BRIEF REPORT
Drive for Activity in Patients with Anorexia Nervosa Lot Sternheim, PhD1,2,3* Unna Danner, PhD1,3 Roger Adan, PhD1,3,4 Annemarie van Elburg, PhD1,2,3
ABSTRACT Objective: Hyperactivity and elevated physical activity are both considered symptom characteristics of anorexia nervosa (AN). It has been suggested that a drive for activity (DFA) may underlie these expressions, yet research into DFA in AN remains scant. This study investigated DFA levels in patients with AN and its relation to AN severity. Furthermore, as physical exercise may be a way to reduce negative affect, the influence of negative affect (anxiety) on the role of DFA in AN was tested. Method: Two hundred and forty female patients with AN completed measures for DFA, eating disorder (ED) pathology, anxiety, and clinical parameters.
Introduction Although elevated physical activity is a well-known feature in anorexia nervosa (AN), its causes and role in AN are still poorly understood. The bulk of physical activity research literature discusses biological approaches to this clinical phenomenon.1 For example, elevated physical activity may be secondary to AN pathology as a consequence of negative energy balance resulting in a “foraging” response to increase physical activity to find food.2 Evidence from the activity-based anorexia (ABA) rodent model supports this theory as food restriction results in increased running behavior when running wheels are available.2 Another factor that may be important for understanding physical activity is an individual’s motivation for physical activity. It has been suggested that physical activity may be a core feature of AN as a Accepted 2 March 2014 Additional Supporting Information may be found in the online version of this article. *Correspondence to: Lot Sternheim, Altrecht Eating Disorders Rintveld, The Netherlands. E-mail: [email protected] 1 Altrecht Eating Disorders Rintveld, The Netherlands 2 Clinical and Health Psychology, University Utrecht, Utrecht, The Netherlands 3 Utrecht Research Group Eating disorders (URGE), Utrecht, The Netherlands 4 Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, The Netherlands Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22272 C 2014 Wiley Periodicals, Inc. V
International Journal of Eating Disorders 00:00 00–00 2014
Results: A strong relation between DFA levels and ED pathology was found, which remained significant even after controlling for negative affect (anxiety). Discussion: After much theorizing about DFA in AN this study provides empirical evidence for DFA as a hallmark feature of AN, independent of anxiety levels. Future research should shed light on the relationships between DFA, actual physical C 2014 activity, and the course of AN. V Wiley Periodicals, Inc. Keywords: anorexia nervosa; drive for activity; eating disorders; negative affect; anxiety (Int J Eat Disord 2014; 00:000–000)
conscious attempt to lose weight. Furthermore, patients with AN report feeling committed, and obliged to exercise, whereby postponing exercise results in guilt.3–6 Moreover, physical activity relieves (negative) affective states, which may resemble the runners “high.”7–10 Interestingly, subjective reports by patients of elevated physical exercise often highlight an egodystonic and compulsive quality, indicating that they may not have their high levels of physical activity under control. This in turn suggests that physical activity is not merely an affect regulation strategy or conscious attempt to lose weight, but instead may derive from an urge to be physically active and an inability to sit still: a drive for activity (DFA). However, despite growing research into physical activity in AN, empirical evidence for the role of a DFA in the course of AN remains scant. This is remarkable considering that not only may a DFA underlie actual physical activity in AN, but moreover, a high DFA has been posited as a possibly important indicator for AN severity.11 A handful of studies has examined concepts closely related to DFA (e.g., motor or inner restlessness) in relation to leptin levels in AN,12–14 finding that lower leptin levels were associated to higher levels of restlessness. Only one study has explored DFA levels in individuals with eating disorders in relation to eating disorder pathology, and found that those with higher DFA levels were characterized by low BMI and chronic negative 1
STERNHEIM ET AL. TABLE 1. Mean (and SD) for demographics (age in years), clinical characteristics (BMI in kg/m2, eating disorders symptomatology (EDE scores), illness duration, age of onset of eating disorder, Drive for Activity (DFAQ) trait anxiety (STAI-T)
Age Age of onset ED Duration of illness BMI EDE scores DFAQ STAI-T
Mean (SD)
Ranges
21.6 (8.9) 15.7 (4.0) 5.5 (7.4) 15.8 (1.7) 3.6 (1.3) 39.21 (10.2) 52.6 (9.8)
11.2–62.9 12.5–42.0 0.2–43.8 11.6–18.5 0.4–6.0 17–60.0 29–76.0
measure (4-point Likert scale17), is based on questions used in previous studies assessing subjective experience of activity.12,13 In the current study, Cronbach’s a was 0.93. Trait anxiety was measured with the State-Trait Anxiety Inventory trait subscale,18,19 a well-validated self-report questionnaire. In the current study Cronbach’s a was 0.85. Statistical Analyses
affect.7 These findings suggest that DFA is particularly relevant to AN. In the current study, we expect to find that higher DFA levels are associated to more severe AN pathology. Furthermore, we expect this relationship to be present independently of negative affect (as measured by anxiety levels).
