The efficacy of aerobic exercise and resistance training as

0 downloads 0 Views 470KB Size Report
Sep 23, 2017 - resistance training, have shown promise for social anxiety disorder, .... met a research assistant certified in personal training in a private gym.
Journal of Anxiety Disorders 52 (2017) 43–52

Contents lists available at ScienceDirect

Journal of Anxiety Disorders journal homepage: www.elsevier.com/locate/janxdis

The efficacy of aerobic exercise and resistance training as transdiagnostic interventions for anxiety-related disorders and constructs: A randomized controlled trial

MARK



Daniel M. LeBouthillier , Gordon J.G. Asmundson University of Regina, Canada

A R T I C L E I N F O

A B S T R A C T

Keywords: Randomized controlled trial Transdiagnostic Anxiety disorders Aerobic exercise Resistance training

Evidence supports exercise as an intervention for many mental health concerns; however, randomized controlled investigations of the efficacy of different exercise modalities and predictors of change are lacking. The purposes of the current trial were to: (1) quantify the effects of aerobic exercise and resistance training on anxiety-related disorder (including anxiety disorders, obsessive-compulsive disorder, and posttraumatic stress disorder) status, symptoms, and constructs, (2) evaluate whether both modalities of exercise were equivalent, and (3) to determine whether exercise enjoyment and physical fitness are associated with symptom reduction. A total of 48 individuals with anxiety-related disorders were randomized to aerobic exercise, resistance training, or a waitlist. Symptoms of anxiety-related disorders, related constructs, and exercise enjoyment were assessed at pre-intervention and weekly during the 4-week intervention. Participants were further assessed 1-week and 1-month post-intervention. Both exercise modalities were efficacious in improving disorder status. As well, aerobic exercise improved general psychological distress and anxiety, while resistance training improved disorder-specific symptoms, anxiety sensitivity, distress tolerance, and intolerance of uncertainty. Physical fitness predicted reductions in general psychological distress for both types of exercise and reductions in stress for aerobic exercise. Results highlight the efficacy of different exercise modalities in uniquely addressing anxiety-related disorder symptoms and constructs.

1. Introduction Anxiety-related disorders (including anxiety disorders, obsessive-compulsive disorder, and posttraumatic stress disorder) are highly prevalent and comorbid, affecting approximately 28.8% of individuals over their lifetime (Kessler et al., 2005). Although these disorders share superficial similarities, such as overlap in diagnostic criteria (Barlow, 2014), evidence suggests that they also share common etiology (Bienvenu, Hettema, Neale, Prescott, & Kendler, 2007; Brown & Naragon-Gainey, 2013; Carleton, 2016; Hettema, 2006), latent structure and higher dimensions (Brown, 2007; Paulus, Talkovsky, Heggeness, & Norton, 2015), and response to treatment (Hadjistavropoulos et al., 2014; Norton & Barrera, 2012). These shared factors have fostered the development of transdiagnostic treatments that can effectively target a relatively wide array of anxiety-related psychopathological concerns and that can be widely and efficiently distributed to the individuals in need of treatment. In recent years, there has been increasing empirical support for the role of exercise as a legitimate standalone or adjuvant treatment for a variety of mental health concerns (Asmundson et al., 2013). ⁎

Particularly, specific types of exercise, such as aerobic exercise and resistance training, have shown promise for social anxiety disorder, panic disorder, generalized anxiety disorder, obsessive-compulsive disorder, and posttraumatic stress disorder (Broocks et al., 1998; Fetzner & Asmundson, 2015; Herring, Jacob, Suveg, & O’Connor, 2011; Jazaieri, Goldin, Werner, Ziv, & Gross, 2012; Powers et al., 2015; Rector, Richter, Lerman, & Regev, 2015). Potential differences in the efficacy of distinct exercise paradigms have seldom been systemically evaluated using robust empirical methodology in the form of randomized controlled trials. The effects of exercise on constructs thought to underlie and maintain anxiety-related disorders have been studied to varied and limited extent, including anxiety sensitivity (i.e., fear of anxiety-related sensations based on negative or catastrophic beliefs regarding their consequences; Reiss & McNally, 1985; Wheaton, Deacon, McGrath, Berman, & Abramowitz, 2012), distress tolerance (i.e., the ability to tolerate emotional distress; Leyro, Zvolensky, & Bernstein, 2010; Simons & Gaher, 2005), and intolerance of uncertainty (i.e., the inability to withstand ambiguous or uncertain situations; Carleton, 2012;

Corresponding author at: Department of Psychology, University of Regina, Regina, Saskatchewan, S4S 0A2, Canada. E-mail address: [email protected] (D.M. LeBouthillier).

http://dx.doi.org/10.1016/j.janxdis.2017.09.005 Received 31 July 2017; Received in revised form 14 September 2017; Accepted 19 September 2017 Available online 23 September 2017 0887-6185/ © 2017 Elsevier Ltd. All rights reserved.

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

2.2. Measures

Laugesen, Dugas, & Bukowski, 2003; McEvoy & Mahoney, 2011). Aerobic exercise appears to reduce anxiety sensitivity (e.g., Fetzner & Asmundson, 2015; LeBouthillier & Asmundson, 2015; Smits et al., 2008), but only one study to date has examined the effects of resistance training (Broman-Fulks, Kelso, & Zawilinski, 2015). Limited evidence suggests that distress tolerance and intolerance of uncertainty are not amenable to change after acute exercise (Broman-Fulks et al., 2015; LeBouthillier & Asmundson, 2015), but the effect of longer exercise interventions remains untested. Greater research on the effects of exercise on these constructs could bolster exercise as a legitimate intervention for anxiety-related disorders and help elucidate mechanisms of action. Likewise, little is known regarding factors that may affect the efficacy of exercise for anxiety-related psychopathology. Evidence suggests a role of physical fitness in predicting the effectiveness of these interventions: While individuals with posttraumatic stress disorder generally benefit from aerobic exercise, those with lower cardiorespiratory fitness experience relatively greater reductions in symptoms (LeBouthillier & Asmundson, 2015). Sedentary individuals also benefit more from acute exercise than active individuals (Ensari, Greenlee, Motl, & Petruzzello, 2015). Enjoyment of exercise is another potentially important factor that is seldom investigated. Individuals have greater enhancements in mood states following their most preferred compared to their least preferred exercise modality (Lane, Jackson, & Terry, 2005) and engaging in exercise that increases self-efficacy appears to effect greater reductions in anxiety (Bodin & Martinsen, 2004). Despite the flourishing and diverse nature of research on the mental health benefits of exercise, integrative and controlled research on the relationship between exercise and anxiety-related disorders and constructs is lacking. Consequently, the purposes of the current trial were to: (1) quantify the effects of aerobic exercise and resistance training on anxiety-related disorder status, symptoms, and constructs, (2) evaluate whether both modalities of exercise are equivalent, and (3) to determine whether exercise enjoyment and physical fitness are associated with symptom reduction. We tested three hypotheses, namely that (1) aerobic exercise and resistance training would both be efficacious in & #8232;improving disorder status, disorder-specific symptoms, and related construct (i.e., general psychological distress, depression, anxiety, stress, anxiety sensitivity, distress tolerance, intolerance of uncertainty) when compared to a waitlist control; (2) the efficacy of both types of exercise would be generally equivalent in reducing disorderspecific symptoms and related constructs, and (3) greater enjoyment of the assigned exercise condition and lower fitness would predict the magnitude of reductions in disorder-specific symptoms and related constructs. The present trial also addressed limitations in current literature by using a transdiagnostic approach to study a mixed sample of individuals with anxiety-related disorders, quantifying changes in other important anxiety-related constructs, and utilizing a randomized controlled trial design.

