EMPIRICAL ARTICLE
Mandating Weekly Weight Gain in a Day Treatment Program for Eating Disorders Lea Thaler, PhD1,2* Samantha Wilson, BA3 Jennifer S. Coelho, PhD4 Juliana Mazanek Antunes, MSc1 Mimi Israel, MD1,2 Howard Steiger, PhD1,2
ABSTRACT Objective: We examined differences in treatment outcome associated with a “self-directed”- versus “externalincentive”-based day treatment protocol for individuals with eating disorders (EDs) with below normal body mass index. The self-directed protocol recommended a weekly 500 g weight gain, but allowed participants freedom to gain weight at their own rate; the “external-incentive” protocol required a minimum weight gain of 500 g/week as a requirement for staying in the program. Method: Weight outcome was available for 49 individuals who were treated with the recommended weight gain protocol, and 40 individuals who were treated with the required weight gain protocol.
Introduction Intensive day-hospitalization has been shown to be an effective treatment for individuals with moderate to severe eating-disorder (ED) symptoms, with relevant studies reporting decreased eating pathology1–4 and comorbid psychopathology5–7 in day hospitalized adults and adolescents. Statistically significant increases in body mass index (BMI) have also been demonstrated in several day hospital treatments.2,3,8–10 Furthermore, available longitudinal research has suggested that treatmentrelated changes are maintained at follow-up, with weight gain obtained during day hospital treatment being maintained 6 months after treatment.4 Accepted 26 December 2013 *Correspondence to: Lea Thaler, Eating Disorders Program, Douglas University Institute in Mental Health, 6875 LaSalle Blvd, Montreal, Quebec, Canada H4H 1R3. E-mail: [email protected] and [email protected] 1 Eating Disorders Program, Douglas University Institute, Montreal, Quebec, Canada 2 Psychiatry Department, McGill University, Montreal, Quebec, Canada 3 Psychology Department, Universit e de Montr eal, Montreal, Quebec, Canada 4 Provincial Specialized Eating Disorders Program, BC Children’s Hospital, Vancouver, British Columbia, Canada Published online 15 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22246 C 2014 Wiley Periodicals, Inc. V
500
Results: Post-treatment weight was significantly higher in patients treated in the required weight gain protocol compared to those treated in the recommended weight gain protocol. All patients, regardless of which program they completed, showed improvements in the measures of eating-disorder and comorbid-psychiatric symptoms. Discussion: This study demonstrated the benefits of an alteration in protocol that required no additional financial or human resources. Further studies are needed to validate the effectiveness of various day treatment protocol parameC 2014 Wiley Periodicals, Inc. ters. V Keywords: day treatment; weight gain; eating disorders (Int J Eat Disord 2014; 47:500–506)
In treating eating disorders (EDs), day programs offer various clinical and economic advantages over inpatient programs: day treatments promote greater autonomy; they help patients maintain active social, occupational or academic roles5; patients return to their home environment on evenings and weekends, and therefore have more opportunities to integrate changes that they are making during the intensive treatment program11; and financial costs associated with day treatment are substantially lower than those of inpatient treatment. However, the increased time spent away from the supports of treatment in a day hospital program also allows patients more occasions to engage in disordered eating behaviors in their home environments (e.g., restriction, binge-eating, or compensatory behaviors; see Ref. 11). The available literature includes some studies that have explored the elements that may contribute to the effectiveness of day-hospital programs in promoting weight gain in underweight patients. Olmsted et al.7 found a lengthier (5 days/week) day program to be superior overall to a shorter (4 days/ week) program, but to yield comparable levels of weight gain in underweight patients. Motivation to change and initial BMI have been found to be predictors of weight maintenance following day International Journal of Eating Disorders 47:5 500–506 2014
MANDATING WEEKLY WEIGHT GAIN IN DAY TREATMENT
hospital treatment in adolescents with anorexia nervosa (AN).12 The present study evaluated the effectiveness of two types of day-treatment programs for underweight patients with EDs: one that emphasized self-directedness (in which weight gain was recommended but not mandated); the other applying external incentives (by mandating the attainment of weight-gain goals as a requirement for staying in the program). We expected the required weight gain protocol to result in larger weight gains, as well as in more improvement in eating and concurrent symptoms when compared to the recommended weight gain protocol.
