Cultural adaptations of a group exercise intervention for diverse outpatients in an urban community to improve exercise engagement and mood outcomes

Abstract

Background

Exercise is an accessible intervention that can improve mental and physical health. However, few people meet exercise recommendations, particularly those with anxiety or depressive symptoms. Individuals living in marginalized communities may have more barriers to exercise engagement. Affect-based interventions highlight the immediate mood benefits of exercise. They may perform better than standard health-based prescriptions at improving exercise engagement and mental health outcomes. Including important cultural adaptations may improve the efficacy of affect-based interventions for racially and economically marginalized communities.

Purpose

This pilot study aimed to establish feasibility, acceptability, and preliminary effectiveness of a culturally adapted affect-based intervention with culturally diverse outpatients with mood or anxiety symptoms and primary diagnoses ranging in severity.

Method

Twenty adult outpatients engaged in a 3-week Exercise for Mood group with culturally informed adaptations to reduce structural barriers, increase accessibility, and increase intervention engagement. Feasibility, acceptability, exercise, mood, and quality of life were examined.

Results

The intervention was deemed feasible (80% attended all sessions) and highly acceptable (average credibility ratings from 7.39 to 8.17 of 9). Total and moderate/vigorous exercise minutes improved over time (P = 0.015, P = 0.007; d = 0.33–0.42). Some changes in stress and ability to tolerate distress occurred (d = 0.64–0.82), but there were no noticeable improvements in depression and quality of life.

Conclusion

Cultural adaptations to an affect-based intervention were effective at retaining participants and increasing exercise in an at-risk, culturally diverse population. Preliminary results suggest potential for improvement in distress intolerance and stress, which may importantly target mental and physical health.

An effective intervention highlighting immediate mood benefits of exercise was culturally adapted to better meet the needs of at-risk individuals in the Bronx, NY with depression and anxiety symptoms. The adapted intervention resulted in increases in exercise and some improvements in stress after three weeks compared to baseline.

Implications.

Practice: Cultural adaptations to reduce structural barriers and increase accessibility of a 3-week affect-based Exercise for Mood intervention may be helpful in encouraging consistent group attendance and exercise engagement.

Policy: Effective hospital-based outpatient group programs should consider important cultural adaptations to existing effective psychosocial and health-based interventions to improve attendance, health behaviors (e.g. exercise), and mental health.

Research: Future research is needed to understand core necessary adaptations across various hospital settings and patients and to understand longevity of effects on exercise engagement and mental health symptoms.

Introduction

Mood and anxiety-related disorders are some of the most common (e.g. lifetime prevalence of 33.7% among individuals 18–64) [1] and chronic mental health conditions, resulting in devastating outcomes including higher mortality [2], worse overall health [3], and higher healthcare costs [4]. These mental health conditions and their detrimental outcomes have higher odds of occurring and are more chronic in marginalized populations, such as racial and ethnic minorities and those with lower income [5, 6]. Although typical evidenced-based treatments (e.g. psychotherapy, pharmacotherapy) for mood and anxiety disorders demonstrate some effectiveness, only approximately 41% and 45%–65% of individuals respond to initial depression and anxiety treatment, respectively [7, 8]. Further, barriers related to access, cost, and stigma may prevent individuals from seeking care [9]. These barriers may be more pronounced in underserved communities that face multiple disparities in social determinants of health (i.e. poor housing conditions, poverty) [10].

As such, exercise may be an alternative, accessible option that has demonstrated effectiveness in treating mood and anxiety-related disorders [11, 12], including in vulnerable and underserved populations [13]. Engaging in regular exercise also reduces likelihood of future onset of depression or anxiety symptoms and disorders [14, 15]. Yet, despite its benefits to mental and physical health, only approximately 23% of US adults adhere to exercise recommendations [16], with even lower rates in individuals with anxiety and/or mood-related disorders [17, 18] and in individuals who identify as Hispanic, are obese, report worse health, or have lower socioeconomic status [19].

Several factors may serve to increase exercise engagement. For example, anticipation or experience of positive affect during or after exercise predicts long-term exercise engagement including in new exercisers and sedentary individuals [20] as well as contributes to health behavior change [21] and reduces known intention-behavior gaps for exercise [22]. Psychological interventions aimed at adapting exercise approaches to focus on improving immediate mood or affect, known as affect-based interventions, have shown promise in increasing exercise engagement [23, 24]. Specifically, affect-based interventions highlight immediate mood benefits of exercise rather than the delayed long-term physical health benefits to improve the contingency between exercise behavior and immediate rewards. For example, individuals may be instructed to pay attention to immediate mood improvement, stress reduction, energy enhancement, or sense of accomplishment following exercise. These mood benefits, if attended to, can improve mood acutely and also lead to more exercise engagement in the future (as exercise will be seen as inherently rewarding). On the other hand, health-focused interventions encourage patients to focus on long-term health benefits as a rationale for exercise (e.g. weight loss, reducing cardiovascular risk, fitness), which require many bouts of exercise to realize and may lower exercise motivation. In comparison to health-focused interventions, affect-based interventions result in increased exercise intention, greater exercise engagement, and better mood outcomes [23, 24] Exercise for Mood [25] is a specific affect-based intervention based in cognitive behavioral theory, which was developed to improve exercise engagement and mood/anxiety symptoms for individuals with mood and anxiety disorders. The Exercise for Mood framework highlights the importance of paying attention to mood benefits and also integrates cognitive-behavioral motivation strategies. Exercise for Mood has shown early interventional success in improving exercise engagement and depressive symptoms [23, 26].

