BRIEF REPORT
Eating Disorders: A Hidden Phenomenon in Outpatient Mental Health? Anthea Fursland, PhD1 Hunna J. Watson, PhD1,2,3,4*
ABSTRACT Background: Eating disorders are common but underdiagnosed illnesses. Helpseeking for co-occurring issues, such as anxiety and depression, are common.
of the sample met criteria for a DSM-IV eating disorder. Those scoring 2 on the SCOFF were more likely to: be female (p 5 0.001), younger (p 5 0.003), and have a history of self-harm (p < 0.001).
Objectives: To identify the prevalence of eating problems, using the SCOFF, and eating disorders when screening positive on the SCOFF (i.e., 2), among patients seeking help for anxiety and depression at a community-based mental health service.
Discussion: This study confirms that eating disorders are a hidden phenomenon in general outpatient mental health. By using a standardized diagnostic interview to establish diagnosis rather than self- or staff-report, the study builds on limited previous findings. The naturalistic study setting shows that screening for eating disorders can be easily built into routine intake practice, and successC 2013 fully identifies treatment need. V Wiley Periodicals, Inc.
Method: Patients (N 5 260) consecutively referred and assessed for anxiety and depression treatment were administered the SCOFF screening questionnaire and a semi-structured standardized diagnostic interview during routine intake. Results: 18.5% (48/260) scored 2 on the SCOFF, indicating eating problems. Of these, 41% (19/48) met criteria for an eating disorder. Thus, overall, 7.3% (19/260)
Introduction Community-based studies indicate a 15% lifetime prevalence of eating disorders in women1 and 10.5% of adolescents have an eating disorder.2 In “high-risk” populations such as outpatient mental health settings, the estimates of eating disorders are even higher, with point prevalence estimates of 16.2% for women and 2.0% for men using staff reports3 and 18.7% for women and 14.7% for men using self-reports.4 Yet, the majority of people with eating disorders go undetected. Most do not seek treatment,5,6 and Accepted 14 September 2013 *Correspondence to: Dr. Hunna Watson, Centre for Clinical Interventions, 223 James Street, Northbridge, Western Australia, Australia. E-mail: [email protected] 1 Department of Health in Western Australia, Centre for Clinical Interventions, Perth, Australia 2 Eating Disorders Program, Princess Margaret Hospital for Children, Perth, Australia 3 School of Paediatrics and Child Health, The University of Western Australia, Perth, Australia 4 School of Psychology and Speech Pathology, Curtin University, Perth, Australia Published online 17 October 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22205 C 2013 Wiley Periodicals, Inc. V
422
Keywords: anxiety; co-morbidity; depression; eating disorders; SCOFF; screening (Int J Eat Disord 2014; 47:422–425)
among those who do, the tendency is to present for help with dieting and weight loss or co-morbid mental or physical illness.7 Since the eating disorder is masked, there is a need to screen for eating disorders and the mental health clinician is a valuable gatekeeper.8 Screening for eating disorders enables improves prognosis and enables early detection.9,10 Early detection averts the physical and mental consequences of disease progression11 and ameliorates individual, family, and health care burdens.6 There is limited research on screening and point prevalence of eating problems in outpatient mental health populations, besides a program treating cooccurring substance abuse and PTSD,12 and two Norwegian4,13 studies. These studies relied on selfor staff-report and did neither use validated screening instruments nor clinical diagnostic interviews to establish diagnosis. This study will address these shortcomings by using a short, reliable and validated screening instrument to detect possible eating disorder, and a structured clinical diagnostic interview to establish diagnostic status among outpatients with anxiety and depression in an outpatient mental health International Journal of Eating Disorders 47:4 422–425 2014
EATING DISORDER SCREENING
setting. We predicted the point prevalence of eating disorders to be between the previous estimates of 4.7%13 and 17.2%.4 A second objective was to investigate the sociodemographic and clinical characteristics associated with positive screening i.e., “at-risk” for an eating disorder. We hypothesised that at-risk cases would be more likely to be female and younger, based on the epidemiological profile of these illnesses.