2.2.1. Structured Clinical Interview for DSM-5, Research Version (SCID-5-RV; First, Williams, Karg, & Spitzer, 2015). The SCID-5-RV was used to screen for and establish diagnosis of anxiety-related disorders and to rule out symptoms of psychosis when suspected. Participants answered screening questions in a yes/no format. Disorderspecific sections of the SCID-5-RV were administered based on responses to the screening questions and disorder-specific symptom measures. The SCID-5-RV was also use to assess changes in disorder status (i.e., a change from meeting full criteria to no longer meeting full criteria for the primary disorder). 2.2.2. Disorder-specific outcome measures Disorder-specific measures were administered to participants based on endorsement of screening questions and outcome of the SCID-5-RV. Questionnaires included the Severity Measure for Specific Phobia–Adult (SMSP–A; Craske et al., 2013b), Social Interaction Phobia Scale (SIPS; Carleton et al., 2009), Panic Disorder Severity Scale–Self Report (PDSS–SR; Shear et al., 1997), Severity Measure for Agoraphobia–Adult (SMAA; Craske et al., 2013a), Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990), Obsessive-Compulsive Inventory–Revised (OCI–R; Foa et al., 2002), and Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5; Weathers et al., 2013). As described in the procedures below, participants who reported substantial symptoms at screening were invited to complete the SCID-5-RV to determine eligibility for the trial. Scores deemed to represent substantial symptoms were as follows: 20 (moderate) on the SMSP–A (Craske et al., 2013b); 21 on the SIPS (Carleton et al., 2009); 8 on the PDSS–SR (Shear et al., 2000); 20 (moderate) on the SMAA (Craske et al., 2013a); 58 on the PSWQ (Behar, Alcaine, Zuellig, & Borkovec, 2003); and 30 on the PCL-5 (Weathers et al., 2013). 2.2.3. Transdiagnostic outcome measures In addition to relevant disorder-specific measures, all participants completed transdiagnostic measures, including the Depression Anxiety Stress Scales–21 (DASS–21; Lovibond & Lovibond, 1995), Anxiety Sensitivity Index–3 (ASI–3; Taylor et al., 2007), Distress Tolerance Scale (DTS; Simons & Gaher, 2005), and Intolerance of Uncertainty Scale, Short Form (IUS–12; Carleton, Norton, & Asmundson, 2007). For screening purposes, a score of 5 on the DASS–21 Anxiety scale (Henry & Crawford, 2005) was deemed to represent substantial anxiety. 2.2.4. Exercise and fitness-related measures Physical Activity Readiness Questionnaire for Everyone (PAR-Q +; Warburton, Jamnik, Bredin, & Gledhill, 2014). The PAR-Q+ is a 16item self-report measure of potential barriers to safe engagement in physical activity. Items were answered in a yes/no format. The measure includes seven items relating to general health explicitly (e.g., heart condition, dizziness, joint problems). Individuals who endorse any of these items are directed to answer nine follow-up items relating to chronic health conditions (e.g., back problems, respiratory disease). Individuals who denied all screening items or all follow-up items were deemed to be able to physically engage in physical exercise. The PAR-Q + has high test-retest reliability, sensitivity, and specificity, and it is more robustly evidence-based than its predecessor, the PAR-Q (Bredin, Gledhill, Jamnik, & Warburton, 2013; Warburton, Bredin, Jamnik, & Gledhill, 2011). Six-Minute Walk Test (6MWT). The 6MWT is a standardized submaximal test of functional capacity used as a general measure of exercise capacity (Crapo et al., 2002). The test assesses the global and integrated functioning of pulmonary, cardiovascular, circulatory, and muscular systems. Scores represent the distance that participants can walk (in meters) in 6 min. Exercise enjoyment Participants rated enjoyment of exercise using the following statements created for the trial: (a) “I enjoy engaging in

2. Methods 2.1. Participants Ethical approval for the trial was obtained from the University of Regina Research Ethics Board. Data were from 48 community individuals diagnosed with an anxiety-related disorder (see below). Participants were eligible for the trial if they were between 18 and 65 years of age, resided in the Regina, Saskatchewan area, reported fewer than 150 min of moderate to vigorous exercise weekly, could safely engage in exercise, were not engaged in empirically supported therapy for anxiety (e.g., cognitive behavioural therapy), were not taking benzodiazepines or antipsychotic medication, were on at least 6 weeks of stable dosage of any other psychotropic medication, did not have symptoms of psychosis, were currently at low risk for suicide, and were not involved in litigation. 44

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

aerobic exercise, such as running, cycling, walking, or hiking”; (b) “I enjoy engaging in anaerobic exercise, such as lifting free weights or using weight machines.” Answers choices ranged from 0 (strongly disagree) to 10 (strongly agree). Session ratings Participants rated their sessions using the following statements: (a) “I have enjoyed this week's sessions”; (b) “I put a lot of effort during my sessions”; (c) “I found this week's sessions difficult”; (d) “This week's sessions were useful and it was worth coming”; and (e) “I am motivated to continue going with future sessions.” Items were rated on a Likert scale ranging from 0 (not at all) to 10 (very much).

2016), using the R programming language (R Core Team, 2017) and the following packages: boot (Canty & Ripley, 2016), Deducer (Fellows, 2012), lme4 (Bates, Mächler, Bolker, & Walker, 2015), lmerTest (Kuznetsova, Brockhoff, Christensen, & Bojesen, 2016), MuMIn (Bartoń, 2016), pbkrtest (Halekoh & Højsgaard, 2014), plotly (Sievert et al., 2017), and psych (Revelle, 2017). Chi-square tests were used to determine differences in disorder status at post-intervention. Two-level multilevel models were used to test the effects of exercise on continuous outcome variables. These models included measurement occasions (level 1) nested within individuals (level 2). Scores for disorder-specific outcome measures were converted to T scores based on means and standard deviations from published non-clinical samples for the SMSP–A (Lebeau et al., 2012), SIPS (Carleton et al., 2009), PDSS–SR (Newman, Holmes, Zuellig, Kachin, & Behar, 2006), SMAA (Lebeau et al., 2012), PSWQ (Behar et al., 2003), OCI–R (Foa et al., 2002), and PCL-5 (Blevins, Weathers, Davis, Witte, & Domino, 2015) and combined into a single variable for analysis. All models were computed using maximum likelihood estimation and included both fixed and random effects of intercept. Time was coded in 0.25-unit increments representing each week of the intervention, ranging 0 (baseline) to 1 (post-intervention). For the model testing T scores, a fixed effect of primary anxiety-related disorder was also included. Changes from post-intervention to follow-up were quantified when the efficacy of exercise from pre- to post-intervention was established. For pre- to post-intervention models, we computed two types of effect sizes. The variance accounted for by each model (R2) was estimated based on recommendations by Nakagawa and Schielzeth (2013). Additionally, standardized mean differences (Cohen’s d) for the exercise groups and the waitlist group were extracted by standardizing the outcome variable of each model. The analytic plan for equivalence testing in the multilevel models followed recommendations by Mascha and Sessler (Mascha & Sessler, 2011) and Greene and Morland (Greene, Morland, Durkalski, & Frueh, 2008). The equivalence criterion (δ) was set at 0.25, such that the efficacy of both types of exercise was deemed equivalent if the 90% confidence interval for the change over time in the resistance training group overlapped a value that was between 0.75 times (1 − δ) and 1.25 times (1 + δ) the magnitude of the effect for aerobic exercise. Equivalence tests were conducted only in cases where both aerobic exercise and resistance training were efficacious compared to the waitlist (Greene et al., 2008). To determine whether exercise enjoyment and physical fitness predicted the effectiveness of exercise, likelihood ratio tests (LRTs) were computed. These tests evaluated the addition of fixed 3-way interactions between each of these two predictors, condition, and time in models that included only the two exercise conditions. Non-significant LRTs suggested that the additional predictors did not substantially improve model fit and were not related to changes in the dependent variable. In multilevel models, 6MWT scores were expressed in 100 m units and centered to a mean of 0. Ancillary analyses were conducted as needed to determine the effectiveness of the exercise groups, plot and interpret complex interaction effects, and test variables of interest that were not included in primary analyses. Ancillary analyses were specified in the results if and when they were conducted. Quantitative ancillary analyses employed multilevel modelling as discussed above. All analyses were bootstrapped using 1000 replicates. For multilevel models, nonparametric percentile confidence intervals were computed based on the bootstrap. Null hypothesis testing involved two-tailed tests with an α of 0.05. In cases where a confidence interval and corresponding probability value conflicted (e.g., a 95% confidence interval overlapped zero but the corresponding p-value was less than 0.05), the nonparametric confidence interval was used to test the null hypothesis.