Method Study Design The current study used a field-research and sequential-cohort design. The first cohort consisted of patients who participated in the Day Treatment Program (DTP) of the Douglas Institute Eating Disorders Program (EDP) from January 1, 2007, to August 31, 2009. During that time, the protocol of the DTP recommended, but did not require, weight gain for patients with a subnormal pretreatment BMI. On September 1, 2009, the DTP amended its protocol, introducing a minimum weight gain of 500 g/week (as per NICE guidelines13) for those beginning treatment with a BMI lower than 20.0. Failure to realize the weight-gain target on three consecutive or nonconsecutive weeks was established as a rationale for discharging the patient from the DTP, on the grounds that the individual involved was not yet ready to accept a mandatory weight gain goal. Both versions of the protocol were intended for a transdiagnostic sample of patients suffering from moderate-to-severe EDs and all patients, regardless of when they participated in the DTP, received the same assessment procedures and treatment. Participants All participants were recommended for day treatment at the EDP following a diagnostic assessment and a multidisciplinary treatment planning discussion. Assessments were performed by trained clinicians who used semistructured clinical interviews to draw Diagnostic and Statistical Manual of Mental Disorders14 ED diagnoses. To assess the impact of introducing mandatory weight gain as part of the treatment protocol, weight gain was used as the primary outcome measure for all patients with a BMI of 18.5 at admission. The cutoff of 18.5 was designed to allow for examination of weight gain in patients who had a significant amount of weight to be gained during treatment. For patients who had repeat admissions to the DTP during the study period, only data from their first admission were included. An International Journal of Eating Disorders 47:5 500–506 2014
additional inclusion criterion was completion of at least 6 weeks of treatment in the day program, to ensure that there was sufficient time to apply the consequences of the required weight gain protocol system, and to compare differences in retention rates in treatment across the two protocols. Ethical approval was obtained from the Research Ethics Board of the Douglas Institute, and all patients who chose to take part in the treatment outcome research provided written informed consent. From January 1, 2007, to August 31, 2009 (recommended weight gain protocol), data on 58 new admissions with a BMI of 18.5 or less were available. Of these individuals, 49 (84%) completed at least 6 weeks of treatment. Complete outcome data were available for all 49 participants. From September 1, 2009, to May 1, 2012 (required weight gain protocol), data on 49 new admissions with a BMI of 18.5 or less were available. Of these, 40 (82%) completed at least 6 weeks of treatment. Complete outcome data were available for 34 (85%) of the 40. Data were retained for patients in the required weight gain protocol who, after at least 6 weeks of treatment, were discharged due to failure to meet the required weigh gains. Therefore, data from these patient’s last week in the DTP were used as their post-treatment scores. Means, standard deviations (SDs), and ranges, for all patients with a BMI under 18.5 who completed treatment under the recommended and required protocols, are depicted in Table 1. Measures Eating Disorder Examination-Self-Report Questionnaire Version. The Eating Disorder Examination-Self-Report Questionnaire Version (EDE-Q)15 is a 41-item, self-report measure adapted from the Eating Disorder Examination interview.16 The EDE-Q assesses the key behavioral features and associated psychopathology of EDs and yields scores on four subscales: Restraint, Shape Concern, Weight Concern, and Eating Concern. The four subscales demonstrate excellent internal consistency and test– retest reliability.17 The EDE-Q was also used to assess frequency of binge eating and purging behaviors (vomiting, laxative, and diuretic use) in the past month. Cronbach’s a for our sample was .96. Center for Epidemiologic Studies Depression Scale. The Center for Epidemiologic Studies Depression Scale (CESD)18 is a 20-item self-report scale assessing depressive symptoms during the past week. Scores range from 0 to 60, with a cut-point of 16 used to indicate the likely presence of depression. The CES-D has been shown to demonstrate concurrent and discriminant validity and good reliability. In the current study the CES-D demonstrated good reliability (Cronbach’s a 5 .70). Behavior and Symptom Identification Scale. The Behavior and Symptom Identification Scale (BASIS-32)19
501
THALER ET AL. TABLE 1.