Method Participants Participants (N 5 260) were individuals with Diagnostic and Statistical Manual (DSM-IV)14 anxiety or depressive disorders consecutively referred to the Anxiety and Depression Program at a mental health outpatient clinic, the Centre for Clinical Interventions (CCI), in Western Australia. CCI is a free, state-wide service administered through the public health system and has two separate service streams of anxiety and mood disorders and eating disorders. Referrals were made by psychiatrists, primary care physicians, and clinical psychologists to the Anxiety and Depression Program of CCI rather than to the Eating Disorders Program, thus all participants were being assessed for inclusion in the treatments offered by the Anxiety and Depression team. Participants provided informed consent for de-identified data to be stored on a database and used for research. Procedure As part of routine clinic procedure for the Anxiety and Depression Program, each patient attended an intake assessment which included the MINI International Neuropsychiatric Interview (MINI PLUS Version 5.0)15 and the SCOFF.16 All participants who exceeded the threshold of 2 positive SCOFF responses were administered the MINI eating disorder modules to establish diagnostic status. Participants completed a sociodemographic questionnaire and various psychometric assessments. Measures Screening. The SCOFF contains five questions addressing features of eating disorders and takes a few minutes to complete and score. The SCOFF is valid17–19 and reliable.20 A threshold of 2 positive responses detected clinical eating disorders with 100% sensitivity (95% CI: 96.9–100%) and 87.5% specificity (95% CI: 79.2–93.4%).16 Diagnosis. Diagnoses were ascribed by Clinical Psychologists who administered the MINI,15 a standardized diagnostic interview with good concordance with International Journal of Eating Disorders 47:4 422–425 2014
TABLE 1. Sociodemographic and clinical characteristics of referrals to a specialist anxiety and depression clinical stream within a public mental health outpatient service Variable Sex (female), n (%) Age (years), M (SD) Employed, n (%) Married/de facto, n (%) Years of school, M (SD) Primary diagnosis, n (%) Major depression Social anxiety disorder Generalized anxiety disorder Panic disorder Dysthymia Bipolar Obsessive–compulsive disorder Post-traumatic stress disorder Other Problem medicated, n (%) Previous psychological treatment, n (%) Previous psychiatric hospitalization, n (%) History of self-harm, n (%) History of suicide attempt, n (%)
N
Sample (N 5 260)
260 260 259 258 246 260
156 (60) 34.89 (13.27) 139 (53.7) 85 (32.9) 11.39 (1.13)
256 259
112 (43.0) 80 (30.8) 32 (12.3) 10 (3.9) 7 (2.7) 4 (1.6) 3 (1.2) 2 (0.8) 10 (3.9) 163 (63.7) 232 (89.6)
255
70 (27.5)
240 240
53 (22) 62 (26)
established, lengthier interviews.21,22 The MINI has outstanding inter-rater reliability.21 Psychometric Assessments. The Beck Depression Inventory-2,23 the Beck Anxiety Inventory,24 the Rosenberg Self Esteem Scale25 and the Quality of Life Enjoyment and Satisfaction Questionnaire-Short Form26 were administered to measure depressive symptoms, anxiety symptoms, self-esteem, and quality of life, respectively. Statistical Analysis Analysis of variance (ANOVA) and the chi square test of independence were used to compare groups who screened positive or negative on the SCOFF on a priori specified sociodemographic and clinical features. When homogeneity of variance was violated in ANOVA, Welch’s F test was used. The Holm–Bonferroni procedure corrected for multiple testing.27
