**2. Randomised controlled trial of an eating disorder (bulimia nervosa) mental health literacy intervention (BN-MHL)**

#### **2.1 Aims of BN-MHL trial**

The study aims were to test the efficacy of a mental health literacy intervention for eating disorders in a non-clinical sample of adult women. Outcomes included mental health literacy regarding treatments for a common eating disorder, bulimia, perceived health related quality of life and general and specific eating disorder psychological symptoms.

#### **2.2 Methods of BN-MHL trial**

The sample was derived from a longitudinal survey of women with disordered eating recruited through advertisements in four universities and colleges of higher education in two Australian States (Queensland and Victoria). Details of the total sample at baseline have been reported in Mond et al. (2010). Recruitment strategies varied and included approaches via central University email/web mail, printed advertisements in student bulletins and halls of residence and direct approach to students in University common areas. For individuals approached via email, participants were given the option of completing an on-line questionnaire. For other participants, questionnaires were provided in hard copy with reply-paid envelopes. The questionnaire included measures of eating disorder psychopathology and health-related quality of life (as completed by the first sample, see below).

The sample for the trial comprised 217 symptomatic young women (all > 18 years, mean age 24.5 years SD 7.6) who agreed to follow-up. They were included if they had current extreme weight/shape concerns and/or current regular (e.g. occurring weekly over the past three months) binge eating and/or any extreme weight control behaviours such as self-induced

Striegel-Moore, 2006; Hepworth & Paxton, 2007), and a belief that one could or should

Many of these reasons for the under-utilisation of health care in eating disorders are features of 'mental health literacy', a term introduced and defined by Jorm as "knowledge and beliefs about mental disorders that may aid in their recognition, management and treatment" (Jorm et al, 1997). Jorm and colleagues, and others, have argued that poor mental health literacy is a major factor in the individual, social and economic burden of mental health problems (Andrews et al., 2000; Jorm et al., 2000). There have been attempts to evaluate the efficacy of mental health literacy interventions in improving outcomes for patients with problems such as depression. In one study Jorm and colleagues (2003) reported a large community-based RCT (n=1094) for an evidenced based guide to treatments versus a general brochure for people with depressive symptoms. They found

In the area of eating disorders we have conducted a small randomized controlled study of a brief postal mental health literacy intervention in community women with bulimic eating disorders. At the end of a year symptomatic improvement, less pessimism about how difficult eating disorders are to treat, improved recognition and knowledge, as well as increased help-seeking were observed in both groups (Hay et al., 2007a). Those randomized to receive the mental health literacy intervention also had improved mental health related quality of life. The study supported further investigations of the role of targeted health

**2. Randomised controlled trial of an eating disorder (bulimia nervosa) mental** 

The study aims were to test the efficacy of a mental health literacy intervention for eating disorders in a non-clinical sample of adult women. Outcomes included mental health literacy regarding treatments for a common eating disorder, bulimia, perceived health related quality of life and general and specific eating disorder psychological symptoms.

The sample was derived from a longitudinal survey of women with disordered eating recruited through advertisements in four universities and colleges of higher education in two Australian States (Queensland and Victoria). Details of the total sample at baseline have been reported in Mond et al. (2010). Recruitment strategies varied and included approaches via central University email/web mail, printed advertisements in student bulletins and halls of residence and direct approach to students in University common areas. For individuals approached via email, participants were given the option of completing an on-line questionnaire. For other participants, questionnaires were provided in hard copy with reply-paid envelopes. The questionnaire included measures of eating disorder psychopathology and health-related quality of life (as completed by the first sample, see

The sample for the trial comprised 217 symptomatic young women (all > 18 years, mean age 24.5 years SD 7.6) who agreed to follow-up. They were included if they had current extreme weight/shape concerns and/or current regular (e.g. occurring weekly over the past three months) binge eating and/or any extreme weight control behaviours such as self-induced

handle the problem alone (Becker et al., 2004; Cachelin & Striegel-Moore, 2006).

more positive outcomes in the former group but the effects were not large.

literacy interventions in eating disorders described in this chapter.

**health literacy intervention (BN-MHL)** 

**2.1 Aims of BN-MHL trial** 

**2.2 Methods of BN-MHL trial** 

below).

vomiting and/or laxative/diuretic use and/or fasting or severe food restriction and/or 'driven' exercise and/or who self-identified on the BN-MHL survey as currently having a problem like that of 'Naomi' (see below – only one was included on this criteria alone). The majority of students (179, 84%) were Australian born and 150 (72%) were never married.