Demographic and clinical variables by protocol Recommended Weight Gain (n 5 49)
Age (years) Baseline body mass index (BMI; kg/m2) Treatment (number of weeks) Medication use Eating disorder diagnoses Anorexia nervosa (AN) subtypes Restrictive (AN-R) Binge and purge (AN-B/P) Bulimia nervosa (BN) Eating disorders not otherwise specified (EDNOS) Comorbid diagnoses Alcohol abuse or dependence Substance abuse or dependence Major depression Post-traumatic stress disorder (PTSD) Obsessive-compulsive disorder (OCD) Panic disorder Borderline personality disorder (BPD) Obsessive-compulsive personality disorder (OCPD)
Required Weight Gain (n 5 40)
Mean (SD), range 26.7 (9.0) 26.2 (8.4) 16.6 (1.4), 14.0–18.4 16.7 (1.3), 14.1–18.5 13.6 (3.1) 11.1 (3.9) n (%) 28 (60.9%) 24 (64.9%) 19 (38.8%) 21 (42.8%) 5 (10.2%) 4 (8.2%)
14 (35.0%) 18 (45.0%) 4 (10.0%) 4 (10.0%)
2 (4.1%) 5 (10.2%) 11 (22.4%) 1 (2.0%) 1 (2.0%) 5 (10.2%) 6 (12.2%) 7 (14.3%)
5 (12.8%) 2 (5.1%) 14 (35.9%) 0 (0%) 1 (2.6%) 2 (5.1%) 2 (5.1%) 2 (5.1%)
Significance
Effect size
ns ns p < .05a
d 5 0.06 d 5 0.14 d 5 0.69
ns ns ns ns ns ns ns ns ns ns ns ns ns ns
ns, nonsignificant. a The difference in number of weeks of treatment was significant across the weight protocol groups. This difference was carried by the fact that significantly more people had treatment terminated prematurely in the required weight gain protocol. After excluding individuals who were asked to leave treatment for failure to gain weight, and those who dropped out, there were no significant group differences in the number of weeks of treatment, with an average of 13.5 weeks of treatment (SD 5 3.2) for treatment completers in the required weight gain protocol and 12.6 weeks (SD 5 3.8) for those in the recommended weight gain protocol (t(72) 5 1.12, p 5 .26).
is a 32-item questionnaire that provides an overall assessment of psychiatric symptoms and functional abilities. It contains four subscales: anxiety/depression, problems in daily living, impulsivity, and problems in relationship to self and others. The questionnaire possesses good internal consistency. In the current sample, the scales demonstrated good reliability (Cronbach’s a 5 .65–.80). Autonomous and Controlled Motivations for Treatment Questionnaire. The Autonomous and Controlled Motivations for Treatment Questionnaire (ACMTQ)20 is a 12-item questionnaire with subscales measuring autonomous motivation (personally driven or internal intentions for change) and controlled motivation (externally driven intentions for change). For the current study, the word “depression” (used in a previous version of the scale in a study of depressed patients) was replaced with “ED.” A previous study from our group21 found the ACMTQ, modified for an ED population, to have good reliability (a 5 .78) for both the autonomous and controlled subscales. The ACMTQ asks respondents to rate why they participate in treatment for their ED. Sample items include: “I have chosen to make ED treatment an important part of my weekly life” (autonomous motivation) and “I would feel guilty if I didn’t do what my therapist said” (controlled motivation). In the present study, both subscales of the ACMTQ demonstrated good reliability (Cronbach’s a was .77 for the “controlled” subscale and .81 for the “autonomous” subscale).