2.3. Procedure The trial was registered on the Open Science Framework (available at https://osf.io/dj3ba) in June 2016, after commencing data collection but prior to any statistical analyses.1 Data collection ended in January 2017. Interested individuals were directed to a secure website (Qualtrics.com) to complete eligibility questionnaires, including anxiety-related disorders screening questions, relevant disorder-specific outcome measures (according to their endorsement of screening questions), DASS–21, PAR-Q+, and exercise enjoyment questions. Those who met at least one cut score on a disorder-specific questionnaire or the DASS–21 Anxiety subscale and met other inclusion criteria completed relevant sections of the SCID-5-RV by phone. Subsequently, participants met a research assistant certified in personal training in a private gym to undergo the 6MWT. Participants completed baseline self-report questionnaires and were randomized to one of the experimental conditions. To prevent imbalanced randomization, the maximum number of participants randomized to each group was capped at 23. The trial used a partial cross-over design (Handley, Schillinger, Shiboski, Shihoski, & Shiboski, 2011; Lenhard et al., 2017), such that waitlisted individuals were re-randomized to one of the exercise conditions upon completion. Exercise was completed three times a week for 4 weeks under the guidance of a personal trainer. The aerobic exercise condition was similar to that used in previous literature (e.g., Broman-Fulks & Storey, 2008; Fetzner & Asmundson, 2015; LeBouthillier & Asmundson, 2015; Smits et al., 2008). Participants completed 40 min of aerobic exercise on a spin cycle at 60–80% age-adjusted maximum heart rate reserve. The resistance training condition was similar to that used in recent literature (Broman-Fulks et al., 2015), involving 2–3 sets of 10–12 repetitions of machine leg press, machine chest press, machine hamstring curl, dumbbell single arm row, machine shoulder press, machine triceps extension, and machine bicep curl. Exercise sessions began and ended with 5 min of stretching. Questionnaires were completed weekly (i.e., approximately every third session). Participants in the waitlist condition completed online questionnaires weekly from home for the duration of the trial phase. At 1-week follow-up, participants were contacted by phone to reverify their disorder status using relevant sections of the SCID-5-RV. They then met with the personal trainer in order to reassess their fitness level using the 6MWT as well as complete self-report questionnaires. Waitlisted participants were re-randomized to exercise at this point. At 1-month follow-up, participants were sent an email link to complete the same set of questionnaires administered during the 1-week follow-up. 2.4. Analyses Analyses were conducted with the RStudio software (RStudio Team, 1 Interim analyses were conducted during the data collection phase (after registration of the trial) to ensure that electronic questionnaires functioned correctly and were completed as expected, as well as to ensure data and data analytic syntax were ready in anticipation for the end of data collection. Additionally, interim analyses were conducted in November 2016 to present findings at the 37th Annual Conference of the Anxiety Disorders Association of America, San Francisco, CA.

45

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

that were more than 14 days late, totalling 66 observations. The remaining data included data from all randomized participants. Internal consistencies for outcome measures, as estimated by Cronbach’s α, were acceptable (0.77; SIPS) to excellent (0.93; PSWQ) in the waitlist group; acceptable (0.72; PDSS–SR) to excellent (0.95; DTS) in the aerobic exercise group; and good (0.80; DASS–21 Anxiety) to excellent (0.96; IUS–12) in the resistance training group.

Table 1 Demographics at Pre-Intervention by Trial Group. Waitlist

Aerobic exercise

Resistance training

M/n

SD/%

M/n

SD/%

M/n

SD/%

Age Body mass index

33.4 28.03

10.36 8.21

33.0 27.57

8.83 4.48

31.39 28.98

9.22 6.85

Sex Male Female

3 12

20 80

7 16

30 70

3 15

17 83

12

80

23

100

15 1

83 6

1 1 1

7 7 7

1

6

1

6

9 9

50 50

Ethnicity White Black/African Canadian Asian Indian Multiracial Other Relationship status Single Married or cohabiting Divorced Widowed Employment status Student (part time) Student (full time) Employed (part time) Employed (full time) Self-employed Unemployed On disability leave Primary disorder Panic disorder Social anxiety disorder Generalized anxiety disordera Posttraumatic stress disorder Secondary disorder Panic disorder Agoraphobia Social anxiety disorder Specific phobia Generalized anxiety disorder Obsessivecompulsive disorder

8 5

53 33

10 12

43 52

1 1

7 7

1

4

1 3 1

7 20 7

1 3 1

4 13 4

10

67

11 2 4 1

48 9 17 4

5 3

28 17

9

50

1

6

8

53

2 7

9 30

6

33

6

40

12

52

10

56

1

7

2

9

2

11

3

17

3

20

1 4

4 17

6

33

5

33

1 2

4 9

1 4

6 22

1

4

3.2. Disorder-specific outcomes 3.2.1. Disorder status One individual in the waitlist (9%) showed changes in disorder status, while 56% of individuals in the aerobic exercise and 100% of those in the resistance training group showed improvements in disorder status. Chi-square tests revealed overall differences in disorder status between groups (χ2 = 12.09, p = 0.001, Cramér’s V = 0.502). Compared to the waitlist, improvements in disorder status were significant in the aerobic exercise group (χ2 = 5.66, p = 0.032, Cramér’s V = 0.561) as well as in the resistance training group (χ2 = 12, p = 0.001, Cramér’s V = 0.926). 3.2.2. T scores For disorder-specific symptoms, the model explained a large proportion of variance (see Table 3). Results revealed no significant change in symptoms over time in the waitlist group or in the aerobic exercise group (Cohen’s d = 0.05, 95% CI [−0.26, 0.36]). Compared to the waitlist, resistance training was associated with a significant relative reduction of 7.6 T from pre- to post-intervention (Cohen’s d = -0.39, 95% CI [−0.72, −0.07]). Neither the addition of exercise enjoyment nor physical fitness significantly contributed to model fit (LRTs ≤ 4.03, ps ≥ 0.331), suggesting that they were unrelated to changes in symptoms during the intervention. For resistance training, scores were not significantly different at 1-week (b = 3.49, 95% CI [−13.47, 19.60], p = 0.650) or 1-month follow-up (b = −2.19, 95% CI [−12.82, 7.17], p = 0.658) compared to post-intervention. 3.3. Transdiagnostic outcomes 3.3.1. DASS–21 For general psychological distress, the model explained a large proportion of variance (see Table 4). Results suggested an increase in symptoms over time in the waitlist group of 6.63 points from pre- to post-intervention. Both the aerobic exercise and resistance training groups showed significant reductions relative to the waitlist of 12.84 points (Cohen’s d = −1.07, 95% CI [−1.72, −0.45]), and 12.00 points (Cohen’s d = −1.00, 95% CI [−1.74, −0.23]), respectively. Ancillary analyses testing whether the change in symptoms differed from zero (rather than from the relative change in the waitlist) revealed a significant change in the aerobic exercise group (b = −6.20, 95% CI [−11.21, −1.52], p = 0.012, Cohen’s d = −0.51, 95% CI [−0.93, −0.13]), but not in the resistance training group (b = −5.37, 95% CI [−12.06, 1.03], p = 0.100 Cohen’s d = −0.45, 95% CI [−1.00, 0.09]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. The addition of physical fitness significantly contributed to model fit (LRT = 9.51, p = 0.014); therefore, the model was refit, selecting only the exercise conditions to include the additional parameters. Results suggested that individuals with average fitness (i.e., a 6MWT score of approximately 609.93 m) on average experienced significant reduction in symptoms in the aerobic exercise group of 7.48 points over time (95% CI [−12.79, −2.37], p = 0.002), but there was no significant reduction in the resistance training group (b = −2.58, 95% CI [−9.19, 4.12], p = 0.446). For both groups, scores below the mean on the 6MWT were associated with greater reductions in symptoms (b = 5.80, 95% CI [0.33, 11.22], p = 0.032 for aerobic exercise; b = 12.39, 95% CI [2.43, 22.94], p = 0.016 for resistance training). Although this interaction

a One waitlist participant met criteria for other specified anxiety disorder: generalized anxiety disorder with physical symptoms not occurring more days than not.