For racially and economically marginalized populations, cultural adaptations may improve the effectiveness of mood and anxiety treatments, including engagement with adaptive health behaviors, such as exercise. Cultural adaptations are defined as “systematic modifications of an evidence-based treatment or intervention protocol to consider language, culture, and context in such a way, that it is compatible with the client’s cultural patterns, meanings, and values.” [27]. Though evidence-based interventions, such as cognitive behavior therapy (CBT), are efficacious across races and ethnicities for improving psychiatric symptoms, nonsignificant, but notable trends suggest they may not be as beneficial for ethnic minority patients [28]. Indeed meta-analyses of cultural adaptations to treatments demonstrate efficacy in improving psychotherapy outcomes [29–31], including moderate to large effects for anxiety and depressive disorders [32, 33]. Further, when compared to their unchanged counterparts, culturally adapted interventions may be more effective [34].

Several frameworks exist to guide specific cultural adaptations to treatment [35–37]. Frameworks highlight potential important malleable elements. Though elements vary across frameworks, they tend to include strategies for treatment delivery (e.g. language, easy-to-understand text, culturally relevant illustrations/metaphors), treatment components (e.g. in-session techniques), and highlight surface/peripheral versus deep/core adaptations. Some specific adaptations leading to larger effect sizes for mental health outcomes in one meta-analysis included but were not limited to utilizing cultural metaphors, basing goals on clients’ cultural values, and using the clients’ preferred language. The studies implementing higher amounts of cultural adaptations, including the aforementioned and others, led to larger effect sizes for mental health outcomes [31]. Further, the framework of cultural humility should be considered in adaptations. It suggests that to develop strong bonds with patients and to enhance outcomes, the therapist must be open to the beliefs, values, and worldview of diverse clients [38]. This openness and self-reflection then can be utilized to adapt interventions for more cultural relevance to the clients [39].

Evidence also suggests that cultural adaptations to promote health behaviors may be useful, including a focus on social support barriers, motives, language, and adaptations in the setting, which demonstrate effectiveness at improving health behaviors in marginalized populations [40]. Reducing financial barriers, increasing access to safe alternatives to outdoor activity, cultivating interventions linked to group members’ values, and considering different types of physical activity that may resonate with group members may also be important [41, 42]. Further, anticipating and addressing cultural barriers to exercise participation was identified as an important factor in promoting exercise engagement in a review of culturally adapted exercise interventions [43].

The current pilot study examined a culturally adapted Exercise for Mood group intervention intended to enhance exercise engagement and improve psychiatric symptoms and quality of life in an adult hospital-based outpatient psychiatric clinic in the Bronx, New York (NY). The Bronx is predominately home to historically racially/ethnically minoritized individuals with the population being 57% Hispanic/Latino and 28% Black/African American [44], and is the poorest of the boroughs in New York City [45]. It is ranked 62nd out of the 62 counties in NY in several poor health outcomes (e.g. length of life, quality of life, and obesity) [46] and health factors (e.g. access to care, income, diet, and exercise) [47]. Structural racism has been implicated, although understudied, in contributing to health disparities [48]. The health disparities seen in the Bronx are complicated by poor air quality and pollution [49], and a disproportionate access to unhealthy foods [50] which may amplify physical and mental health concerns. In general, NY state is below the national average for adults who meet physical activity guidelines at 18.9% (vs. 22.9%) [16], making the unique Bronx population a particularly vulnerable group who may benefit immensely from physical and mental health benefits of exercise. As such, using a framework of cultural humility and building on past research examining effective cultural adaptions, the current pilot study primarily aimed to establish feasibility and acceptability of a culturally adapted group format of the Exercise for Mood intervention with culturally diverse outpatients in the Bronx who had mood or anxiety-related symptoms. Additionally, this pilot study assessed preliminary effectiveness of the intervention on exercise engagement, depressive symptoms, distress intolerance, perceived stress, and quality of life.

Methods

Design

Participants were outpatients enrolled in an open pilot trial of 3 weekly 1-h sessions of an Exercise for Mood group intervention implemented through the Adult Outpatient Department at Montefiore Hospital in the Bronx, New York. The 3-week intervention was administered in five cohorts of 3–5 members each over the course of 1 year (2017–2018), without substantial modifications between cohorts. This study was approved by the local ethics committee, and all participants provided written informed consent prior to completion of study procedures.

Setting and group leaders

The hospital-based outpatient psychiatric clinic, which provides care to an average census of 1665 patients, is diverse with respect to race (24.56% of patients identifying as Black; 13.57% White, 47% Multiracial) and ethnicity (55.98% Hispanic/Latino). It largely serves individuals of lower socioeconomic status, with most patients eligible for Medicaid (75%) or Medicare (9.8%), which they use to cover clinical services. There is a high degree of medical comorbidities present in this patient population (e.g. asthma, diabetes, and hypertension). At the time of data collection for this study, visits were primarily conducted in-person, including for this group.