502
Treatment The DTP of the Douglas Institute’s EDP provides intensive treatment for patients suffering from AN and bulimia nervosa in threshold and subthreshold forms. A maximum of 15 patients are enrolled in the DTP at a time. Admission criteria include a BMI of 14.0 or greater and medical stability. The DTP consists of two streams: a day program where patients are in treatment for 6 h/day, 4 days/week; and a day hospital for patients requiring more intensive treatment, running 10 h/day, 4 days/ week. Recommended treatment lengths are 16 weeks for underweight patients, 12 weeks for normal weight patients, and 8 weeks for patients participating in the DTP immediately after hospitalization in the EDP inpatient unit. The DTP is directed mainly at the normalization of eating, reduction in dietary restraint, elimination of binge eating and purging, and challenging distorted thoughts related to food, shape, and weight, primarily accomplished through group therapy and meal support.
Analyses Patients in the different protocols (recommended vs. required) were compared on pretreatment demographic variables (i.e., age and ED diagnosis) using t tests and chi square tests, as appropriate. To examine differences in post-treatment variables (i.e., BMI, binge eating, vomiting, ED, and comorbid symptoms measures), a series of repeated measures analysis of variances (ANOVAs) was International Journal of Eating Disorders 47:5 500–506 2014
MANDATING WEEKLY WEIGHT GAIN IN DAY TREATMENT
run, with assessment time (pre- vs. post-treatment) as the within subjects factor and protocol (recommended vs. required weight gain) as the between subjects factor. If significant main or interaction effects were observed, Bonferroni-corrected post hoc tests were conducted. Intent-to-treat (ITT) analyses were conducted on all participants with a BMI of 18.5 or less, regardless of number of weeks of treatment completed.
Results Treatment Dropouts and Manipulation Check
Of those who completed at least 6 weeks of treatment, one individual dropped out after 14 weeks of treatment under the recommended weight gain protocol, whereas no patients dropped out under the required weight gain protocol. These rates were not significantly different (p 5 1.00, Fisher’s exact test). To evaluate whether or not those in the required weight-gain protocol had a significantly higher rate of early treatment terminations (i.e., being asked to leave the program due to not meeting weight gain requirements) than did those in the recommended weight gain protocol, a chi-square analysis was done on the number of individuals who were asked to leave treatment. A total of 14 individuals (35%) under the required weight gain protocol had their treatment terminated early due to lack of weight gain, whereas no patients under the recommended weight gain protocol were asked to leave treatment early [v(1)2 5 20.35, p < .001]. Differences on Demographic and Treatment Characteristic Variables
Patients in the recommended and required protocols were compared on demographic and clinical variables (see Table 1). No differences were detected for psychoactive-medication use, comorbid diagnoses, age or BMI. No differences between the two protocols were seen for number of weeks spent in day program [t(81) 5 .97, p 5 .33] or day hospital [t(81) 5 .91, p 5.37]. There was a significant difference as to total number of weeks of treatment completed, due to the fact that significantly more people were required to terminate treatment in the required weight gain protocol versus the recommended protocol. After excluding individuals who were asked to leave treatment for failure to gain weight, and those who dropped out, there were no significant group differences as to number of weeks of treatment, with an average of 13.5 weeks of treatment (SD 5 3.2) for treatment completers in the required weight gain protocol International Journal of Eating Disorders 47:5 500–506 2014
versus 12.6 weeks (SD 5 3.8) for those in the recommended weight gain protocol [t(72) 5 1.12, p5.26]. Effect of Different Protocols on Weight Gain
A repeated measures ANOVA was conducted to compare participants’ BMIs at baseline and at post-treatment. There was a significant main effect of time, and a significant interaction between time (pre- vs. post-treatment) and protocol. Post hoc tests show that participants in both protocols gained weight from pre- to post-treatment (p values
Mandating Weekly Weight Gain in a Day Treatment Program for Eating Disorders Lea Thaler, PhD1,2* Samantha Wilson, BA3 Jennifer S. Coelho, PhD4 Juliana Mazanek Antunes, MSc1 Mimi Israel, MD1,2 Howard Steiger, PhD1,2
ABSTRACT Objective: We examined differences in treatment outcome associated with a “self-directed”- versus “externalincentive”-based day treatment protocol for individuals with eating disorders (EDs) with below normal body mass index. The self-directed protocol recommended a weekly 500 g weight gain, but allowed participants freedom to gain weight at their own rate; the “external-incentive” protocol required a minimum weight gain of 500 g/week as a requirement for staying in the program. Method: Weight outcome was available for 49 individuals who were treated with the recommended weight gain protocol, and 40 individuals who were treated with the required weight gain protocol.