3. Results 3.1. Descriptive statistics Participant demographics at pre-intervention by trial group are presented in Table 1. Scores for disorder-specific and transdiagnostic outcomes measures, as well as exercise and fitness-related measures at baseline are presented in Table 2. Individuals enrolled in the trial were primary female, White, single, married or cohabiting, and employed full time. In all trial groups, participants were most likely to meet criteria for social anxiety disorder or generalized anxiety disorder. An overview of participant flow is provided in Fig. 1. The majority of participants completed their assigned condition. Observations were removed for weekly surveys that were more than 4 days late, for 1-week follow-ups that were more than 7 days late, or for 1-month follow-ups 46

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

Table 2 Descriptive Statistics for Continuous Outcomes and Predictors at Pre-Intervention for Each Trial Group. Waitlist

T1 T2 DASS DASS-D DASS-A DASS-S ASI–3 IUS–12 DTS 6MWT AE enjoyment RT enjoyment

Aerobic exercise

Resistance training

n

M

SD

n

M

SD

n

M

SD

15 8 15 15 15 15 15 15 15 15 15 15

88.56 95.08 26.60 6.13 8.33 12.13 39.13 38.93 2.30 604.03 6.31 6.85

17.23 20.49 11.95 4.49 5.16 5.01 14.74 9.43 0.75 105.20 2.64 2.83

23 9 23 23 23 23 23 23 23 23 23 23

96.56 87.18 28.83 8.91 7.39 12.52 35.96 39.09 2.55 634.48 6.73 6.70

18.65 17.01 12.04 6.61 3.90 4.32 17.02 11.65 0.82 90.45 2.20 2.27

18 14 18 18 18 18 18 18 18 18 18 18

89.15 83.43 29.06 8.78 8.11 12.17 35.61 41.28 2.46 586.28 5.90 6.79

15.68 21.11 11.36 4.87 4.54 4.46 17.13 11.24 0.88 64.83 2.85 2.36

Note: T1 = primary disorder T score. T2 = secondary disorder T score. DASS = Depression Anxiety Stress Scales–21. D = depression. A = anxiety. S = stress. ASI–3 = Anxiety Sensitivity Index–3. IUS–12 = Intolerance of Uncertainty Scale–Short Form. DTS = Distress Tolerance Scale. 6MWT = Six-Minute Walk Test. AE = aerobic exercise. RT = resistance training.

group (b = −2.94, 95% CI [−5.99, 0.00], p = 0.050, Cohen’s d = −0.51, 95% CI [−1.04, 0.00]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. Neither the addition of exercise enjoyment nor physical fitness significantly contributed to model fit (LRTs ≤ 5.38, ps ≥ 0.209), suggesting that they were unrelated to changes in symptoms.

could also suggest that greater physical fitness was associated with an increase in symptoms over time, ancillary analyses involving visual inspection of three-dimensional plots of the data suggested that this was not the case. Scores were significantly lower in the aerobic exercise group at 1-week follow-up (b = −6.14, 95% CI [−11.17, −1.29], p = 0.018), but not at 1-month follow-up (b = −0.78, 95% CI [−5.70, 3.90], p = 0.722).

3.3.3. DASS–21 anxiety For anxiety, the model explained a large proportion of variance (see Table 4). Results suggested no change in symptoms over time in the waitlist group. Aerobic exercise was associated with a significant reduction relative to the waitlist of 4.75 points from pre- to post-intervention (Cohen’s d = −1.03, 95% CI [−1.70, −0.38]); however, no significant changes were observed for resistance training (Cohen's d = −0.52 95% CI [−1.29, 0.28]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. Neither the addition of exercise enjoyment nor physical fitness significantly contributed to model fit (LRTs ≤ 6.1, ps ≥ 0.076), suggesting that they were unrelated to changes in symptoms. For aerobic exercise, no significant changes compared to post-intervention were

3.3.2. DASS–21 depression For depression, the model explained a large proportion of variance (see Table 4). Results revealed a significant increase in symptoms in the waitlist group, corresponding to 3.98 points from pre- to post-intervention. Both aerobic exercise and resistance training were associated with a relative reductions of 4.95 points (Cohen’s d = −0.86, 95% CI [−1.49, −0.26]) and 6.92 points (Cohen’s d = −1.20, 95% CI [−1.92, −0.47]), respectively. Ancillary analyses testing whether the change in symptoms differed from zero (rather than from the relative change in the waitlist) revealed no significant changes in either the aerobic exercise group (b = −0.97, 95% CI [−3.26, 1.19], p = 0.444, Cohen’s d = −0.17, 95% CI [−0.57, 0.21]), or the resistance training

Fig. 1. Participant flow. A total of 48 individuals took part in the trial. Sample sizes for aerobic exercise and resistance training include the eight participants who completed the rerandomization process.

47

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

Table 3 Multilevel Model for T Scores from Pre- to Post-Intervention. Fixed effects

Intercept Group (AE) Group (RT) Time OCD PTSD PD SAD Group × time (AE) Group × time (RT)

Random effects

b

95% CI

p

b

95% CI

p

R2

102.48 1.08 −0.52 −1.68 4.12 −28.20 −21.82 −25.90 0.93 −7.60

[97.22, 107.45] [−2.84, 5.08] [−5.78, 4.82] [−6.09, 2.71] [−19.87, 30.70] [−40.53, −14.84] [−34.31, −10.86] [−32.32, −19.89] [−4.99, 6.96] [−13.89, −1.42]

< .001*** .602 .810 .444 .740 < .001*** < .001*** < .001*** .766 .012*

−21.82

[−34.31, −10.86]

< .001***

.901

Note: OCD = obsessive-compulsive disorder. PTSD = posttraumatic stress disorder. PD = panic disorder. SAD = social anxiety disorder. AE = aerobic exercise. RT = resistance training. Waitlist was used as the reference group. Generalized anxiety disorder was used as the reference disorder. Confidence intervals were bootstrapped using 1000 replicates. * p < 0.05. *** p < 0.001.

contributed to model fit (LRT = 9.01, p = 0.018); therefore, the model was refit, selecting only the exercise conditions to include the additional parameters. Results suggested that individuals with average fitness (i.e., a 6MWT score of approximately 609.93 m) on average experienced a significant reduction in stress in the aerobic exercise group (b = −2.68, 95% CI [−4.96, −0.47], p = 0.010), but not the resistance training group (b = −0.44. 95% CI [−3.27, 2.46], p = 0.768). The interaction between physical fitness and time was not significant in the resistance training group (b = 4.10, 95% CI [−0.16, 8.61], p = 0.064) but it was significant in the aerobic exercise group, such that scores below the mean for 6MWT were associated with

observed at 1-week (b = −1.70, 95% CI [−3.73, 0.19], p = 0.074) or 1-month follow-up (b = 0.89, 95% CI [−1.02, 2.66], p = 0.332). 3.3.4. DASS–21 stress For stress, the model explained a large proportion of variance (see Table 4). Results suggested no changes over time in any of the trial groups. Aerobic exercise was associated with a Cohen’s d of −0.72 (95% CI [−1.49, 0.03]), while resistance training was associated with a Cohen's d of −0.66 (95% CI [−1.54, 0.24]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. The addition of physical fitness significantly Table 4 Multilevel Model for DASS-21 Scores from Pre- to Post-Intervention. Fixed effects

Random effects

b

95% CI

p

b

95% CI

p

R2

Total Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

24.90 3.20 0.06 6.63 −12.84 −12.00

[20.32, 29.77] [−1.41, 8.32] [−6.37, 6.84] [1.08, 12.26] [−20.67, −5.38] [−20.92, −2.77]

< .001*** .192 .966 .026* < .001*** .004**

9.50

[6.99, 11.44]

< .001***

.714

Depression Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

5.89 2.59 1.74 3.98 −4.95 −6.92

[3.71, 8.20] [0.47, 4.95] [−1.28, 4.91] [1.39, 6.58] [−8.58, −1.52] [−11.02, −2.70]

< .001*** .022* .254 .004** .006** < .001***

4.69

[3.48, 5.63]

< .001***

.738

Anxiety Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

7.54 0.74 −0.63 1.38 −4.75 −2.41

[5.73, 9.45] [−1.09, 2.77] [−3.17, 2.05] [−0.83, 3.63] [−7.86, -1.77] [−5.95, 1.28]

< .001*** .444 .634 .226 < .001*** .182

3.65

[2.65, 4.40]

< .001***

.695

Stress Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

11.42 −0.05 −0.91 1.37 −3.20 −2.94

[9.55, 13.30] [−2.07, 2.10] [−3.56, 1.69] [−1.09, 3.85] [−6.59, 0.14] [−6.82, 1.08]

< .001*** .918 .496 .278 .062 .132

3.22

[2.22, 3.96]

< .001***

.594

Note: AE = aerobic exercise. RT = resistance training. Waitlist was used as the reference group. Confidence intervals were bootstrapped using 1000 replicates. * p < 0.05. ** p < 0.01. *** p < 0.001.