Groups were led by two psychologists, one identifying as a white/Latina (Cuban) woman with experience in exercise interventions and one identifying as a white woman who has worked with the Bronx population for over 10 years, with a dedication to developing culturally responsive, evidence-based care to decrease mental health disparities. Both group leaders had training and experience working with evidence-based interventions and adapting for diverse communities [51–53].

Participants

Participants were outpatients ages 18–65 experiencing current symptoms of anxiety and/or depression and a variety of primary diagnoses ranging in severity. Participants were recruited through flyers placed in the hospital-based outpatient psychiatric clinic as well as referrals from outpatient mental health providers, who were informed of the study through provider email list serves and word-of-mouth. To increase external validity, no specific primary psychiatric diagnoses were required. Exclusion criteria to minimize risk were: current anorexia or bulimia; acute episode of psychosis; intoxication by drugs or alcohol at the time of intervention; risk for exercise according to the physical activity readiness questionnaire (PAR-Q) or excluded medical conditions (e.g. moderate to severe heart conditions, bone/joint/muscle problems, seizure disorders) as assessed by study physician; and women who were currently pregnant or planned to become pregnant during the intervention. There were no exclusion criteria related to current physical activity levels. Participants were recruited through referrals from clinicians, screened over the phone for appropriateness, and medically screened in-person by a study physician for ability to safely participate in exercise. Group sessions were billed to insurance under the group psychotherapy standard billing codes since the group intervention aimed to improve mood and anxiety through discussion of cognitive behavioral motivation enhancement strategies to engage in exercise. Participant insurance coverage was verified ahead of group participation.

Intervention

The group intervention was completed in three, 1-h weekly sessions. Groups consisted of 3–5 members. The intervention included principles from Otto & Smits’ Exercise for Mood and Anxiety [25] modified for a low income, racially and ethnically diverse population served in this Bronx hospital-based outpatient psychiatric clinic. Each session included setting specific goals for exercise for the following week and identifying potential motivation enhancement strategies to utilize as well as problem-solving potential barriers that may be encountered. Core elements of the intervention as well as adaptations for structural barriers are detailed in Table 1 and briefly summarized below. The intervention was primarily didactic and did not involve in-session exercise, besides the walk around the hospital (see below). At-home exercise goals were identified individually based on access and what patients enjoyed to increase potential for long-term engagement and generalizability across varying sociodemographic populations with differential access to exercise equipment. Patients completed a worksheet to identify 1–2 types of exercise and set goals for number of days and amount of time completing the exercise for the coming week. Patients were encouraged to choose activities that would be enjoyable and would likely improve their mood. They shared their goals with the group at the end of the session and instructors aided in goal setting if needed.

Table 1.

Core elements of Exercise for Mood and cultural adaptations

Core elements of Exercise for Mood
Session 1	Rationale for attending to mood benefits of exercise Introduce mood log Discuss possible exercises Defining motivation as an active construct Motivation enhancement: breaking tasks into smaller parts Set exercise goals
Session 2	Review exercise logs Review concepts from previous week Motivation enhancement: combining motivations, using mood goals, creating habits, identifying low motivation situations Key elements for enjoying exercise Set exercise goals
Session 3	Review exercise logs Review concepts from previous week Motivation enhancement: using rewards Reducing self-critical thoughts about exercise Planning for maintaining exercise routine Set exercise goals
Cultural adaptations for specialized population
Barrier targeted	Modification example
Treatment attendance	Reduction from typical 6–8 session format to 3 1-h group sessions Providing metro cards for transportation
Treatment engagement: culturally, temporally, and geographically relevant examples	Motivation examples: “recertify for food stamps, care for family, watch a rerun of Law & Order” Daily effort examples: “long day of appointments, dealing with the landlord, paying bills” Combining motivations example: Pokemon Go; going to the Oval and attending to scenery; getting of the bus or subway a stop earlier
Accessibility: financial	Provided list of free neighborhood exercise resources (e.g. classes, running tracks, fitness equipment, pools, outside fitness locations) and other freely available resources (e.g. YouTube videos) Sessions offered at no cost and billed to insurance
Accessibility: exercise types and locations	Provided examples of exercise types to allow for patient choice As above, provided varying locations for exercise, including within the home and outside the home
Accessibility: comprehension and language	Provided outline for important points from each session Added structured worksheets with core concepts (e.g. goal setting worksheet) Added structured and directive homework worksheets Modified language for comprehension level
Accessibility: in-session experiential activities	Walk Montefiore mile and fill out mood log pre and post walk Mindfulness practice in session

Core elements

Session 1 laid the foundation for the rationale for exercise as an intervention to improve mood (depression and anxiety symptoms), the core feature of the Exercise for Mood intervention that was emphasized throughout all three group sessions. The focus on immediate mood benefits resulting from exercise, for example, specifically attending to mood improvement, anxiety/stress reduction, energy enhancement, or a sense of accomplishment, is distinct from typical health-based reasons for exercising (e.g. to lose weight, prevent cardiovascular disease). Focusing on immediate mood benefits is a strategy to promote better momentary mood, increase exercise engagement, and ultimately reduce depression and anxiety symptoms. Other core elements of the Exercise for Mood intervention included motivation enhancement strategies rooted in evidence-based cognitive behavioral and motivational interviewing principles. Strategies are applied to facilitate motivation to start and continue with an exercise routine and reduce barriers to exercise. The final session concluded with creating a maintenance plan.