Visual examination of histograms and Kolmogorov– Smirnov (K-S) tests confirmed normal distribution of data. Pearson’s correlation coefficients were used to explore relationships between DFA levels and EDE scores. A hierarchical regression analysis was done to test DFA as predictor of eating pathology controlling for age, BMI, and anxiety. Step 1 thus included age, BMI, and anxiety, and Step 2 included DFA. Before regressions were run, independent variables were standardized.
Methods
Results
Participants
The sample consisted of 240 women with AN (ANR 5 145, ANBP 5 95; it was found that AN subtypes did not differ on any of the outcome measures). The mean age for the patients with AN was 21.6 years (SD 5 8.9) and mean BMI was 15.8 kg/ m2 (SD 5 1.7). The mean age of onset was 15.7 years (SD 5 4.0) and the mean duration of illness was 5.5 (SD 5 7.4) (71.4% had a illness duration equal to or less than 2 years) (see Table 1). Data were obtained during intake, after which some patients continued with in-patient treatment but the majority of patients continued with outpatient treatment. Whilst DSM-5 criteria were used for the purposes of this study, 41.9% of the patients fulfilled DSM-IV criteria for Eating Disorders Not Otherwise Specified subtype AN (EDNOS-AN). DFA levels were positively correlated to both EDE scores (r 5 .421, p < .001) and levels of anxiety (r 5 .420, p < .001). Anxiety levels were also positively correlated with EDE scores (r 5 .333, p < .012). BMI did not significantly correlate to any of the other factors (EDE scores, DFA, or anxiety). Seeing that there were (strong) correlations of EDE with DFA and anxiety levels, and there was a variation in age and BMI, it was essential to test whether DFA was predictive of EDE independent of these important factors (anxiety levels, age, and BMI). Hence, a hierarchical regression was conducted with EDE scores as dependent variable in which the first step included all control variables, and in the second step DFA was added. The
Participants were recruited via Altrecht Eating Disorders Rintveld, in the Netherlands and included if they had received a DSM-5 diagnosis of AN (restrictive or binge-purge subtype). Following DSM-5 criteria, patients had to have a body mass index (BMI) of 18.5 kg/m2 or lower. Eating disorder diagnoses were made by experienced clinicians (all medical doctors) and confirmed using the Eating Disorder Examination (EDE15), a semistructured interview widely used to assess psychopathology associated with eating disorders. Procedure This research project was reviewed and approved by the Medical Ethical Committee of the University Medical Centre Utrecht (UMCU), and the Committee Scientific Research of Altrecht Mental Health Institute. All participants provided informed consent according to the ethical standards laid down in the 2008 World Medical Association Declaration of Seoul.16 Assessments Participants’ age in years was recorded and all participants were weighed on a digital Tanita scale (Tanita Cooperation of America, Inc, Arlington Heights, IL), and their height was measured with a stadiometer in order to calculate BMI (kg/m2). Patients were weighed without their clothes. AN pathology was assessed with the EDE,15 a semi-structured interview widely used to assess psychopathology associated with eating disorders. The Drive for Activity Questionnaire (DFAQ), a 15 self-report
2
International Journal of Eating Disorders 00:00 00–00 2014
DRIVE FOR ACTIVITY TABLE 2. Hierarchical regression analyses relating eating disorder symptoms (EDE) to drive for activity (DFA), controlling for age, BMI, and anxiety
Step 1 Age BMI Anxiety Step 2 DFA
b
t
p
20.04 0.17 0.33
20.67 2.79 5.46
.50
Drive for Activity in Patients with Anorexia Nervosa Lot Sternheim, PhD1,2,3* Unna Danner, PhD1,3 Roger Adan, PhD1,3,4 Annemarie van Elburg, PhD1,2,3
ABSTRACT Objective: Hyperactivity and elevated physical activity are both considered symptom characteristics of anorexia nervosa (AN). It has been suggested that a drive for activity (DFA) may underlie these expressions, yet research into DFA in AN remains scant. This study investigated DFA levels in patients with AN and its relation to AN severity. Furthermore, as physical exercise may be a way to reduce negative affect, the influence of negative affect (anxiety) on the role of DFA in AN was tested. Method: Two hundred and forty female patients with AN completed measures for DFA, eating disorder (ED) pathology, anxiety, and clinical parameters.