At the start of the first year (baseline) the participants who agreed to follow-up were randomised to receive either a bulimia nervosa mental health literacy (BN-MHL) intervention (n=97) or information about their symptom scores and local mental health services only, with the comparison group (as required by ethical consideration) receiving the intervention at the end of the first year. The intervention comprised a single posted package of information about treatment of BN and related disorders, purchasing information on the book "Binge eating and Bulimia nervosa: A guide to recovery" (Cooper, 1995). The recommended book included a detailed psycho-educational section and a selfdirected cognitive-behaviour therapy. The package also provided recommended websites for further information on treatments, lists and contact details of local eating disorder specialist treatment facilities, and contact details for the (local) eating disorders support group and consumer organisation. At baseline the control group (n=120) received information about local mental health services only.

Randomisation was by means of SPSS RV.BINOM (1,0.5) function and allocation was concealed from the research officer who communicated with the participants. In the covering letter informed consent was obtained, along with permission for follow-up in order to "find out how health issues and general health and well-being impacts on people's quality of life over time". Participants were not told they were part of a randomised controlled trial. Three respective institutional ethics committees approved the research (namely James Cook, La Trobe and Western Sydney universities), with the proviso that control participants were provided with the intervention at one year.

ED symptoms were assessed with the Eating Disorder Examination Questionnaire (EDE-Q). The EDE-Q has been validated in community and clinic samples of people with EDs (Fairburn & Beglin, 1994; Mond et al., 2004). It yields a global score of ED attitudes and restraint, and four sub-scales (i.e. shape, weight and eating concern and dietary restraint) and also frequency of ED behaviours such as binge-eating over the preceding four weeks.

BN-MHL was assessed with a questionnaire designed for this research (Mond et al., 2010). A vignette describing a (fictional) 19-year-old female suffering from BN called Naomi (N) was presented. Care was taken to ensure that the core features of the disorder were present while avoiding the use of medical terminology. The text of the vignette was: *N is a 19-yearold second year arts student. Although mildly overweight as an adolescent, N's current weight is within the normal range for her age and height. However, she thinks she is overweight. Upon starting university, N joined a fitness program at the gym and also started running regularly. Through this effort she gradually began to lose weight. N then started to "diet," avoiding all fatty foods, not eating between meals, and trying to eat set portions of "healthy foods," mainly fruit and vegetables and bread or rice, each day. N also continued with the exercise program, losing several more kilograms. However, she has found it difficult to maintain the weight loss and for the past 18 months her weight has been continually fluctuating, sometimes by as much as 5 kilograms within a few weeks. N has also found it difficult to control her eating. While able to restrict her dietary intake during the day, at night she is often unable to stop eating, bingeing on, for example, a block of chocolate and several pieces of fruit. To counteract the effects of this bingeing, N takes water tablets. On other occasions, she vomits after overeating. Because of her strict routines of eating and exercising, N has become isolated from her friends.* 

Targeted Prevention in Bulimic Eating Disorders:

**Main problem** 

**Most helpful therapy** 

**Most helpful medication** 

**Most helpful professional** 

group to have fewer subjective binges at 6-months.)

**Baseline 6-months 12-months** 

**BN-MHL I-only BN-MHL I-only** 

Bulimia nervosa 39 (18%) 18 (28%) 26 (33%) 10 (16%) 8 (6.7%) Other ED 35 (16%) 15 (23%) 8 (6.7%) 16 (26%) 19 (16%) Low self-esteem 58 (27%) 14 (22%) 16 (13%) 13 (21%) 19 (16%) Other 82 (38%) 18 (28%) 28 (36%) 22 (36%) 25 (35%) Not answered 3 (1.4%) 0 0 1 (1.6%) 1 (1.3%)

 Getting information 42 (20%) 23 (36%) 24 (31%) 20 (32%) 27 (38%) Cognitive-behaviour 39 (19%) 10 (15%) 12 (15%) 12 (19%) 11 (16%) Other psychotherapy 33 (15%) 7 (11%) 6 (8%) 5 (8%) 4 (6%) Other 92 (42%) 24 (37%) 36 (46%) 34 (55%) 29 (40%) Not answered 11 (5%) 1 0 1 (1.6%) 1 (1.3%)