Cultural adaptations

Several elements were modified or added to adapt the intervention to be most effective to the Bronx-based population seen in this hospital-based outpatient psychiatric clinic. Cultural adaptations were guided by existing frameworks suggesting various elements that can be targeted in adaptations (e.g. metaphors, content of interventions, treatment goals, delivering methods, and context) [27, 35–38]. See Table 1 for specific examples of the different modifications. Adaptations focused on strategies to reduce the impact of structural barriers, increase accessibility, and increase engagement with the intervention. For example, a common barrier is treatment attendance, which may be due to conflicting appointments for supportive services, problems with childcare, lacking finances for transportation, or other psychosocial stressors. Indeed, the rate for missed psychiatric appointments in the Montefiore Adult Outpatient Department typically averages around 25%. To limit missed appointments, we modified the group to be just three sessions (typically ranges from 6 to 8 sessions, but has been delivered in a single group session previously [23] and modified brief group interventions have been demonstrated to be effective for other interventions, for example, CBT [54]) and provided free Metro cards for transportation.

To target engagement, we modified examples to be more culturally relevant to the population, using language that is user-friendly and accessible, responsive to the context of those residing in the urban community of the Bronx, and to the current events occurring at the time of treatment delivery (see Table 1 for specific examples). Additionally, we included experiential activities together in session to demonstrate the concepts in real-time. This included practicing rating pre and post exercise mood in the context of walking the “Montefiore Mile,” a ¼ mile walk through a loop marked within the hospital, which participants were encouraged to walk four times if they were able before or after group sessions or around other appointments. We also completed an in-session mindfulness exercise to help members understand how to incorporate mindfulness into exercise to promote enjoyment.

Several modifications targeted accessibility. These included modifications for financial accessibility, including the development of a list of resources that focused on access to exercise that could be completed for free in safe environments. We also considered modifications targeting accessibility of the intervention materials. For example, worksheets and handouts, which were printed for participants, were modified for comprehension level, with attention to the fact that many group members’ first language was not English. They also provided more structured, directive, and concrete strategies for learning, remembering, and implementing core concepts. For example, patients were provided with worksheets summarizing important points from each session in brief bullet points in lay language. Additionally, worksheets were modified to be in the form of checklists to reduce reliance on memory alone for core motivational concepts. Sessions were offered at no cost as they were billed to insurance. Compensation for participation was not provided; however, participants who completed all sessions were provided with a healthy snack pack (e.g. fruit, granola bars) at the end of Session 3.

Measures

Medical information and demographics

Self-reported age, gender, educational attainment, race, and ethnicity were collected at Week 1. Information about psychiatric and medical diagnoses were obtained from the medical record and confirmed by study physician at screening. Height and weight to calculate body mass index (BMI) were obtained at screening.

Primary outcomes: feasibility and acceptability of the culturally adapted intervention

Feasibility was determined by attendance at all three sessions. Acceptability was determined by the validated Credibility-Expectancy Questionnaire (CEQ) [55], a six-item measure assessing outcome expectancy and credibility, which was administered after Week 3 (final group session). From the CEQ, we used its three credibility ratings, measured on a 0–9 Likert scale, which were (i) how logical the intervention seemed, (ii) how successful participants felt the intervention would be at reducing anxiety/depression symptoms, and (iii) how confident participants would be in recommending this treatment to a friend. We also used one expectancy rating on the CEQ that measured how much improvement in symptoms participants thought would occur measured from 0% to 100% in 10% increments. Higher scores indicate greater acceptability of the intervention.

Exercise

The International Physical Activity Questionnaire (IPAQ) [56] is a validated seven-item measure of physical activity at various intensities completed over the previous week. The IPAQ instructs consideration of the following types of activities: “activities you do at work, as part of your house and yard work, to get from place to place, and in your spare time for recreation, exercise or sport.” Metabolic equivalents (METs) were calculated for total activity (low, moderate, and vigorous intensity) and for the combination of moderate and vigorous intensity exercise. Aligned with recommendations for IPAQ scoring, responses for each variable of time spent: (i) walking, (ii) in moderate exercise, and (iii) in vigorous exercise were limited to 4 h per day each (if endorsed >4 h of time) to account for potential incorrect completion (e.g. putting total minutes per week of exercise participation rather than daily exercise participation) [57].