Introduction Intensive day-hospitalization has been shown to be an effective treatment for individuals with moderate to severe eating-disorder (ED) symptoms, with relevant studies reporting decreased eating pathology1–4 and comorbid psychopathology5–7 in day hospitalized adults and adolescents. Statistically significant increases in body mass index (BMI) have also been demonstrated in several day hospital treatments.2,3,8–10 Furthermore, available longitudinal research has suggested that treatmentrelated changes are maintained at follow-up, with weight gain obtained during day hospital treatment being maintained 6 months after treatment.4 Accepted 26 December 2013 *Correspondence to: Lea Thaler, Eating Disorders Program, Douglas University Institute in Mental Health, 6875 LaSalle Blvd, Montreal, Quebec, Canada H4H 1R3. E-mail: [email protected] and [email protected] 1 Eating Disorders Program, Douglas University Institute, Montreal, Quebec, Canada 2 Psychiatry Department, McGill University, Montreal, Quebec, Canada 3 Psychology Department, Universit e de Montr eal, Montreal, Quebec, Canada 4 Provincial Specialized Eating Disorders Program, BC Children’s Hospital, Vancouver, British Columbia, Canada Published online 15 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22246 C 2014 Wiley Periodicals, Inc. V
500
Results: Post-treatment weight was significantly higher in patients treated in the required weight gain protocol compared to those treated in the recommended weight gain protocol. All patients, regardless of which program they completed, showed improvements in the measures of eating-disorder and comorbid-psychiatric symptoms. Discussion: This study demonstrated the benefits of an alteration in protocol that required no additional financial or human resources. Further studies are needed to validate the effectiveness of various day treatment protocol parameC 2014 Wiley Periodicals, Inc. ters. V Keywords: day treatment; weight gain; eating disorders (Int J Eat Disord 2014; 47:500–506)
In treating eating disorders (EDs), day programs offer various clinical and economic advantages over inpatient programs: day treatments promote greater autonomy; they help patients maintain active social, occupational or academic roles5; patients return to their home environment on evenings and weekends, and therefore have more opportunities to integrate changes that they are making during the intensive treatment program11; and financial costs associated with day treatment are substantially lower than those of inpatient treatment. However, the increased time spent away from the supports of treatment in a day hospital program also allows patients more occasions to engage in disordered eating behaviors in their home environments (e.g., restriction, binge-eating, or compensatory behaviors; see Ref. 11). The available literature includes some studies that have explored the elements that may contribute to the effectiveness of day-hospital programs in promoting weight gain in underweight patients. Olmsted et al.7 found a lengthier (5 days/week) day program to be superior overall to a shorter (4 days/ week) program, but to yield comparable levels of weight gain in underweight patients. Motivation to change and initial BMI have been found to be predictors of weight maintenance following day International Journal of Eating Disorders 47:5 500–506 2014
MANDATING WEEKLY WEIGHT GAIN IN DAY TREATMENT
hospital treatment in adolescents with anorexia nervosa (AN).12 The present study evaluated the effectiveness of two types of day-treatment programs for underweight patients with EDs: one that emphasized self-directedness (in which weight gain was recommended but not mandated); the other applying external incentives (by mandating the attainment of weight-gain goals as a requirement for staying in the program). We expected the required weight gain protocol to result in larger weight gains, as well as in more improvement in eating and concurrent symptoms when compared to the recommended weight gain protocol.