Vitamins/minerals 116 (54%) 36 (55%) 43 (55%) 40 (65%) 30 (42%) Anti-depressant 37 (17%) 14 (22%) 22 (28%) 12 (19%) 19 (26%) Herbal 29 (14%) 4 (6%) 4 (5%) 4 (7%) 12 (17%) Other 1 (0.4%) 0 2 (3%) 1 (1.6%) 0 Unsure/none 19 (9%) 7 (11%) 4 (5%) 2 (3%) 6 (8%) Not answered 14 (7%) 4 (6%) 3 (4%) 3 (5%) 5 (7%)

 Dietitian 51 (24%) 4 (6%) 14 (18%) 14 (23%) 22 (31%) Specialist 48 (22%) 13 (20%) 18 (23%) 12 (19%) 17 (24%) Non-specialist 30 (14%) 17 (26%) 13 (17%) 13 (21%) 12 (17%) Family doctor 32 (15%) 12 (19%) 7 (9%) 7 (11%) 7 (10%) Other 46 (21%) 16 (25%) 23 (30%) 15 (24%) 13 (18%) Not answered 10 (5%) 3 (5%) 3 (4%) 1 (1.6%) 1 (1.3%)

Table 1. **Bulimia nervosa mental health literacy (BN-MHL) outcomes following a BN-MHL intervention.** All data is in the form of n (%), I=information, ED=eating disorder, specialist refers to psychiatrist or psychologist, non-specialist refers to a counsellor or social

worker, all between group differences not significant.

**N** 217 65 78 62 72

Randomized Controlled Trials of a Mental Health Literacy and Self-Help Intervention 73

sensitivity analysis (to test for completer only analysis bias) was therefore not done. (Whilst on inspection it appeared that those in the intervention group were more likely to identify the problem as BN or another eating disorder ED at 6 and 12 months these differences were did not reach significance. There was a significant trend for those in the information only

Following presentation of the vignette, participants were asked: "What would you say is N's *main* problem?" They were required to choose one answer only from a list of options provided. Options, listed in a pre-determined, random order, were: "bulimia nervosa"; "anorexia nervosa"; "an eating disorder, but not anorexia or bulimia"; "yo-yo dieting"; "poor diet"; "low self-esteem*/*lack of self-confidence"; "depression"; "an anxiety disorder or problem"; "stress"; "a nervous breakdown"; "a mental health problem"; and "no real problem, just a phase." Participants were asked to indicate which of a number of possible interventions within each of three categories—people (15 options), treatments*/*activities (12 options), and medicines*/*pills (4 options)—they believed would be most helpful for N as well as the person that they would first approach for advice or help were they to have a problem such as the one described. At 6 and 12 months the name and age of the person in the vignette was changed but gender remained female and the symptom profile remained that of purging type BN.

Mental health related quality of life was assessed with the well-validated 12-item Short Form-12 Health Status Questionnaire (SF-12; Ware et al., 1996). This provides a mental health related component score presented in this chapter. A score below 50 indicates impairment and below 40 moderate to severe impairment. General psychiatric symptoms were assessed with the Kessler-10 item distress scale (K-10). It is designed to detect cases of anxiety and affective disorders in the general population (Andrews & Slade, 2001) and it has been used in our previous research (e.g. Mond et al., 2004b). Scores range from 10 to 50 as there are ten items scored from 1 to 5. Scores of 19 or above indicate likely psychiatric disorder such as major depression or an anxiety disorder. Body Mass Index (BMI; kg/m2) was calculated from self-reported height and weight.

Differences between groups were tested statistically using SPSS v 18 and with independent t-test and chi square or independent sample Mann-Whitney U tests respectively. Due to multiple testing significance was set at alpha < 0.01.

#### **2.3 Results of BN-MHL trial**

At baseline the participants' BN-MHL and ED symptoms did not differ between groups. Eighteen percent correctly identified the problem in the vignette as BN and the most common response (27%) response was that the person's problem was low self-esteem (Table 1). Regard for evidence based treatments or specialists was modest. Only one person at baseline, two at 6-months and five at 12-months thought a self-help treatment manual would be helpful.

ED symptoms were high with mean (SD) scores on the EDE-Q subscales of eating concern 2.4 (1.4), shape concern 4.2 (1.2), weight concern 3.8 (1.2), and restraint 3.0 (1.5). The majority (80%) were binge eating (objective and /or subjective type), 32 (15%) were vomiting for weight control, 30 (14%) were using laxatives and three (1.4%) had used diuretics in the past four weeks. Follow-up responses at 6 months were 66% and 62% at 12 months. There were no significant differences at baseline on outcome variables between those who were and were not followed to 12-months.