Depression, distress intolerance, perceived stress, and quality of life

Past week depressive symptom severity was measured with the 20-item Center for Epidemiological Studies Depression Scale (CES-D) [58]. Current distress intolerance, the inability to tolerate stressful states, was measured with the 10-item Distress Intolerance Index (DII) [59] and perceived stress in the past month was measured with the 10-item Perceived Stress Scale (PSS) [60]. Quality of life (e.g. physical health, mood, activities of daily living, and life satisfaction) over the past week was measured by the Quality of Life Enjoyment and Satisfaction Questionnaire scored as a percent of 100 (Q-LES-Q) [61]. Higher scores on CES-D, DII, and PSS indicate poorer functioning (i.e. worse depression, higher distress intolerance, and higher perceived stress). Higher scores on Q-LES-Q and IPAQ indicate better quality of life and more exercise engagement. Self-report assessments of exercise and psychiatric symptoms were administered weekly during the 3-week intervention and at 1-month postintervention follow-up.

Data analysis

Baseline demographic, clinical characteristics, and feasibility/acceptability were summarized using mean and standard deviation for continuous variables and proportion for categorical variables. Feasibility was determined by proportion of participants attending all three intervention sessions. Acceptability was determined by mean ratings on CEQ items and proportion of participants reporting expectancy of >50% improvement in symptoms. To understand preliminary effectiveness of this culturally adapted intervention, we examined changes in exercise (total METs and moderate/vigorous intensity activity METs), depressive symptoms (CES-D), distress intolerance (DII), perceived stress (PSS), and quality of life (Q-LES-Q percent) over time using separate longitudinal mixed effects models. Models had time as a predictor with all time points (Weeks 1–3, 1-month follow-up) included and utilized restricted maximum likelihood estimation and an unstructured variance-covariance matrix. Pairwise comparisons of estimated marginal means were examined to identify significant differences at individual time points. Intent-to-treat models were used for all participants with at least one time point of data. All significance tests were two-tailed with a P level of 0.05. Analyses were conducted in SPSS Version 28.0.

Results

Participants

Twenty-one outpatients enrolled in the study; however, only twenty outpatients (90% female, mean age [SD] = 48.0 [11.7]) completed baseline measures and attended at least one group session. Aligned with the composition of the clinic and the Bronx more broadly, our participants were diverse with respect to race and ethnicity (25% Black, 65% Hispanic), primarily had high school or some college education (70%), and were likely to have a medical comorbidity (87.5% with ≥1 medical comorbidity). Mean baseline BMI was in the obese range (33.7 ± 7.2). Baseline depressive symptoms and distress intolerance were in the severe range and perceived stress was moderate. Mean METs of moderate/vigorous exercise was high at 1235.0 ± 2016.9 (representing ∼308 moderate intensity minutes per week); however, 50% of the sample was below exercise recommendations (150 min of moderate intensity exercise) at baseline. See Table 2 for detailed baseline information.

Table 2.

Baseline demographics and characteristics

Baseline characteristics	N   =  20 Mean (SD)/n (%)
Age	48.0 (11.7)
Gender
Female	18 (90.0)
Male	2 (10.0)
Race
White	5 (25.0)
African American/Black	5 (25.0)
Other	8 (40.0)
Missing	2 (10.0)
Ethnicity
Hispanic	13 (65.0)
Non-Hispanic	7 (35.0)
Education
Graduate degree	2 (10.0)
Completed college	3 (15.0)
Some college	9 (45.0)
Completed high school	4 (20.0)
Less than high school	2 (10.0)
Principal diagnoses
Major depressive disorder	4 (20.0)
Persistent depressive disorder	3 (15.0)
Schizoaffective disorder	3 (15.0)
Anxiety disorders	2 (10.0)
Bipolar II disorder	1 (5.0)
Borderline personality disorder	1 (5.0)
Adjustment disorder	1 (5.0)
Missinga	5 (25.0)
Medical comorbiditiesa
Hypertension	5 (25.0)
Headache/migraines	5 (25.0)
Diabetes/prediabetes	5 (25.0)
Hyperlipidemia	5 (25.0)
Pain syndromes (e.g. fibromyalgia)	5 (25.0)
Arthritis	4 (20.0)
Asthma	3 (15.0)
Hypothyroidism	3 (15.0)
Sickle cell anemia	3 (15.0)
GERD	2 (10.0)
Sleep apnea	1 (5.0)
Axenfield Rieger syndrome	1 (5.0)
PCOS	1 (5.0)
Hypoglycemia	1 (5.0)
Psychiatric medication usage
Antidepressants	11 (55.0)
Antipsychotics	6 (30.0)
Benzodiazepines	4 (20.0)
Missinga	5 (25.0)
Body mass index	33.7 (7.2)
Center for Epidemiological Studies Depression Scale	28.2 (10.3) Severe range
Distress Intolerance Index	25.0 (6.5) High range
Perceived Stress Scale	23.1 (5.1) Moderate range
Quality of Life Enjoyment and Satisfaction Questionnaire (%)	44.0 (15.1)
Total metabolic equivalents of moderate and vigorous exercise	1235.0 (2016.9)
Percent above exercise recommendations (150 min of moderate intensity per week)	10 (50.0)

^aMissing medical record information for n = 5 participants.

Feasibility and acceptability of culturally adapted exercise for mood group intervention

Of the 21 participants enrolled in the study, only one participant dropped out prior to completing baseline measures or attending any group sessions, suggesting a strong conversion rate from initial phone screen to group participation. This participant was not included in further analyses. Of the 20 participants who attended at least one group session, 80% (n = 16) completed all three sessions. Three individuals attended two sessions and one attended one session. Eight participants were lost to follow-up for the final assessment visit.