Method Study Design The current study used a field-research and sequential-cohort design. The first cohort consisted of patients who participated in the Day Treatment Program (DTP) of the Douglas Institute Eating Disorders Program (EDP) from January 1, 2007, to August 31, 2009. During that time, the protocol of the DTP recommended, but did not require, weight gain for patients with a subnormal pretreatment BMI. On September 1, 2009, the DTP amended its protocol, introducing a minimum weight gain of 500 g/week (as per NICE guidelines13) for those beginning treatment with a BMI lower than 20.0. Failure to realize the weight-gain target on three consecutive or nonconsecutive weeks was established as a rationale for discharging the patient from the DTP, on the grounds that the individual involved was not yet ready to accept a mandatory weight gain goal. Both versions of the protocol were intended for a transdiagnostic sample of patients suffering from moderate-to-severe EDs and all patients, regardless of when they participated in the DTP, received the same assessment procedures and treatment. Participants All participants were recommended for day treatment at the EDP following a diagnostic assessment and a multidisciplinary treatment planning discussion. Assessments were performed by trained clinicians who used semistructured clinical interviews to draw Diagnostic and Statistical Manual of Mental Disorders14 ED diagnoses. To assess the impact of introducing mandatory weight gain as part of the treatment protocol, weight gain was used as the primary outcome measure for all patients with a BMI of 18.5 at admission. The cutoff of 18.5 was designed to allow for examination of weight gain in patients who had a significant amount of weight to be gained during treatment. For patients who had repeat admissions to the DTP during the study period, only data from their first admission were included. An International Journal of Eating Disorders 47:5 500–506 2014
additional inclusion criterion was completion of at least 6 weeks of treatment in the day program, to ensure that there was sufficient time to apply the consequences of the required weight gain protocol system, and to compare differences in retention rates in treatment across the two protocols. Ethical approval was obtained from the Research Ethics Board of the Douglas Institute, and all patients who chose to take part in the treatment outcome research provided written informed consent. From January 1, 2007, to August 31, 2009 (recommended weight gain protocol), data on 58 new admissions with a BMI of 18.5 or less were available. Of these individuals, 49 (84%) completed at least 6 weeks of treatment. Complete outcome data were available for all 49 participants. From September 1, 2009, to May 1, 2012 (required weight gain protocol), data on 49 new admissions with a BMI of 18.5 or less were available. Of these, 40 (82%) completed at least 6 weeks of treatment. Complete outcome data were available for 34 (85%) of the 40. Data were retained for patients in the required weight gain protocol who, after at least 6 weeks of treatment, were discharged due to failure to meet the required weigh gains. Therefore, data from these patient’s last week in the DTP were used as their post-treatment scores. Means, standard deviations (SDs), and ranges, for all patients with a BMI under 18.5 who completed treatment under the recommended and required protocols, are depicted in Table 1. Measures Eating Disorder Examination-Self-Report Questionnaire Version. The Eating Disorder Examination-Self-Report Questionnaire Version (EDE-Q)15 is a 41-item, self-report measure adapted from the Eating Disorder Examination interview.16 The EDE-Q assesses the key behavioral features and associated psychopathology of EDs and yields scores on four subscales: Restraint, Shape Concern, Weight Concern, and Eating Concern. The four subscales demonstrate excellent internal consistency and test– retest reliability.17 The EDE-Q was also used to assess frequency of binge eating and purging behaviors (vomiting, laxative, and diuretic use) in the past month. Cronbach’s a for our sample was .96. Center for Epidemiologic Studies Depression Scale. The Center for Epidemiologic Studies Depression Scale (CESD)18 is a 20-item self-report scale assessing depressive symptoms during the past week. Scores range from 0 to 60, with a cut-point of 16 used to indicate the likely presence of depression. The CES-D has been shown to demonstrate concurrent and discriminant validity and good reliability. In the current study the CES-D demonstrated good reliability (Cronbach’s a 5 .70). Behavior and Symptom Identification Scale. The Behavior and Symptom Identification Scale (BASIS-32)19
501
THALER ET AL. TABLE 1.