Introduction Although elevated physical activity is a well-known feature in anorexia nervosa (AN), its causes and role in AN are still poorly understood. The bulk of physical activity research literature discusses biological approaches to this clinical phenomenon.1 For example, elevated physical activity may be secondary to AN pathology as a consequence of negative energy balance resulting in a “foraging” response to increase physical activity to find food.2 Evidence from the activity-based anorexia (ABA) rodent model supports this theory as food restriction results in increased running behavior when running wheels are available.2 Another factor that may be important for understanding physical activity is an individual’s motivation for physical activity. It has been suggested that physical activity may be a core feature of AN as a Accepted 2 March 2014 Additional Supporting Information may be found in the online version of this article. *Correspondence to: Lot Sternheim, Altrecht Eating Disorders Rintveld, The Netherlands. E-mail: [email protected] 1 Altrecht Eating Disorders Rintveld, The Netherlands 2 Clinical and Health Psychology, University Utrecht, Utrecht, The Netherlands 3 Utrecht Research Group Eating disorders (URGE), Utrecht, The Netherlands 4 Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, The Netherlands Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22272 C 2014 Wiley Periodicals, Inc. V
International Journal of Eating Disorders 00:00 00–00 2014
Results: A strong relation between DFA levels and ED pathology was found, which remained significant even after controlling for negative affect (anxiety). Discussion: After much theorizing about DFA in AN this study provides empirical evidence for DFA as a hallmark feature of AN, independent of anxiety levels. Future research should shed light on the relationships between DFA, actual physical C 2014 activity, and the course of AN. V Wiley Periodicals, Inc. Keywords: anorexia nervosa; drive for activity; eating disorders; negative affect; anxiety (Int J Eat Disord 2014; 00:000–000)
conscious attempt to lose weight. Furthermore, patients with AN report feeling committed, and obliged to exercise, whereby postponing exercise results in guilt.3–6 Moreover, physical activity relieves (negative) affective states, which may resemble the runners “high.”7–10 Interestingly, subjective reports by patients of elevated physical exercise often highlight an egodystonic and compulsive quality, indicating that they may not have their high levels of physical activity under control. This in turn suggests that physical activity is not merely an affect regulation strategy or conscious attempt to lose weight, but instead may derive from an urge to be physically active and an inability to sit still: a drive for activity (DFA). However, despite growing research into physical activity in AN, empirical evidence for the role of a DFA in the course of AN remains scant. This is remarkable considering that not only may a DFA underlie actual physical activity in AN, but moreover, a high DFA has been posited as a possibly important indicator for AN severity.11 A handful of studies has examined concepts closely related to DFA (e.g., motor or inner restlessness) in relation to leptin levels in AN,12–14 finding that lower leptin levels were associated to higher levels of restlessness. Only one study has explored DFA levels in individuals with eating disorders in relation to eating disorder pathology, and found that those with higher DFA levels were characterized by low BMI and chronic negative 1
STERNHEIM ET AL. TABLE 1. Mean (and SD) for demographics (age in years), clinical characteristics (BMI in kg/m2, eating disorders symptomatology (EDE scores), illness duration, age of onset of eating disorder, Drive for Activity (DFAQ) trait anxiety (STAI-T)
Age Age of onset ED Duration of illness BMI EDE scores DFAQ STAI-T
Mean (SD)
Ranges
21.6 (8.9) 15.7 (4.0) 5.5 (7.4) 15.8 (1.7) 3.6 (1.3) 39.21 (10.2) 52.6 (9.8)
11.2–62.9 12.5–42.0 0.2–43.8 11.6–18.5 0.4–6.0 17–60.0 29–76.0
measure (4-point Likert scale17), is based on questions used in previous studies assessing subjective experience of activity.12,13 In the current study, Cronbach’s a was 0.93. Trait anxiety was measured with the State-Trait Anxiety Inventory trait subscale,18,19 a well-validated self-report questionnaire. In the current study Cronbach’s a was 0.85. Statistical Analyses
affect.7 These findings suggest that DFA is particularly relevant to AN. In the current study, we expect to find that higher DFA levels are associated to more severe AN pathology. Furthermore, we expect this relationship to be present independently of negative affect (as measured by anxiety levels).