The intervention was rated as highly credible with mean ratings of 7.39 (SD: 1.7), 7.61 (SD: 1.3), and 8.17 (SD: 1.4) out of nine for how logical the intervention seemed, how successful they felt it would be at reducing symptoms, and how confident they were in recommending it to a friend, respectively. Expectancy was reasonable, with a mean expected symptom improvement of 61.7% (SD: 20.9). Of note, 72% (n = 13) of participants with expectancy ratings reported thinking their symptoms would improve by at least 50%.

Preliminary effectiveness: changes in exercise

Mixed effects models for total [F(1,3) = 16.6, P = 0.015] and moderate/vigorous [F(1,3) = 9.7, P = 0.007] METs of exercise demonstrated a significant effect of time overall (Fig. 1). For moderate/vigorous METs, pairwise comparisons of estimated marginal means between Week 1 [M(SE) = 1235.0 ± 451.0; ∼308 moderate intensity minutes] and Week 3 (2132.4 ± 554.8; ∼533 moderate intensity minutes) were significant at a medium effect size (d = 0.42; P = 0.006). Comparisons between Week 1 and follow-up (2163.3 ± 788.5; ∼540 moderate intensity minutes, d = 0.40, P = 0.251) did not reach significance, likely due to smaller sample size, but did demonstrate a similar effect size. No pairwise comparisons of mean differences for total METs reached significance. Effect sizes for changes in estimated marginal means of total METs from Week 1 (M ± SE: 2125.2 ± 559.8; ∼531 moderate intensity minutes) to Week 3 (2935.9 ± 597.6; 733 moderate intensity minutes; d = 0.33, P = 0.066) and follow-up (3010.8 ± 839.3; 752 moderate intensity minutes; d = 0.33, P = 0.393) were in the small to medium range.

Figure 1.

Amount of exercise calculated in metabolic equivalents. Total includes low, moderate, and vigorous intensity exercise reported in previous week.

Preliminary effectiveness: changes in depression, distress intolerance, stress, and quality of life

No mixed effects models for psychiatric outcomes demonstrated significant overall time effects; however, the model for DII had a trend-level effect [F(1,3) = 2.8, P = 0.076]. In examining pairwise comparisons of estimated marginal means, significant mean differences between Week 1 and follow-up emerged for DII (Week 1: 24.9 ± 1.5, follow-up: 18.9 ± 2.4, d = 0.82, P = 0.021) and PSS (Week 1: 23.1 ± 1.1; follow-up: 19.4 ± 1.9, d = 0.64, P = 0.046) at large and medium to large effect sizes, respectively. Changes from Week 1 to Week 3 in DII (Week 3: 24.9 ± 2.1, d = 0.01, P = 0.950) and PSS (22.1 ± 1.2, d = 0.19, P = 0.218) were minimal. Though pairwise comparisons of estimated marginal means did not indicate any significant effects, mean differences from Week 1 (28.2 ± 2.3) to follow-up (24.5 ± 3.2, d = 0.35, P = 0.199) for CES-D indicated a small to medium effect, but changes to Week 3 (28.3 ± 3.7, d = 0.01, P = 0.949) were minimal. This pattern was similar for quality of life (Week 1: 44.0 ± 3.4; Week 3: 46.5 ± 4.3, d = 0.26, P = 0.459; follow-up: 47.7 ± 4.9, d = 0.42, P = 0.301), though with a small effect evident between Weeks 1 and 3.

Discussion

Exercise is an accessible, cost-effective intervention with known efficacy to improve anxiety and depressive symptoms [11, 12]. Specifically, Exercise for Mood, which highlights attention to the immediate mood benefits of exercise rather than the long-term health benefits, has shown early promise in improving exercise engagement and psychiatric symptoms [23, 26]. Our current study aimed to examine the feasibility, acceptability, and preliminary effectiveness of a culturally adapted group-based Exercise for Mood program, modified specifically for our unique Bronx-based sample. Aligned with the general composition of the Bronx, our sample was diverse with respect to race, ethnicity, and socioeconomic status. They also reported significant physical and mental health challenges. Indeed, 87.5% of our sample reported at least one medical comorbidity and overall our sample reported severe initial depressive symptoms and moderate levels of distress intolerance and perceived stress. Overall, and consistent with our hypotheses, our results supported the feasibility and acceptability of this culturally modified Exercise for Mood program. Further, as hypothesized, participants increased their total amount of exercise and specifically moderate and vigorous exercise by Week 3 of the intervention. Improvements in distress intolerance and perceived stress were evident by 1-month follow-up, though there was less support for impact on depressive symptoms or quality of life, contrary to hypotheses.