Demographic and clinical variables by protocol Recommended Weight Gain (n 5 49)
Age (years) Baseline body mass index (BMI; kg/m2) Treatment (number of weeks) Medication use Eating disorder diagnoses Anorexia nervosa (AN) subtypes Restrictive (AN-R) Binge and purge (AN-B/P) Bulimia nervosa (BN) Eating disorders not otherwise specified (EDNOS) Comorbid diagnoses Alcohol abuse or dependence Substance abuse or dependence Major depression Post-traumatic stress disorder (PTSD) Obsessive-compulsive disorder (OCD) Panic disorder Borderline personality disorder (BPD) Obsessive-compulsive personality disorder (OCPD)
Required Weight Gain (n 5 40)
Mean (SD), range 26.7 (9.0) 26.2 (8.4) 16.6 (1.4), 14.0–18.4 16.7 (1.3), 14.1–18.5 13.6 (3.1) 11.1 (3.9) n (%) 28 (60.9%) 24 (64.9%) 19 (38.8%) 21 (42.8%) 5 (10.2%) 4 (8.2%)
14 (35.0%) 18 (45.0%) 4 (10.0%) 4 (10.0%)
2 (4.1%) 5 (10.2%) 11 (22.4%) 1 (2.0%) 1 (2.0%) 5 (10.2%) 6 (12.2%) 7 (14.3%)
5 (12.8%) 2 (5.1%) 14 (35.9%) 0 (0%) 1 (2.6%) 2 (5.1%) 2 (5.1%) 2 (5.1%)
Significance
Effect size
ns ns p < .05a
d 5 0.06 d 5 0.14 d 5 0.69
ns ns ns ns ns ns ns ns ns ns ns ns ns ns
ns, nonsignificant. a The difference in number of weeks of treatment was significant across the weight protocol groups. This difference was carried by the fact that significantly more people had treatment terminated prematurely in the required weight gain protocol. After excluding individuals who were asked to leave treatment for failure to gain weight, and those who dropped out, there were no significant group differences in the number of weeks of treatment, with an average of 13.5 weeks of treatment (SD 5 3.2) for treatment completers in the required weight gain protocol and 12.6 weeks (SD 5 3.8) for those in the recommended weight gain protocol (t(72) 5 1.12, p 5 .26).
is a 32-item questionnaire that provides an overall assessment of psychiatric symptoms and functional abilities. It contains four subscales: anxiety/depression, problems in daily living, impulsivity, and problems in relationship to self and others. The questionnaire possesses good internal consistency. In the current sample, the scales demonstrated good reliability (Cronbach’s a 5 .65–.80). Autonomous and Controlled Motivations for Treatment Questionnaire. The Autonomous and Controlled Motivations for Treatment Questionnaire (ACMTQ)20 is a 12-item questionnaire with subscales measuring autonomous motivation (personally driven or internal intentions for change) and controlled motivation (externally driven intentions for change). For the current study, the word “depression” (used in a previous version of the scale in a study of depressed patients) was replaced with “ED.” A previous study from our group21 found the ACMTQ, modified for an ED population, to have good reliability (a 5 .78) for both the autonomous and controlled subscales. The ACMTQ asks respondents to rate why they participate in treatment for their ED. Sample items include: “I have chosen to make ED treatment an important part of my weekly life” (autonomous motivation) and “I would feel guilty if I didn’t do what my therapist said” (controlled motivation). In the present study, both subscales of the ACMTQ demonstrated good reliability (Cronbach’s a was .77 for the “controlled” subscale and .81 for the “autonomous” subscale).