48

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

Table 5 Multilevel Model for ASI-3, DTS, and IUS-12 Scores from Pre- to Post-Intervention. Fixed effects

Random effects

b

95% CI

p

b

95% CI

p

R2

ASI-3 Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

29.97 8.01 5.11 3.62 −3.78 −13.10

[23.88, 36.15] [3.23, 13.39] [−2.53, 12.53] [−2.00, 9.34] [−11.76, 3.73] [−22.10, -4.24]

< .001*** .002** .180 .212 .338 .002**

15.45

[11.84, 18.36]

< .001***

.865

DTS Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

2.73 −0.30 −0.41 −0.10 0.07 0.68

[2.42, 3.04] [−0.55, -0.02] [−0.80, -0.03] [−0.39, 0.20] [−0.34, 0.46] [0.22, 1.14]

< .001*** .020* .042* .524 .740 < .001***

0.78

[0.59, 0.92]

< .001***

.856

IUS-12 Intercept Group (AE) Group (RT) Time Group × time (AE) Group × time (RT)

35.57 3.76 5.76 3.50 0.04 −7.06

[31.58, 39.62] [0.64, 7.27] [0.77, 10.61] [−0.17, 7.24] [−5.17, 4.95] [−12.94, −1.27]

< .001*** .022* .022* .064 .950 .018*

10.12

[7.76, 12.03]

< .001***

.863

associated with the waitlist or aerobic exercise (Cohen’s d = 0.00, 95% CI [−0.46, 0.44]); however, resistance training was associated with a reduction of 7.06 points compared to the waitlist (Cohen’s d = −0.63, 95% CI [−1.16, −0.11]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. Neither the addition of exercise enjoyment nor physical fitness significantly contributed to model fit (LRTs ≤ 3.44, ps ≥ 0.204), suggesting that they were unrelated to changes in symptoms. In the resistance training group, there were no changes compared to postintervention at 1-week (b = -5.42, 95% CI [−11.75, 0.64], p = 0.080) or 1-month follow-up (b = −4.55, 95% CI [−9.18, 0.33], p = 0.056).

reductions in stress (b = 2.89, 95% CI [0.54, 5.08], p = 0.004). Although this interaction could also suggest that greater physical fitness was associated with an increase in symptoms over time, ancillary analyses involving visual inspection of a three-dimensional plot of the data suggested that this was not the case. 3.3.5. ASI–3 For anxiety sensitivity, the model explained a large proportion of variance (see Table 5). Results revealed no changes associated with the waitlist or aerobic exercise (Cohen’s d = −0.23, 95% CI [−0.71, 0.23]); however, resistance training was associated with a relative decrease of 13.10 points compared to the waitlist (Cohen’s d = −0.79 (95% CI [−1.34, −0.26]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. Neither the addition of physical fitness nor exercise enjoyment significantly contributed to model fit (LRTs ≤ 5.12, ps ≥ 0.175), suggesting that they were unrelated to changes in symptoms. For resistance training, no significant changes compared to post-intervention were observed at 1-week (b = 5.24, 95% CI [−6.71, 16.57], p = 0.380) or 1-month follow-up (b = 3.25, 95% CI [−5.49, 12.54], p = 0.464).

3.4. Ancillary analyses To better understand differences in the efficacy of aerobic exercise and resistance training, ancillary analyses were conducted to test differences in session enjoyment, difficulty, and worth, as well as selfrated participant effort and motivation (n = 34). Compared to aerobic exercise, resistance training was associated with greater session enjoyment (b = 2.25, 95% CI [0.65, 3.77], p = 0.002), but not greater session difficulty (b = 1.18, 95% CI [−0.69, 2.97], p = 0.198). The resistance training group did not differ in terms of session worth (b = 1.34, 95% CI [−0.07, 2.70], p = 0.058), self-rated effort (b = 0.95, 95% CI [0.00, 1.86], p = 0.034), or motivation (b = 1.46, 95% CI [−0.03, 2.90], p = 0.044) when compared to aerobic exercise. Although probability values for the latter two results were below 0.05, confidence intervals overlapped zero, suggesting non-significant differences. Given that enjoyment did not predict the efficacy of either exercise type, a multilevel model was also computed to test whether exercise enjoyment during the trial predicted adherence. No significant differences in enjoyment of the assigned exercise were observed in those who completed all exercise sessions when compared to those who did not complete all sessions (b = −0.27, 95% CI [−1.85, 1.34], p = 0.748).

3.3.6. DTS For distress tolerance, the model explained a large proportion of variance (see Table 5). There were no changes over time in the waitlist or aerobic exercise groups (Cohen’s d = 0.07, 95% CI [−0.38, 0.51]); however, resistance training was associated with an increase in distress tolerance relative to the waitlist of 0.68 points compared to the waitlist (Cohen’s d = 0.76, 95% CI [0.24, 1.27]). Because the efficacy of both exercise groups could not be established, equivalency analyses were not conducted. Neither the addition of exercise enjoyment nor physical fitness significantly contributed to model fit (LRTs ≤ 4.16, ps ≥ 0.281), suggesting that they were unrelated to changes in symptoms. For the resistance training group, no significant changes compared to post-intervention were observed at 1-week (b = 0.37, 95% CI [−0.71, 1.40], p = 0.498) or 1-month follow-up (b = 0.28, 95% CI [−0.52, 1.12], p = 0.462).

4. Discussion The primary purpose of the present trial was to determine and compare the efficacy of two types of exercise in reducing anxiety-related disorder symptoms and constructs. Additionally, a secondary

3.3.7. IUS–12 For intolerance of uncertainty, the model explained a large proportion of variance (see Table 5). Results revealed no changes 49

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

sessions more highly than those in aerobic exercise. These perceptions of the session could include not only enjoyment of exercise, but factors such as working alliance with the trainer and satisfaction with the amenities used in the trial. Greater overall satisfaction with trial sessions may have facilitated greater overall improvements in anxiety-related symptoms and constructs. The causality of this relationship remains uncertain, as perceived improvements may also have facilitated satisfaction with the exercise sessions. This latter interpretation is consistent with other ancillary analyses revealing no differences between groups in terms of session difficulty, session worth, self-rated effort, and motivation. Moreover, research suggests that exercise that increases self-efficacy (e.g., martial arts) effects greater reductions in anxiety when compared to exercise that simply maintains it (e.g., cycling; Bodin & Martinsen, 2004). Due to the more varied nature of tasks involved in resistance training, this type of exercise may have been associated with greater self-efficacy and, thereby, improvements in a greater number of anxiety-related symptoms and constructs. The present trial was subject to several limitations. First, the size of the sample was relatively small. While results of the trial were promising, null findings should be interpreted with adequate caution and should be replicated in a larger sample. Second, follow-up was limited. As with previous research (Fetzner & Asmundson, 2015), attrition of trial participants at follow-up was greater than during the trial. As a consequence, estimates of symptoms after the intervention are less reliable. Future research should strive to understand and subsequently minimize attrition while respecting the autonomy of participants. Third, the trial sample was uniform, comprising mostly White females, ranging mostly between 20 and 40 years old, and who were students or working full time. Future research should strive to test the efficacy of exercise in other groups of individuals. Fourth, individuals in the present trial primarily presented with generalized anxiety or social anxiety disorder as their primary or secondary diagnosis. While all anxiety-related disorders of interest were represented in the present trial, future research should seek to replicate the findings in a sample with a more balanced proportion of disorders, as recent evidence suggests some anxiety-related disorders may not respond as well to exercise (see Jacquart et al., 2017). Fifth, while we were blind to participants’ group assignment at initial assessment of diagnostic status, re-assessment could not be completely blinded and could therefore have been prone to unintended bias. Nonetheless, self-report questionnaires were completed without involvement of the researchers. Notwithstanding these limitations, the unique patterns of improvements associated with aerobic exercise and resistance training in the present trial highlight the importance of continued research to elucidate the mechanisms of action in exercise-related interventions. Future research should explicitly and simultaneously investigate physiological and psychological mechanisms that may be responsible for the effects of exercise on an anxietyrelated disorders and constructs.