Results Characteristics of the sample are described in Table 1. Patients presented rather evenly with a primary anxiety disorder (49%) or depressive illness (46%). Some presented with other primary diagnoses (5%; e.g., bipolar disorder, hypochondriasis), but had secondary anxiety and depression. Of the sample, 18.5% (48/260) scored 2 on the SCOFF, indicating eating problems. Of these, 41% (19/48) met criteria for an eating disorder (Table 2) as yielded by the MINI. The point prevalence of DSM-IV eating disorders was 7.3% (19/260); 9.0% for females and 4.8% for males. One met criteria 423
FURSLAND AND WATSON TABLE 2. Comparison of patients referred to a specialist anxiety and depression treatment program in a communitybased mental health outpatient setting who screened positive and negative for eating problems on the SCOFF Characteristic Sex (female), n (%) Age (years), M (SD) Employed, n (%) Problem medicated, n (%) Previous psychiatric treatment, n (%) Previous psychiatric hospitalization, n (%) History of self-harm, n (%) History of suicide attempt, n (%) Primary anxiety disorder diagnosis, n (%) Primary depressive disorder diagnosis, n (%) Beck depression inventory-2, M (SD) Beck anxiety inventory, M (SD) Rosenberg self esteem scale, M (SD) Quality of life enjoyment and satisfaction Questionnaire – short form, M (SD) a
Sample (N 5 260)
Screened Negative (n 5 212)
Screened Positive (n 5 48)
p
260 260 259 256 259
117 (55.2) 35.96 (13.41) 112 (53.1) 133 (63.9) 194 (91.9)
39 (81.2) 30.17 (11.61) 27 (56.2) 30 (62.5) 38 (79.2)
0.001a 0.003a 0.69 0.85 0.009b
255
55 (26.6)
15 (31.2)
0.51
240 240
34 (17.4) 49 (25.0)
19 (42.2) 13 (29.5)
260
107 (50.5)
22 (45.8)
0.56
260
95 (44.8)
24 (50.0)
0.51
210
25.68 (10.92)
31.19 (12.78)
0.005b
204 203
18.26 (11.07) 22.36 (5.09)
25.12 (16.00) 20.05 (5.66)
0.01b 0.01b
250
46.58 (14.80)
42.83 (15.20)
0.12
Eating Disorders: A Hidden Phenomenon in Outpatient Mental Health? Anthea Fursland, PhD1 Hunna J. Watson, PhD1,2,3,4*
ABSTRACT Background: Eating disorders are common but underdiagnosed illnesses. Helpseeking for co-occurring issues, such as anxiety and depression, are common.
of the sample met criteria for a DSM-IV eating disorder. Those scoring 2 on the SCOFF were more likely to: be female (p 5 0.001), younger (p 5 0.003), and have a history of self-harm (p < 0.001).
Objectives: To identify the prevalence of eating problems, using the SCOFF, and eating disorders when screening positive on the SCOFF (i.e., 2), among patients seeking help for anxiety and depression at a community-based mental health service.
Discussion: This study confirms that eating disorders are a hidden phenomenon in general outpatient mental health. By using a standardized diagnostic interview to establish diagnosis rather than self- or staff-report, the study builds on limited previous findings. The naturalistic study setting shows that screening for eating disorders can be easily built into routine intake practice, and successC 2013 fully identifies treatment need. V Wiley Periodicals, Inc.