Visual examination of histograms and Kolmogorov– Smirnov (K-S) tests confirmed normal distribution of data. Pearson’s correlation coefficients were used to explore relationships between DFA levels and EDE scores. A hierarchical regression analysis was done to test DFA as predictor of eating pathology controlling for age, BMI, and anxiety. Step 1 thus included age, BMI, and anxiety, and Step 2 included DFA. Before regressions were run, independent variables were standardized.
Methods
Results
Participants
The sample consisted of 240 women with AN (ANR 5 145, ANBP 5 95; it was found that AN subtypes did not differ on any of the outcome measures). The mean age for the patients with AN was 21.6 years (SD 5 8.9) and mean BMI was 15.8 kg/ m2 (SD 5 1.7). The mean age of onset was 15.7 years (SD 5 4.0) and the mean duration of illness was 5.5 (SD 5 7.4) (71.4% had a illness duration equal to or less than 2 years) (see Table 1). Data were obtained during intake, after which some patients continued with in-patient treatment but the majority of patients continued with outpatient treatment. Whilst DSM-5 criteria were used for the purposes of this study, 41.9% of the patients fulfilled DSM-IV criteria for Eating Disorders Not Otherwise Specified subtype AN (EDNOS-AN). DFA levels were positively correlated to both EDE scores (r 5 .421, p < .001) and levels of anxiety (r 5 .420, p < .001). Anxiety levels were also positively correlated with EDE scores (r 5 .333, p < .012). BMI did not significantly correlate to any of the other factors (EDE scores, DFA, or anxiety). Seeing that there were (strong) correlations of EDE with DFA and anxiety levels, and there was a variation in age and BMI, it was essential to test whether DFA was predictive of EDE independent of these important factors (anxiety levels, age, and BMI). Hence, a hierarchical regression was conducted with EDE scores as dependent variable in which the first step included all control variables, and in the second step DFA was added. The
Participants were recruited via Altrecht Eating Disorders Rintveld, in the Netherlands and included if they had received a DSM-5 diagnosis of AN (restrictive or binge-purge subtype). Following DSM-5 criteria, patients had to have a body mass index (BMI) of 18.5 kg/m2 or lower. Eating disorder diagnoses were made by experienced clinicians (all medical doctors) and confirmed using the Eating Disorder Examination (EDE15), a semistructured interview widely used to assess psychopathology associated with eating disorders. Procedure This research project was reviewed and approved by the Medical Ethical Committee of the University Medical Centre Utrecht (UMCU), and the Committee Scientific Research of Altrecht Mental Health Institute. All participants provided informed consent according to the ethical standards laid down in the 2008 World Medical Association Declaration of Seoul.16 Assessments Participants’ age in years was recorded and all participants were weighed on a digital Tanita scale (Tanita Cooperation of America, Inc, Arlington Heights, IL), and their height was measured with a stadiometer in order to calculate BMI (kg/m2). Patients were weighed without their clothes. AN pathology was assessed with the EDE,15 a semi-structured interview widely used to assess psychopathology associated with eating disorders. The Drive for Activity Questionnaire (DFAQ), a 15 self-report
2
International Journal of Eating Disorders 00:00 00–00 2014
DRIVE FOR ACTIVITY TABLE 2. Hierarchical regression analyses relating eating disorder symptoms (EDE) to drive for activity (DFA), controlling for age, BMI, and anxiety
Step 1 Age BMI Anxiety Step 2 DFA
b
t
p
20.04 0.17 0.33
20.67 2.79 5.46
.50
