorexia and bulimia, while 14% met thresholds for EDNOS, which was twice the rate of non-diabetic women in this cross-sectional sample.³

AETIOLOGY OF EATING DISORDERS
Overall risk factors for the development of an eating disorder are perfectionist traits, placing a high value on thinness, body image dissatisfaction and low self-esteem.

Specific aetiological factors contributing to the development of eating disorders in women with T1DM include:

• The onset of a diagnosis of diabetes as a life stress⁴ and the effects of this on family functioning and dynamics⁵
• The daily challenge of managing diabetes-related tasks such as self-monitoring and insulin injections
• The ease by which insulin can be misused⁵
• Rapid weight changes that occur following diagnosis,⁶ which can trigger dietary restraint and attempts at weight loss. Average weight gain after diagnosis was reported as 6.9kg (range  0–15kg), with significant increases in body image dissatisfaction and drive for thinness in the sample. Two of 17 women developed bulimia nervosa with insulin misuse at 12-month follow-up⁷
• A feeling of loss of control over one’s body, with control factors being a central feature of eating disorders⁵
• Women with T1DM are two BMI points higher than their peers, predisposing them to greater body image dissatisfaction^3,8
• When women who have been misusing insulin start complying with insulin injections, there can be initial weight gain related to oedema,⁹ as well as longer-term weight gain related to improved blood glucose levels. As HbA_1c levels improve, weight also increases.¹⁰ In the Diabetes Control and Complications Trial, women with intensive treatment and well-controlled diabetes were 4.75kg heavier and more likely to be overweight (BMI > 27.3).

The short- and long-term changes in weight that can occur when insulin is recommenced can quickly de-motivate young people from continuing with their insulin injections.

In fact, studies have found that the key issues associated with both the emergence and resolution of insulin restriction are fear of weight gain and problems with diabetes self-care.¹¹

INSULIN MISUSE
In T1DM, insulin omission is the most common eating disorder behaviour observed.¹² There can often be associated beliefs that omitting insulin is a successful way to control weight.

A recent study showed that women who engaged in insulin misuse had a higher BMI than those who stopped restricting insulin.¹¹

An Australian self-report online survey of 148 individuals¹³ revealed that up to 60% of respondents had restricted insulin at some time, and 11% were currently doing so at the time of the survey.

The research also showed that only 12% of people with T1DM had ever had a healthcare professional directly ask them if they were skipping or restricting insulin.

Signs of insulin misuse for weight control are:⁹

• Increasing neglect of diabetes management, both blood glucose monitoring and insulin omission
• Binge eating and other purging behaviours such as vomiting
• Deterioration in psychosocial functioning, including work functioning and interpersonal relationships
• Elevated HbA_1c
• Recurrent hospital admissions for diabetic ketoacidosis
• Erratic clinic attendance
• Significant weight gain or loss
• Poor body image/low self-esteem
• Depressive symptoms – depression and anxiety are known to be more prevalent in T1DM.¹⁴

There may also be inflexibility about food choices in young adults who may cite the reason for their inflexibility as ‘it’s because of my diabetes’ and not because of weight and shape.⁵

Insulin restriction has been associated with greater eating disorder symptoms, diabetes-specific distress, overall psychological symptoms, and fear of hypoglycaemia at baseline.

Microvascular complications have also been shown to develop earlier and with more severity than in those without an eating disorder.

Retinopathy at four-year follow-up was shown to present at a rate of 86% in T1DM with an eating disorder compared to 24% in T1DM without one.¹⁵

Furthermore, the relative risk of death with self-reported insulin restriction has been found to be 3.2 times higher than non-insulin restrictors.¹⁶ In this study, age of death was also younger among insulin restrictors (mean age = 45 years) compared to those reporting appropriate insulin use (mean age = 58 years). Other reasons for non-compliance with insulin injections and diabetes self-care are related to a fear of hypoglycaemia, with individuals running glucose levels high to avoid episodes.

SCREENING AND DETECTION
There are no standardised and validated methods of detecting insulin misuse.

The clinical observations of patients presenting to a specialist eating disorder clinic in Sydney suggest that:

• some women don’t take any insulin
• some women take basal insulin, but not their bolus doses when they eat
• other women may adhere to injections where possible, but intentionally underdose when they break their dietary rules, binge or overeat.