Method: Patients (N 5 260) consecutively referred and assessed for anxiety and depression treatment were administered the SCOFF screening questionnaire and a semi-structured standardized diagnostic interview during routine intake. Results: 18.5% (48/260) scored 2 on the SCOFF, indicating eating problems. Of these, 41% (19/48) met criteria for an eating disorder. Thus, overall, 7.3% (19/260)
Introduction Community-based studies indicate a 15% lifetime prevalence of eating disorders in women1 and 10.5% of adolescents have an eating disorder.2 In “high-risk” populations such as outpatient mental health settings, the estimates of eating disorders are even higher, with point prevalence estimates of 16.2% for women and 2.0% for men using staff reports3 and 18.7% for women and 14.7% for men using self-reports.4 Yet, the majority of people with eating disorders go undetected. Most do not seek treatment,5,6 and Accepted 14 September 2013 *Correspondence to: Dr. Hunna Watson, Centre for Clinical Interventions, 223 James Street, Northbridge, Western Australia, Australia. E-mail: [email protected] 1 Department of Health in Western Australia, Centre for Clinical Interventions, Perth, Australia 2 Eating Disorders Program, Princess Margaret Hospital for Children, Perth, Australia 3 School of Paediatrics and Child Health, The University of Western Australia, Perth, Australia 4 School of Psychology and Speech Pathology, Curtin University, Perth, Australia Published online 17 October 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22205 C 2013 Wiley Periodicals, Inc. V
422
Keywords: anxiety; co-morbidity; depression; eating disorders; SCOFF; screening (Int J Eat Disord 2014; 47:422–425)
among those who do, the tendency is to present for help with dieting and weight loss or co-morbid mental or physical illness.7 Since the eating disorder is masked, there is a need to screen for eating disorders and the mental health clinician is a valuable gatekeeper.8 Screening for eating disorders enables improves prognosis and enables early detection.9,10 Early detection averts the physical and mental consequences of disease progression11 and ameliorates individual, family, and health care burdens.6 There is limited research on screening and point prevalence of eating problems in outpatient mental health populations, besides a program treating cooccurring substance abuse and PTSD,12 and two Norwegian4,13 studies. These studies relied on selfor staff-report and did neither use validated screening instruments nor clinical diagnostic interviews to establish diagnosis. This study will address these shortcomings by using a short, reliable and validated screening instrument to detect possible eating disorder, and a structured clinical diagnostic interview to establish diagnostic status among outpatients with anxiety and depression in an outpatient mental health International Journal of Eating Disorders 47:4 422–425 2014
EATING DISORDER SCREENING
setting. We predicted the point prevalence of eating disorders to be between the previous estimates of 4.7%13 and 17.2%.4 A second objective was to investigate the sociodemographic and clinical characteristics associated with positive screening i.e., “at-risk” for an eating disorder. We hypothesised that at-risk cases would be more likely to be female and younger, based on the epidemiological profile of these illnesses.
Method Participants Participants (N 5 260) were individuals with Diagnostic and Statistical Manual (DSM-IV)14 anxiety or depressive disorders consecutively referred to the Anxiety and Depression Program at a mental health outpatient clinic, the Centre for Clinical Interventions (CCI), in Western Australia. CCI is a free, state-wide service administered through the public health system and has two separate service streams of anxiety and mood disorders and eating disorders. Referrals were made by psychiatrists, primary care physicians, and clinical psychologists to the Anxiety and Depression Program of CCI rather than to the Eating Disorders Program, thus all participants were being assessed for inclusion in the treatments offered by the Anxiety and Depression team. Participants provided informed consent for de-identified data to be stored on a database and used for research. Procedure As part of routine clinic procedure for the Anxiety and Depression Program, each patient attended an intake assessment which included the MINI