502
Treatment The DTP of the Douglas Institute’s EDP provides intensive treatment for patients suffering from AN and bulimia nervosa in threshold and subthreshold forms. A maximum of 15 patients are enrolled in the DTP at a time. Admission criteria include a BMI of 14.0 or greater and medical stability. The DTP consists of two streams: a day program where patients are in treatment for 6 h/day, 4 days/week; and a day hospital for patients requiring more intensive treatment, running 10 h/day, 4 days/ week. Recommended treatment lengths are 16 weeks for underweight patients, 12 weeks for normal weight patients, and 8 weeks for patients participating in the DTP immediately after hospitalization in the EDP inpatient unit. The DTP is directed mainly at the normalization of eating, reduction in dietary restraint, elimination of binge eating and purging, and challenging distorted thoughts related to food, shape, and weight, primarily accomplished through group therapy and meal support.
Analyses Patients in the different protocols (recommended vs. required) were compared on pretreatment demographic variables (i.e., age and ED diagnosis) using t tests and chi square tests, as appropriate. To examine differences in post-treatment variables (i.e., BMI, binge eating, vomiting, ED, and comorbid symptoms measures), a series of repeated measures analysis of variances (ANOVAs) was International Journal of Eating Disorders 47:5 500–506 2014
MANDATING WEEKLY WEIGHT GAIN IN DAY TREATMENT
run, with assessment time (pre- vs. post-treatment) as the within subjects factor and protocol (recommended vs. required weight gain) as the between subjects factor. If significant main or interaction effects were observed, Bonferroni-corrected post hoc tests were conducted. Intent-to-treat (ITT) analyses were conducted on all participants with a BMI of 18.5 or less, regardless of number of weeks of treatment completed.
Results Treatment Dropouts and Manipulation Check
Of those who completed at least 6 weeks of treatment, one individual dropped out after 14 weeks of treatment under the recommended weight gain protocol, whereas no patients dropped out under the required weight gain protocol. These rates were not significantly different (p 5 1.00, Fisher’s exact test). To evaluate whether or not those in the required weight-gain protocol had a significantly higher rate of early treatment terminations (i.e., being asked to leave the program due to not meeting weight gain requirements) than did those in the recommended weight gain protocol, a chi-square analysis was done on the number of individuals who were asked to leave treatment. A total of 14 individuals (35%) under the required weight gain protocol had their treatment terminated early due to lack of weight gain, whereas no patients under the recommended weight gain protocol were asked to leave treatment early [v(1)2 5 20.35, p < .001]. Differences on Demographic and Treatment Characteristic Variables
Patients in the recommended and required protocols were compared on demographic and clinical variables (see Table 1). No differences were detected for psychoactive-medication use, comorbid diagnoses, age or BMI. No differences between the two protocols were seen for number of weeks spent in day program [t(81) 5 .97, p 5 .33] or day hospital [t(81) 5 .91, p 5.37]. There was a significant difference as to total number of weeks of treatment completed, due to the fact that significantly more people were required to terminate treatment in the required weight gain protocol versus the recommended protocol. After excluding individuals who were asked to leave treatment for failure to gain weight, and those who dropped out, there were no significant group differences as to number of weeks of treatment, with an average of 13.5 weeks of treatment (SD 5 3.2) for treatment completers in the required weight gain protocol International Journal of Eating Disorders 47:5 500–506 2014
versus 12.6 weeks (SD 5 3.8) for those in the recommended weight gain protocol [t(72) 5 1.12, p5.26]. Effect of Different Protocols on Weight Gain
A repeated measures ANOVA was conducted to compare participants’ BMIs at baseline and at post-treatment. There was a significant main effect of time, and a significant interaction between time (pre- vs. post-treatment) and protocol. Post hoc tests show that participants in both protocols gained weight from pre- to post-treatment (p values