Regarding feasibility, in our study, 80% of participants attended all three consecutive intervention sessions; only 1 individual attended just 1 session. Rates align with other evidence-based group interventions (e.g. 24.6% dropout) [62] and exercise interventions (e.g. 18.1%–22.4% dropout) [63, 64]. Further, our attendance rates were significantly better than general attendance at the hospital-based outpatient psychiatric clinic as a whole, which has a 25% no-show rate. Our high rates of attendance to a three session treatment against the backdrop of generally modest attendance suggest the importance of modifications that enable individuals to participate in such group interventions. For example, these changes could reduce financial barriers, including billing sessions to insurance, decreasing number of sessions needed to travel for, and providing free travel options, and limit time necessary to participate (only 3 sessions as opposed to months-long protocols) may be important considerations. While Exercise for Mood aims to increase exercise, the primary skills provided in the group relate to cognitive behavioral motivation enhancement strategies and the primary goal is mood and anxiety improvement. Therefore, sessions can be billed under group psychotherapy codes, which is important given those with insurance coverage tend to access mental healthcare more than those without [65]. If a potential participant does not have insurance coverage, sliding scale fees may be considered to provide broader access. A shorter group, such as this three-session version, may still be feasible for those who have to pay out-of-pocket as well.

Other strategies to increase access are to consider telehealth only or hybrid options. This study was completed prior to the significant increase in use of telehealth for mental health services provision. Currently, approximately 43%–54% of mental health visits generally are conducted via telehealth [66]. In the Montefiore Adult Outpatient Department approximately 72% of visits are conducted by telehealth. Thus, offering groups via telehealth has been a way to increase access to care; however, although the clinic currently offers both individual therapy and group therapy both virtually and in person, missed appointments related to chronic psychosocial stressors (i.e. housing instability, comorbid health conditions) remain a challenge for this patient population.

Our diverse Bronx-based participants also rated the Exercise for Mood intervention as highly credible, including that it made logical sense to them, they believed it would be successful at improving their symptoms, and they would recommend it to a friend. Most also expected it would improve their psychiatric symptoms substantially (by at least 50%). These results replicate past research that demonstrates Exercise for Mood as a highly acceptable treatment for mood and anxiety-related symptoms [23, 26] and exercise more generally to improve mental health [11, 12]. Further, the findings related to acceptability of Exercise for Mood, a psychosocial treatment framework to promote exercise engagement and psychiatric symptom improvement, suggest this intervention may be a useful psychosocial alternative to traditional cognitive behavioral therapy or medications to prescribe for individuals with anxiety and depression symptoms. Preference research that indicates that generally adults prefer psychosocial treatments over medications [67] and specifically, those who identify as Hispanic or Black/African American are more likely to prefer psychological-based treatments as compared to medications [68]. Preference for psychological treatment may extend to psychosocial, health-based interventions, such as Exercise for Mood, and it should be considered another potential treatment option.

Further, our acceptability findings suggest the potential importance of the included cultural adaptations to increase health behavior engagement in our sample. Our adaptations targeted strategies for treatment delivery and treatment components. Aligned with meta-analyses suggesting the importance of utilization of cultural metaphors to improve mental health outcomes [31], we modified examples of motivation enhancement strategies to be more culturally, temporally, and geographically relevant. For example, we suggest combining exercise motivation with higher level motivators, such as playing Pokemon Go (a particularly popular app-based game at the time), or with relevant activities, such as getting off the bus or subway a stop earlier. To reduce financial barriers and consider different types of physical activity aligned with previous recommendations [41], we recommended free parks and activities in the area, but also several other free and accessible exercise options (e.g. YouTube videos of salsa dancing). Further, high ratings on the logical credibility dimension may suggest our adaptations to create more structured worksheets, directive homework, and modified language for comprehension were important for increasing accessibility. In all, our findings contribute to the growing yet sparse area of research on cultural adaptions to evidence-based treatment approaches. Future research should obtain more specific feedback on which adaptations were most useful, particularly to improving treatment attendance, treatment access, and understanding and relevance to diverse outpatient populations in the Bronx and elsewhere.

As noted, consistent with expectations, our results indicate preliminary effectiveness of our intervention in improving exercise engagement. Specifically, total amount of exercise as well as moderate and vigorous intensity exercise improved by Week 3 (intervention endpoint) and appear to be maintained at 1-month follow-up. Our findings are consistent with past work in affect-based interventions in nonpsychiatric adults that report change in minutes per week of exercise demonstrate a similar increase (e.g. 240 min per week [69]) to our intervention (∼225 min per week), which included individuals with moderate to severe anxiety and depression symptoms. These results are promising as exercise plays an important role in the prevention and improvement of physical health issues (e.g. cardiovascular disease) [70]. This highlights the importance of considering strategies to promote exercise, especially in medically compromised populations. Affect-based interventions, such as Exercise for Mood, may be particularly helpful frameworks as improving affect during and after exercise as well as promoting exercise enjoyment are implicated in long-term exercise engagement [23, 24]. However, it is important to integrate cultural considerations into effect-based exercise interventions as well. For example, we considered the importance of safe, outdoor spaces for exercise as park-based activity is linked to mental health outcomes, but only if considered safe [42]. We also integrated within-session exercise [e.g. Montefiore mile] and highlighted various types of activity to ensure access and enjoyment of the exercise participants completed, while simultaneously allowing the patient to connect more directly with the content which may have increased the value of the treatment.