purpose of the present trial was to determine whether individual factors, namely physical fitness and enjoyment of exercise, were associated with greater benefits from exercise. Taken together, the primary findings of the present trial provide evidence for the role of exercise in reducing anxiety-related disorder symptoms and underlying constructs. With the exception of depression and stress, all constructs improved from pre- to post-intervention in at least one of the exercise groups. When comparing the magnitude of reduction in scores across constructs, exercise appeared to effect the largest improvements for anxiety, anxiety sensitivity, distress tolerance, and intolerance of uncertainty whereas significant, but more modest, improvements in disorder-specific symptoms were also observed. The magnitude of the improvements were in line with those reported in a recent meta-analysis of randomized controlled trials for exercise in individuals with anxiety and stress-related disorders (Stubbs et al., 2017). For the aerobic exercise group, results related to disorder status and disorder-specific symptoms appeared contradictory: Significant improvements were observed for disorder status, but not for disorderspecific symptoms. This difference is likely due to the dichotomous nature of the disorder status variable. The threshold for change (i.e., going from meeting full criteria to no longer meeting full criteria) may have been relatively low for individuals with milder symptom severity, thereby increasing the likelihood of significant findings. By contrast, significant improvement in disorder-specific symptoms likely required more substantial change in symptoms, regardless of participants’ symptom severity. From post-intervention to follow-up, no significant increases in scores were revealed for any of the constructs. For anxiety, aerobic exercise was associated with a further reduction in scores at 1-week follow-up, suggesting that changes associated with exercise may be maintained for some period of time after intervention; however, these findings should be interpreted with caution given the relatively higher rate of attrition at follow-up. As noted above, none of the constructs under study showed improvements that were simultaneously associated with both types of exercise. These results suggest that aerobic exercise and resistance training are not equivalent in reducing anxiety-related disorder symptoms and constructs. Physical fitness predicted reductions in scores during the intervention for general psychological distress and anxiety. In both cases, lower physical fitness predicted greater improvements. This is consistent with previous research on the role of physical fitness and physical inactivity in exercise interventions (Ensari et al., 2015; LeBouthillier et al., 2015). In contrast with previous research, however, physical fitness was not associated with other constructs, namely disorder-specific symptoms and anxiety sensitivity (LeBouthillier et al., 2015). Additionally, exercise enjoyment was not associated with changes in any of the constructs studied or with adherence in the trial. These findings contrast with research suggesting that individuals have greater enhancements in mood states following their most preferred compared to their least preferred exercise modality (Lane et al., 2005); however, given that the present trial investigated changes in negative affect, they are in line with research suggesting that these benefits are limited to positive mood states (Miller, Bartholomew, & Springer, 2005; Raedeke, 2007). While individual preferences for exercise modality will likely remain a clinical consideration if and when exercise is recommended for anxiety-related concerns, the results of the present trial suggest that the primary consideration ought to be selection of exercise based on the target constructs. Compared to aerobic exercise, resistance training was efficacious for a greater number of constructs. Aerobic exercise improved general psychological distress, anxiety, and stress; while resistance training improved disorder-specific symptoms, general psychological distress, anxiety sensitivity, distress tolerance, and intolerance of uncertainty. A potential explanation of these results may involve participants’ perceptions of their exercise sessions. Ancillary analyses revealed that those engaged in resistance training rated their enjoyment of the

5. Conclusions The present trial is the first to investigate the efficacy of different types of exercise on anxiety-related disorders and related constructs and using a randomized controlled design. Both aerobic exercise and resistance training appear to be efficacious in effecting improvements in anxiety-related disorder symptoms and related constructs. While exercise enjoyment does not appear to predict the effectiveness of exercise, lower physical fitness is associated with greater reductions in specific constructs, namely general psychological distress and anxiety. Knowledge regarding the role of physical fitness in predicting improvements in constructs such as general psychological distress and stress may allow clinicians to tailor interventions to idiosyncratic needs and preferences of those they treat. In turn, tailored exercise interventions may increase the likelihood that individuals will exercise consistently in order to achieve and maintain the greatest mental and physical health benefits. The results also have implications for 50

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

Carleton, R. N. (2012). The intolerance of uncertainty construct in the context of anxiety disorders: Theoretical and practical perspectives. Expert Review of Neurotherapeutics, 12(8), 937–947. http://dx.doi.org/10.1586/ern.12.82. Carleton, R. N. (2016). Into the unknown: A review and synthesis of contemporary models involving uncertainty. Journal of Anxiety Disorders, 39, 30–43. http://dx.doi. org/10.1016/j.janxdis.2016.02.007. Crapo, R. O., Casaburi, R., Coates, A. L., Enright, P. L., MacIntyre, N. R., McKay, R. T., ... Mottram, C. (2002). ATS statement: Guidelines for the Six-Minute Walk Test. American Journal of Respiratory and Critical Care Medicine, 166(1), 111–117. http:// dx.doi.org/10.1164/rccm.166/1/111. Craske, M., Wittchen, U., Bogels, S., Stein, M., Andrews, G., & Lebeu, R. (2013a). Severity Measure for Agoraphobia–Adult [Measurement instrument]. Retrieved from http:// www.psychiatry.org/practice/dsm/dsm5/online-assessment-measures. Craske, M., Wittchen, U., Bogels, S., Stein, M., Andrews, G., & Lebeu, R. (2013b). Severity Measure for Specific Phobia–Adult [Measurement instrument]. Retrieved from http:// www.psychiatry.org/practice/dsm/dsm5/online-assessment-measures. Ensari, I., Greenlee, T. A., Motl, R. W., & Petruzzello, S. J. (2015). Meta-analysis of acute exercise effects on state anxiety: An update of randomized controlled trials over the past 25 years. Depression and Anxiety, 32(8), 624–634. http://dx.doi.org/10.1002/da. 22370. Fellows, I. (2012). Deducer: A data analysis GUI for R. Journal of Statistical Software, 49(8), 1–15. http://dx.doi.org/10.18637/jss.v049.i08. Fetzner, M. G., & Asmundson, G. J. G. (2015). Aerobic exercise reduces symptoms of posttraumatic stress disorder: A randomized controlled trial. Cognitive Behaviour Therapy, 44(4), 301–313. http://dx.doi.org/10.1080/16506073.2014.916745. First, M., Williams, J., Karg, R., & Spitzer, R. (2015). Structured clinical interview for DSM5—Research version. Arlington, VA: American Psychiatric Association. Foa, E. B., Huppert, J. D., Leiberg, S., Langner, R., Kichic, R., Hajcak, G., & Salkovskis, P. M. (2002). The obsessive-complusive inventory: Development and validation of a short version. Psychological Assessment, 14(4), 485–495. http://dx.doi.org/10.1037// 1040-3590.14.4.485. Greene, C. J., Morland, L. A., Durkalski, V. L., & Frueh, B. C. (2008). Noninferiority and equivalence designs: Issues and implications for mental health research. Journal of Traumatic Stress, 21(5), 433–439. http://dx.doi.org/10.1002/jts.20367. Hadjistavropoulos, H. D., Pugh, N. E., Nugent, M. M., Hesser, H., Andersson, G., Ivanov, M., ... Austin, D. W. (2014). Therapist-assisted Internet-delivered cognitive behavior therapy for depression and anxiety: Translating evidence into clinical practice. Journal of Anxiety Disorders, 28(8), 884–893. http://dx.doi.org/10.1016/j.janxdis. 2014.09.018. Halekoh, U., & Højsgaard, S. (2014). A Kenward-Roger approximation and parametric bootstrap methods for tests in linear mixed models: The R package pbkrtest. Journal of Statistical Software, 59(9), 1–30. http://dx.doi.org/10.18637/jss.v059.i09. Handley, M. A., Schillinger, D., Shiboski, S., Shihoski, S., & Shiboski, S. (2011). Quasiexperimental designs in practice-based research settings: Design and implementation considerations. Journal of the American Board of Family Medicine, 24(5), 589–596. http://dx.doi.org/10.3122/jabfm.2011.05.110067. Henry, J. D., & Crawford, J. R. (2005). The short-form version of the Depression Anxiety Stress Scales (DASS-21): Construct validity and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 44(2), 227–239. http://dx.doi.org/10. 1348/014466505X29657. Herring, M. P., Jacob, M. L., Suveg, C., & O’Connor, P. J. (2011). Effects of short-term exercise training on signs and symptoms of generalized anxiety disorder. Mental Health and Physical Activity, 4(2), 71–77. http://dx.doi.org/10.1016/j.mhpa.2011.07. 002. Hettema, J. (2006). A population-based twin study of the relationship between neuroticism and internalizing disorders. American Journal of Psychiatry, 163(5), 857. http:// dx.doi.org/10.1176/appi.ajp.163.5.857. Jacquart, J., Roquet, F. B., Papini, S., Powers, M. B., Rosenfield, D., Smits, J. A. J., & Monfils, M.-H. (2017). Effects of acute exercise on fear extinction in rats and exposure therapy in humans: Null findings from five experiments. Journal of Anxiety Disorders, 50, 76–86. http://dx.doi.org/10.1016/j.janxdis.2017.05.010. Jazaieri, H., Goldin, P. R., Werner, K., Ziv, M., & Gross, J. J. (2012). A randomized trial of MBSR versus aerobic exercise for social anxiety disorder. Journal of Clinical Psychology, 68(7), 715–731. http://dx.doi.org/10.1002/jclp.21863. Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62(6), 593–602. http://dx.doi.org/10.1001/archpsyc.62.6.593. Kuznetsova, A., Brockhoff, P. B., Christensen, R., & Bojesen, H. (2016). lmerTest: Tests in linear mixed effects models [R package version 2.0-33]. Retrieved from https://cran.rproject.org/package=lmerTest. Lane, A. M., Jackson, A., & Terry, P. C. (2005). Preferred modality influences on exerciseinduced mood changes. Journal of Sports Science and Medicine, 4(2), 195–200. Laugesen, N., Dugas, M. J., & Bukowski, W. M. (2003). Understanding adolescent worry: The application of a cognitive model. Journal of Abnormal Child Psychology, 31(1), 55–64. http://dx.doi.org/10.1023/A:1021721332181. LeBouthillier, D. M., & Asmundson, G. J. G. (2015). A single bout of aerobic exercise reduces anxiety sensitivity but not intolerance of uncertainty or distress tolerance: A randomized controlled trial. Cognitive Behaviour Therapy, 44(4), 252–263. http://dx. doi.org/10.1080/16506073.2015.1028094. LeBouthillier, D. M., Fetzner, M. G., & Asmundson, G. J. G. (2015). Lower cardiorespiratory fitness is associated with greater reduction in PTSD symptoms and anxiety sensitivity following aerobic exercise. Mental Health and Physical Activity. http://dx. doi.org/10.1016/j.mhpa.2015.11.001. Lebeau, R. T., Glenn, D. E., Hanover, L. N., Beesdo-Baum, K., Wittchen, H.-U., & Craske, M. G. (2012). A dimensional approach to measuring anxiety for DSM-5. International