Further results and comparative findings of the groups randomised or not to the BN-MHL intervention over the 12–months are shown in Table 1 below. At follow-up there were no significant differences between the intervention and information-only groups in BN-MHL or in symptomatic outcomes or in mental health related quality of life (see Table 2). A

Following presentation of the vignette, participants were asked: "What would you say is N's *main* problem?" They were required to choose one answer only from a list of options provided. Options, listed in a pre-determined, random order, were: "bulimia nervosa"; "anorexia nervosa"; "an eating disorder, but not anorexia or bulimia"; "yo-yo dieting"; "poor diet"; "low self-esteem*/*lack of self-confidence"; "depression"; "an anxiety disorder or problem"; "stress"; "a nervous breakdown"; "a mental health problem"; and "no real problem, just a phase." Participants were asked to indicate which of a number of possible interventions within each of three categories—people (15 options), treatments*/*activities (12 options), and medicines*/*pills (4 options)—they believed would be most helpful for N as well as the person that they would first approach for advice or help were they to have a problem such as the one described. At 6 and 12 months the name and age of the person in the vignette was changed but gender remained female and the symptom profile remained that

Mental health related quality of life was assessed with the well-validated 12-item Short Form-12 Health Status Questionnaire (SF-12; Ware et al., 1996). This provides a mental health related component score presented in this chapter. A score below 50 indicates impairment and below 40 moderate to severe impairment. General psychiatric symptoms were assessed with the Kessler-10 item distress scale (K-10). It is designed to detect cases of anxiety and affective disorders in the general population (Andrews & Slade, 2001) and it has been used in our previous research (e.g. Mond et al., 2004b). Scores range from 10 to 50 as there are ten items scored from 1 to 5. Scores of 19 or above indicate likely psychiatric disorder such as major depression or an anxiety disorder. Body Mass Index (BMI; kg/m2)

Differences between groups were tested statistically using SPSS v 18 and with independent t-test and chi square or independent sample Mann-Whitney U tests respectively. Due to

At baseline the participants' BN-MHL and ED symptoms did not differ between groups. Eighteen percent correctly identified the problem in the vignette as BN and the most common response (27%) response was that the person's problem was low self-esteem (Table 1). Regard for evidence based treatments or specialists was modest. Only one person at baseline, two at 6-months and five at 12-months thought a self-help treatment manual

ED symptoms were high with mean (SD) scores on the EDE-Q subscales of eating concern 2.4 (1.4), shape concern 4.2 (1.2), weight concern 3.8 (1.2), and restraint 3.0 (1.5). The majority (80%) were binge eating (objective and /or subjective type), 32 (15%) were vomiting for weight control, 30 (14%) were using laxatives and three (1.4%) had used diuretics in the past four weeks. Follow-up responses at 6 months were 66% and 62% at 12 months. There were no significant differences at baseline on outcome variables between

Further results and comparative findings of the groups randomised or not to the BN-MHL intervention over the 12–months are shown in Table 1 below. At follow-up there were no significant differences between the intervention and information-only groups in BN-MHL or in symptomatic outcomes or in mental health related quality of life (see Table 2). A

of purging type BN.

was calculated from self-reported height and weight.

multiple testing significance was set at alpha < 0.01.

those who were and were not followed to 12-months.

**2.3 Results of BN-MHL trial** 

would be helpful.

sensitivity analysis (to test for completer only analysis bias) was therefore not done. (Whilst on inspection it appeared that those in the intervention group were more likely to identify the problem as BN or another eating disorder ED at 6 and 12 months these differences were did not reach significance. There was a significant trend for those in the information only group to have fewer subjective binges at 6-months.)


Table 1. **Bulimia nervosa mental health literacy (BN-MHL) outcomes following a BN-MHL intervention.** All data is in the form of n (%), I=information, ED=eating disorder, specialist refers to psychiatrist or psychologist, non-specialist refers to a counsellor or social worker, all between group differences not significant.

Targeted Prevention in Bulimic Eating Disorders:

services or treatments is thus less clear.

1993).

(EWD-HL).

a copy of the 'guide' booklet.