International Neuropsychiatric Interview (MINI PLUS Version 5.0)15 and the SCOFF.16 All participants who exceeded the threshold of 2 positive SCOFF responses were administered the MINI eating disorder modules to establish diagnostic status. Participants completed a sociodemographic questionnaire and various psychometric assessments. Measures Screening. The SCOFF contains five questions addressing features of eating disorders and takes a few minutes to complete and score. The SCOFF is valid17–19 and reliable.20 A threshold of 2 positive responses detected clinical eating disorders with 100% sensitivity (95% CI: 96.9–100%) and 87.5% specificity (95% CI: 79.2–93.4%).16 Diagnosis. Diagnoses were ascribed by Clinical Psychologists who administered the MINI,15 a standardized diagnostic interview with good concordance with International Journal of Eating Disorders 47:4 422–425 2014
TABLE 1. Sociodemographic and clinical characteristics of referrals to a specialist anxiety and depression clinical stream within a public mental health outpatient service Variable Sex (female), n (%) Age (years), M (SD) Employed, n (%) Married/de facto, n (%) Years of school, M (SD) Primary diagnosis, n (%) Major depression Social anxiety disorder Generalized anxiety disorder Panic disorder Dysthymia Bipolar Obsessive–compulsive disorder Post-traumatic stress disorder Other Problem medicated, n (%) Previous psychological treatment, n (%) Previous psychiatric hospitalization, n (%) History of self-harm, n (%) History of suicide attempt, n (%)
N
Sample (N 5 260)
260 260 259 258 246 260
156 (60) 34.89 (13.27) 139 (53.7) 85 (32.9) 11.39 (1.13)
256 259
112 (43.0) 80 (30.8) 32 (12.3) 10 (3.9) 7 (2.7) 4 (1.6) 3 (1.2) 2 (0.8) 10 (3.9) 163 (63.7) 232 (89.6)
255
70 (27.5)
240 240
53 (22) 62 (26)
established, lengthier interviews.21,22 The MINI has outstanding inter-rater reliability.21 Psychometric Assessments. The Beck Depression Inventory-2,23 the Beck Anxiety Inventory,24 the Rosenberg Self Esteem Scale25 and the Quality of Life Enjoyment and Satisfaction Questionnaire-Short Form26 were administered to measure depressive symptoms, anxiety symptoms, self-esteem, and quality of life, respectively. Statistical Analysis Analysis of variance (ANOVA) and the chi square test of independence were used to compare groups who screened positive or negative on the SCOFF on a priori specified sociodemographic and clinical features. When homogeneity of variance was violated in ANOVA, Welch’s F test was used. The Holm–Bonferroni procedure corrected for multiple testing.27
Results Characteristics of the sample are described in Table 1. Patients presented rather evenly with a primary anxiety disorder (49%) or depressive illness (46%). Some presented with other primary diagnoses (5%; e.g., bipolar disorder, hypochondriasis), but had secondary anxiety and depression. Of the sample, 18.5% (48/260) scored 2 on the SCOFF, indicating eating problems. Of these, 41% (19/48) met criteria for an eating disorder (Table 2) as yielded by the MINI. The point prevalence of DSM-IV eating disorders was 7.3% (19/260); 9.0% for females and 4.8% for males. One met criteria 423
FURSLAND AND WATSON TABLE 2. Comparison of patients referred to a specialist anxiety and depression treatment program in a communitybased mental health outpatient setting who screened positive and negative for eating problems on the SCOFF Characteristic Sex (female), n (%) Age (years), M (SD) Employed, n (%) Problem medicated, n (%) Previous psychiatric treatment, n (%) Previous psychiatric hospitalization, n (%) History of self-harm, n (%) History of suicide attempt, n (%) Primary anxiety disorder diagnosis, n (%) Primary depressive disorder diagnosis, n (%) Beck depression inventory-2, M (SD) Beck anxiety inventory, M (SD) Rosenberg self esteem scale, M (SD) Quality of life enjoyment and satisfaction Questionnaire – short form, M (SD) a
Sample (N 5 260)
Screened Negative (n 5 212)
Screened Positive (n 5 48)
p
260 260 259 256 259
117 (55.2) 35.96 (13.41) 112 (53.1) 133 (63.9) 194 (91.9)
39 (81.2) 30.17 (11.61) 27 (56.2) 30 (62.5) 38 (79.2)
0.001a 0.003a 0.69 0.85 0.009b
255
55 (26.6)
15 (31.2)
0.51
240 240
34 (17.4) 49 (25.0)
19 (42.2) 13 (29.5)
260
107 (50.5)
22 (45.8)
0.56
260
95 (44.8)
24 (50.0)
0.51
210
25.68 (10.92)
31.19 (12.78)
0.005b
204 203
18.26 (11.07) 22.36 (5.09)
25.12 (16.00) 20.05 (5.66)
0.01b 0.01b
250
46.58 (14.80)
42.83 (15.20)
0.12