clinicians and public policy by providing evidence for different types of exercise as efficacious, cost-effective, and easily accessible interventions for anxiety-related psychopathology, thereby reducing the personal and economic burdens rooted in these disorders. Conflicts of interest The authors declare no conflicts of interest. Acknowledgments Daniel M. LeBouthillier is supported by a Canadian Institutes of Health Research Doctoral Research Award (FRN: 146270). Gordon J. G. Asmundson was supported, in part, by the University of Regina President’s Chair for Academic Excellence in Adult Mental Health Research. The authors would like to thank Dr. Darren Candow and the Aging Muscle and Bone Health Laboratory at the University of Regina for graciously providing exercise training amenities to participants during the trial. References Asmundson, G. J. G., Fetzner, M. G., Deboer, L. B., Powers, M. B., Otto, M. W., & Smits, J. A. J. (2013). Let's get physical: A contemporary review of the anxiolytic effects of exercise for anxiety and its disorders. Depression and Anxiety, 30(4), 362–373. http:// dx.doi.org/10.1002/da.22043. Barlow, D. H. (2014). Clinical handbook of psychological disorders: A step-by-step treatment manual (5th ed.). Guilford Press. Bartoń, K. (2016). MuMIn: Multi-model inference. [R package version 1.15.6]. Retrieved from https://cran.r-project.org/package=MuMIn. Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. http://dx.doi.org/10.18637/ jss.v067.i01. Behar, E., Alcaine, O., Zuellig, A. R., & Borkovec, T. D. (2003). Screening for generalized anxiety disorder using the Penn State Worry Questionnaire: A receiver operating characteristic analysis. Journal of Behavior Therapy and Experimental Psychiatry, 34(1), 25–43. http://dx.doi.org/10.1016/S0005-7916. Bienvenu, O. J., Hettema, J. M., Neale, M. C., Prescott, C. A., & Kendler, K. S. (2007). Low extraversion and high neuroticism as indices of genetic and environmental risk for social phobia, agoraphobia, and animal phobia. American Journal of Psychiatry, 164(11), 1714–1721. http://dx.doi.org/10.1176/appi.ajp.2007.06101667. Blevins, C. A., Weathers, F. W., Davis, M. T., Witte, T. K., & Domino, J. L. (2015). The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. Journal of Traumatic Stress, 28(6), 489–498. http://dx.doi. org/10.1002/jts.22059. Bodin, T., & Martinsen, E. (2004). Mood and self-efficacy during acute exercise in clinical depression. A randomized, controlled study. Journal of Sport & Exercise Psychology, 623–633. http://dx.doi.org/10.1123/jsep.26.4.623. Bredin, S. S. D., Gledhill, N., Jamnik, V. K., & Warburton, D. E. R. (2013). PAR-Q+ and ePARmed-X+: New risk stratification and physical activity clearance strategy for physicians and patients alike. Canadian Family Physician, 59(3), 273–277. Broman-Fulks, J. J., & Storey, K. M. (2008). Evaluation of a brief aerobic exercise intervention for high anxiety sensitivity. Anxiety, Stress, and Coping, 21(2), 117–128. http://dx.doi.org/10.1080/10615800701762675. Broman-Fulks, J. J., Kelso, K., & Zawilinski, L. (2015). Effects of a single bout of aerobic exercise versus resistance training on cognitive vulnerabilities for anxiety disorders. Cognitive Behaviour Therapy, 44(4), 240–251. http://dx.doi.org/10.1080/16506073. 2015.1020448. Broocks, A., Bandelow, B., Pekrun, G., George, A., Meyer, T., Bartmann, U., ... Rüther, E. (1998). Comparison of aerobic exercise, clomipramine, and placebo in the treatment of panic disorder. American Journal of Psychiatry, 155(5), 603–609. http://dx.doi.org/ 10.1176/ajp.155.5.603. Brown, T. A., & Naragon-Gainey, K. (2013). Evaluation of the unique and specific contributions of dimensions of the triple vulnerability model to the prediction of DSM-IV anxiety and mood disorder constructs. Behavior Therapy, 44(2), 277–292. http://dx. doi.org/10.1016/j.beth.2012.11.002. Brown, T. A. (2007). Temporal course and structural relationships among dimensions of temperament and DSM-IV anxiety and mood disorder constructs. Journal of Abnormal Psychology, 116(2), 313–328. http://dx.doi.org/10.1037/0021-843X.116.2.313. Canty, A., & Ripley, B. (2016). boot: Bootstrap R (S-Plus) functions [R package]. Retrieved from https://cran.r-project.org/package=boot. Carleton, R. N., Norton, P. J., & Asmundson, G. J. G. (2007). Fearing the unknown: A short version of the Intolerance of Uncertainty Scale. Journal of Anxiety Disorders, 21(1), 105–117. http://dx.doi.org/10.1016/j.janxdis.2006.03.014. Carleton, R. N., Collimore, K. C., Asmundson, G. J. G., McCabe, R. E., Rowa, K., & Antony, M. M. (2009). Refining and validating the social interaction anxiety scale and the social phobia scale. Depression and Anxiety, 26(2), 71–81. http://dx.doi.org/10.1002/ da.20480.