Randomized Controlled Trials of a Mental Health Literacy and Self-Help Intervention 75

Manuals studied have included: "*Overcoming Binge Eating*" (Fairburn, 1995) or translations/adaptations of it; the manual: "*Bulimia Nervosa: a guide to recovery*" (Cooper, 1995) since updated; and the manual: "*Getting Better Bit(e) by Bit(e)*" (Schmidt & Treasure

Hay et al. (2004) and Stefano et al. (2006) examined abstinence rates from ED behaviours such as binge eating in meta-analyses of trials pure self-help (PSH) vs waitlist in bulimic disorders such as BN or binge eating disorder. Rates ranged from 30% to 36% for PSH - and were better for GSH which ranged from 33% to 43%, the latter of which can be comparable to full CBT in its outcomes. In all meta-analyses PSH was however favoured over waitlist where abstinence rates were, for example, between 5% and 11%. Despite promising if modest findings, there have been a number of problems with these studies including variable levels of therapist training and variation in evaluation tools and outcome measurements. Whilst it has been argued that self-help can be a first step in management for selected people seeking help for EDs its role in assisting people with EDs not accessing

In addition, as weight concern and seeking help to lose weight is a common feature of women who do not seek help for their ED (Hay et al., 1998; Mond et al., 2007) we thought it important to add nutrition and lifestyle intervention strategies to self-help to assist women who are overweight or obese to reduce further weight gain and/or maintain weight in the healthy range. This included specific advice on healthy exercise. We also chose a vignette of someone with binge eating disorder as that is a common bulimic eating disorder and is more frequently associated with weight disorder (Hudson et al., 2007, Darby et al., 2009). We thus developed the intervention to be for both eating and weight disorder health literacy

We based this second trial in general practice as unrecognised bulimic eating disorders are common in women attending their family doctors (King, 1989; Whitehouse et al. 1992; Hay et al., 1998; Mond et al., 2009). The family doctor is also the point of access for psychological treatments for people in Australia. To inform the present study we conducted an investigation into the dissemination of an EWD-HL intervention into primary care at two general practices in late 2005 (Hay et al., 2006). One hundred and fifty-five women (aged 18- 45 years) attending the two practices (over 3 months) in North Queensland (Australia) were screened through the distribution of an ED symptom and an ED-MHL survey by reception staff. Fourteen (9%) had a bulimic ED, and a further 12 (7.7%) had clinically significant symptoms. Attractive booklets containing information about ED and their treatments, a brief assessment screening questionnaire for Eating Disorders (the SCOFF (Morgan et al., 1999)) and information on local services and consumer groups were left in the waiting rooms, and a poster containing the SCOFF questions was displayed inviting patients to take

This survey confirmed a high level of untreated bulimic EDs in primary care settings as of the 23% women who self-identified an ED problem only one had sought professional help, in this instance from a counselor. In addition, patients reported they were prompted to discuss their ED symptoms with their GP as a result of reading the booklet. However, screening utilising reception staff was problematic and very inefficient compared to our previous method of embedding a research assistant (RA) in the practices (e.g. Hay et al., 1998). We also found the booklets needed to be provided to participants directly as, while many participants (54%) were interested in receiving a copy of the booklet when their attention was drawn to it, very few (14%) had picked it up in the waiting rooms. This


Table 2. **Health outcomes following a mental health literacy intervention in women with disordered eating.** SF-12 MH (mental health component score) measures mental health related quality of life, the K-10 measures psychological distress, BMI=body mass index, mean and SD, all p not significant excepting \*p=0.01

## **2.4 Summary and introduction to trial of self-help approaches**

In the trial of BN-MHL intervention we found the participants' BN-MHL at baseline to be similar to that in our previous surveys (Mond et al., 2010). Participants were most likely to identify the problem for the women with BN as one of low self-esteem and had modest or low regard for evidence based or specialist therapies compared to non-specialists. As we found previously (Hay et al., 2007a) a BN-MHL intervention had no significant impact on changing attitudes or improving symptoms and in this study it also had no significant impact on improving mental health related quality of life.

The findings indicated that merely providing people with information about treatments for bulimic EDs and also advising them to seek help did not result in notable changes in behaviour or beliefs. Our question then was - what interventions might help people with EDs improve recognition and understanding of treatments for their problem and thereby prompt effective help-seeking? We thus planned a second feasibility trial to investigate the impact of enhancing the MHL intervention by adding an evidence-based self-help treatment manual to the MHL intervention.