However, it is unclear how this may be generalised to other clinics or treatment settings, as clinicians generally have poor knowledge of these patterns of behaviour and they have not been examined.

Some researchers have used a single screening item (‘I take less insulin than I should’) to determine insulin restriction status, with responses rated on a six-point Likert scale ranging from ‘never’ to ‘always’.¹¹

The use of this question in routine clinical practice has the potential to identify at-risk women so that interventions can be offered.¹⁶

Other authors have recommended routine screening of all adults with T1DM¹⁷ with questions such as:

• How concerned are you about your weight?
• Are you on a diet to lose or maintain weight?
• Do you ever skip meals to lose or maintain weight?
• How often do you skip insulin?

More research on screening for insulin misuse is needed.

INTERVENTION
There is limited research into psychological interventions in young women with eating disorders and T1DM.

An intervention targeting insulin misuse in T1DM achieved a reduction in eating disorder behaviours at six-month follow-up but no change in HbA_1c.¹

Research has shown that, in itself, being ready to change is a prerequisite for the change process but not sufficient to achieve meaningful glycaemic improvements.¹⁸ An increase in psychological and coping skills is also needed.

Educational intervention alone has been found to be an insufficient approach among eating disorder populations as it does not address the required skills to cope with emotional distress and weight concerns.

Cognitive behavioural therapy
Based on the above considerations, cognitive behavioural therapy (CBT) provides a good framework to improve outcomes in poorly controlled adults with T1DM.¹⁹

The aims of CBT interventions are to facilitate the acquisition of skills and confidence to integrate effective self-management into daily lives, and provide the skills training necessary to manage the psychological and emotional impact of living with diabetes.²⁰

This approach is beneficial as it is designed to help individuals cope more effectively, optimise self-care behaviours and reduce negative emotions towards diabetes, therefore enhancing glycaemic control.¹⁸ CBT is also particularly suitable to improve glycaemic control in patients with high levels of depressive symptoms.¹⁴

Most literature to date has focused on adolescents or adults who are poorly compliant with their diabetes self-management and who don’t have eating disorders. Aspects of these CBT interventions include the use of problem-solving techniques, cognitive restructuring targeting fears of hypoglycaemia, motivational interviewing, family-based counselling, enhancing general coping skills, and building self-efficacy in relation to diabetes management. There have been a number of CBT inventions with positive outcomes.^19,20,21

For eating disorder patients, individual or group interventions aimed at increasing self-esteem, appearance, and body acceptance are also recommended,²² and family-based interventions aimed at developing flexible approaches to food and meal planning are optimal.

FUTURE RESEARCH
Currently, by the time T1DM patients access a specialist mental health or eating disorder team, their disorder is longstanding.

There is also the likelihood that they may already have significant diabetes-related complications.⁵

Early intervention models with appropriate screening to detect at-risk individuals are necessary, as once the pattern of frequent and habitual insulin restriction becomes entrenched, the cycle of negative feelings about body image, shape and weight; chronically elevated blood sugars; depression, anxiety and shame; and poor diabetes self-care can be complex and difficult to treat.²²

Because of the lack of research on this topic, a clinical trial is being designed to target insulin misuse in young adults using a CBT format.

This is being done in collaboration with the Boden Institute of Obesity, Nutrition, Exercise and Eating Disorders (University of Sydney) and the departments of psychiatry and endocrinology at Royal Prince Alfred Hospital.

This project will have two primary aims: first, to develop and validate a screening questionnaire for insulin misuse; and second, to develop and deliver a structured, manualised, group-based education program based on
CBT models and Fairburn’s trans-diagnostic CBT-enhanced model²³ to treat this population.

CONCLUSION
Insulin misuse is a problem in women with T1DM with significantly increased morbidity and mortality.

Both diabetes and eating disorder specialists are somewhat lacking in evidence-based interventions for treatment.

Considering the number of young adults with diabetes and the documented prevalence of insulin misuse, this represents an enormous unmet need.

KEY POINTS

• In type 1 diabetes, insulin omission is the most common eating disorder behaviour observed.
• Insulin misuse is a problem in women with type 1 diabetes with significantly increased morbidity and mortality.
• Despite the prevalence of the behaviour, both diabetes and eating disorder specialists are somewhat lacking in evidence-based interventions for treatment.
• Early intervention is required before the behaviours are entrenched; there are significant diabetes-related complications and the condition is difficult to treat.