51

Journal of Anxiety Disorders 52 (2017) 43–52

D.M. LeBouthillier, G.J.G. Asmundson

Raedeke, T. D. (2007). The relationship between enjoyment and affective responses to exercise. Journal of Applied Sport Psychology, 19(1), 105–115. http://dx.doi.org/10. 1080/10413200601113638. Rector, N. A., Richter, M. A., Lerman, B., & Regev, R. (2015). A pilot test of the additive benefits of physical exercise to CBT for OCD. Cognitive Behaviour Therapy, 44(4), 328–340. http://dx.doi.org/10.1080/16506073.2015.1016448. Reiss, S., & McNally, R. J. (1985). Expectancy model of fear. In S. Reiss, & R. R. Bootzin (Eds.). Theoretical issues in behavior therapy (pp. 107–121). San Diego, CA: Academic Press. Revelle, W. (2017). psych: Procedures for personality and psychological research [R package version 1.7.3]. Retrieved from https://cran.r-project.org/package=psych. Shear, M. K., Brown, T. A., Barlow, D. H., Money, R., Sholomskas, D. E., Woods, S. W., ... Papp, L. A. (1997). Multicenter collaborative panic disorder severity scale. American Journal of Psychiatry, 154(11), 1571–1575. http://dx.doi.org/10.1176/ajp.154.11. 1571. Shear, M. K., Feske, U., Brown, C., Clark, D. B., Mammen, O., & Scotti, J. (2000). Anxiety disorders measures. Handbook of psychiatric measures. Washington, DC: American Psychiatric Association. Sievert, C., Parmer, C., Hocking, T., Chamberlain, S., Ram, K., Corvellec, M., & Despouy, P. (2017). plotly: Create interactive web graphics via plotly.js [R package version 4.6.0]. Retrieved from https://cran.r-project.org/package=plotly. Simons, J. S., & Gaher, R. M. (2005). The distress tolerance scale: Development and validation of a self-report measure. Motivation and Emotion, 29(2), 83–102. http://dx. doi.org/10.1007/s11031-005-7955-3. Smits, J. A. J., Berry, A. C., Rosenfield, D., Powers, M. B., Behar, E., & Otto, M. W. (2008). Reducing anxiety sensitivity with exercise. Depression and Anxiety, 25(8), 689–699. http://dx.doi.org/10.1002/da.20411. Stubbs, B., Vancampfort, D., Rosenbaum, S., Firth, J., Cosco, T., Veronese, N., ... Schuch, F. B. (2017). An examination of the anxiolytic effects of exercise for people with anxiety and stress-related disorders: A meta-analysis. Psychiatry Research, 249(May), 102–108. http://dx.doi.org/10.1016/j.psychres.2016.12.020. Taylor, S., Zvolensky, M. J., Cox, B. J., Deacon, B., Heimberg, R. G., Ledley, D. R., ... Cardenas, S. J. (2007). Robust dimensions of anxiety sensitivity: Development and initial validation of the anxiety sensitivity index-3. Psychological Assessment, 19(2), 176–188. http://dx.doi.org/10.1037/1040-3590.19.2.176. Warburton, D. E. R., Bredin, S. S. D., Jamnik, V. K., & Gledhill, N. (2011). Validation of the PAR-Q+ and ePARmed-X+. Health & Fitness Journal of Canada, 4(2), 38–46. Warburton, D. E. R., Jamnik, V., Bredin, S. S. D., & Gledhill, N. (2014). The 2014 physical activity readiness questionnaire for everyone (PAR-Q+) and electronic physical activity readiness medical examination (ePARmed-X+). Health & Fitness Journal of Canada, 7(1), 80–83. Retrieved from http://www.healthandfitnessjournalofcanada. com/index.php/html/article/view/171. Weathers, F. W., Litz, B. T., Keane, T. M., Palmieri, P. A., Marx, B. P., & Schnurr, P. P. (2013). The PTSD Checklist for DSM-5 (PCL-5). Retrieved from www.ptsd.va.gov. Wheaton, M. G., Deacon, B. J., McGrath, P. B., Berman, N. C., & Abramowitz, J. S. (2012). Dimensions of anxiety sensitivity in the anxiety disorders: Evaluation of the ASI-3. Journal of Anxiety Disorders, 26(3), 401–408. http://dx.doi.org/10.1016/j.janxdis. 2012.01.002.

Journal of Methods in Psychiatric Research, 21(4), 258–272. http://dx.doi.org/10. 1002/mpr.1369. Lenhard, F., Andersson, E., Mataix-Cols, D., Rück, C., Vigerland, S., Högström, J., ... Serlachius, E. (2017). Therapist-guided, internet-delivered cognitive-behavioral therapy for adolescents with obsessive-compulsive disorder: A randomized controlled trial. Journal of the American Academy of Child & Adolescent Psychiatry, 56(1), 10–19 e2. Leyro, T. M., Zvolensky, M. J., & Bernstein, A. (2010). Distress tolerance and psychopathological symptoms and disorders: A review of the empirical literature among adults. Psychological Bulletin, 136(4), 576–600. http://dx.doi.org/10.1037/ a0019712. Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the depression anxiety & stress scales (2nd ed.). Sydney, Australia: Psychology Foundation. Mascha, E. J., & Sessler, D. I. (2011). Equivalence and noninferiority testing in regression models and repeated-measures designs. Anesthesia and Analgesia, 112(3), 678–687. http://dx.doi.org/10.1213/ANE.0b013e318206f872. McEvoy, P. M., & Mahoney, A. E. J. (2011). Achieving certainty about the structure of intolerance of uncertainty in a treatment-seeking sample with anxiety and depression. Journal of Anxiety Disorders, 25, 112–122. http://dx.doi.org/10.1016/j.janxdis. 2010.08.010. Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Development and validation of the Penn State worry questionnaire. Behaviour Research and Therapy, 28(6), 487–495. http://dx.doi.org/10.1016/0005-7967. Miller, B. M., Bartholomew, J. B., & Springer, B. A. (2005). Post-exercise affect: The effect of mode preference. Journal of Applied Sport Psychology, 17(4), 263–272. http://dx. doi.org/10.1080/10413200500313503. Nakagawa, S., & Schielzeth, H. (2013). A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution, 4(2), 133–142. http://dx.doi.org/10.1111/j.2041-210x.2012.00261.x. Newman, M. G., Holmes, M., Zuellig, A. R., Kachin, K. E., & Behar, E. (2006). The reliability and validity of the Panic Disorder Self-Report: A new diagnostic screening measure of panic disorder. Psychological Assessment, 18(1), 49–61. http://dx.doi.org/ 10.1037/1040-3590.18.1.49. Norton, P. J., & Barrera, T. L. (2012). Transdiagnostic versus diagnosis-specific CBT for anxiety disorders: A preliminary randomized controlled noninferiority trial. Depression and Anxiety, 29(10), 874–882. http://dx.doi.org/10.1002/da.21974. Paulus, D. J., Talkovsky, A. M., Heggeness, L. F., & Norton, P. J. (2015). Beyond negative affectivity: A hierarchical model of global and transdiagnostic vulnerabilities for emotional disorders. Cognitive Behaviour Therapy, 6073(October), 1–17. http://dx. doi.org/10.1080/16506073.2015.1017529. Powers, M. B., Medina, J. L., Burns, S., Kauffman, B. Y., Monfils, M., Asmundson, G. J. G., ... Smits, J. A. J. (2015). Exercise augmentation of exposure therapy for PTSD: Rationale and pilot efficacy data. Cognitive Behaviour Therapy, 44(4), 314–327. http://dx.doi.org/10.1080/16506073.2015.1012740. R Core Team (2017). R: A language and environment for statistical computing [Programming language]. RStudio Team (2016). RStudio: Integrated development for R [Computer software]. Retrieved from http://www.rstudio.com/.

52