Though feasibility, acceptability, and efficacy for improving exercise were encouraging, results on psychiatric outcomes were less clear. Distress intolerance and perceived stress improved by the 1-month follow-up but not notably during the 3-week intervention. These findings support claims that propose that exercise acts as a form of exposure [71] to physiological sensations associated with anxiety that may be experienced as uncomfortable/distressing, which may need more repetitions to see clinically significant impacts (i.e. more engagement in exercise, more ability to tolerate these uncomfortable sensations which are also induced by exercise). No notable effects emerged on depressive symptoms or quality of life which was unexpected given the known effects of exercise on depressive symptoms and quality of life [11, 72]. The lack of evidence supporting changes in depressive symptoms or quality of life may be due to the complexity of this populations’ mental and physical health concerns, the chronic unremitting psychosocial stressors in the contexts in which they reside, or potentially the brief duration of the intervention (3 weeks).

Despite the encouraging initial findings from our pilot study, there are a few limitations to acknowledge. First, the sample size was small and there was no comparison group. Initial signals from this small sample suggest utility to further examine this culturally modified Exercise for Mood group intervention and to consider adding a viable comparison such as an exercise for health framework (to test the differences in attending to immediate mood vs. long-term health benefits) or to compare to other evidence-based interventions, such as cognitive behavior therapy, medications, or mindfulness-based interventions. Additionally, it may be useful to assess pre-existing mood-based versus health-based motivations for engaging in exercise to better understand how to tailor exercise-related prescriptions in the future. Health outcomes may also be assessed, especially if longer term follow-ups are included. Second, the intervention period was only three weeks with a 1-month follow-up. This brief intervention period may not have allowed enough time to see significant changes in psychiatric symptoms particularly for this severe and highly stressed population. However, increasing the intervention period should be balanced with our results that suggest limited sessions may help with attendance. Testing varying interventions lengths that optimize attendance may be useful. Other considerations for flexibility include a modular approach allowing for multiple entry points and incorporation of booster sessions or periodic check-ins (e.g. 3 months, 6 months, 1 year), which may improve exercise adherence and benefits to psychiatric symptoms to allow more skill practice for those who need it, with more empirical support in the future. Additionally, future studies should implement longer follow-up periods to understand maintenance of gains, particularly if effects on exercise engagement are maintained over longer periods (e.g. 6–12 months) and if effects on psychiatric symptoms are either maintained (e.g. for perceived stress, distress intolerance) or evidence delayed improvement (e.g. depressive symptoms). Third, physical activity was only measured by self-report. Though the IPAQ is well-validated and widely used, self-reported physical activity may be biased and device-based measurement of objective physical activity, or regular exercise logs if devices are unavailable, should be considered in future studies. Fourth, our sample was heterogeneous with respect to primary diagnosis, medical comorbidities, and amount of baseline exercise. While this was a strength in terms of examining this intervention in a naturalistic environment, it limits ability to make recommendations for specific types of individuals who may benefit most from the intervention. Of note, it may be especially important for future studies to consider limiting participation to individuals who are not currently exercising as they may receive most benefit from beginning an exercise routine, or to a more homogenous participant group (e.g. just anxiety diagnoses or mood disorder diagnoses), which might enhance effect of the intervention on psychiatric symptom change. Fifth, while cultural adaptations were carefully considered and implemented in our pilot study, we did not collect specific feedback about the modified elements. Future research would benefit from including quantitative and qualitative measures to capture the specific impact of the different cultural adaptations. This information can be used to determine what adaptations were most effective and for whom and will allow for ongoing adaptation to optimize the intervention for various unique populations. In particular, it may be helpful to understand modifications that may better target depressive symptoms and quality of life, as these dimensions remained relatively stagnant in our sample. Sixth, the majority of the sample were female, and while consistent with our studies examining exercise of mental health conditions [12] future research would benefit from examining Exercise for Mood with a more gender diverse population. Finally, future research should continue the progression of this study and examine the implementation of this culturally adapted Exercise for Mood protocol in other outpatient hospital settings to understand if promising initial results are maintained across systems, settings, and populations.

In summary, the findings demonstrated the feasibility and acceptability of a culturally adapted group-based Exercise for Mood protocol in a hospital-based outpatient psychiatric clinic in the Bronx, NY. Overall, these promising findings as well as the early evidence of improving exercise engagement and some stress-related psychiatric symptoms suggest the viability of implementing culturally adapted, affect-based exercise programs in hospital settings. Since exercise as a treatment is not a classic mental health approach, this format may result in decreased stigma, which is an important consideration in marginalized populations [9]. These findings help propel the field’s goal of person-centered approaches, by leveraging cultural adaptations to enhance treatment. Exercise may also be used to augment other evidence-based approaches which could be helpful for vulnerable populations with several medical comorbidities and chronic health issues.

Funding Source

K.L.S. effort on this manuscript was supported by the National Institute of Mental Health (NIMH: K23MH122773). The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Conflicts of Interest

K.L.S. and L.G. receive support from the National Institutes of Health. K.L.S. serves as an associate editor for Mental Health and Physical Activity. The other authors declare they have no conflicts of interest that are relevant to the content of this article.

Human Rights

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. IRB approval from Montefiore Medical Center was obtained prior to conducting the study.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

Transparency Statements

This study was not formally registered. The analysis plan was not formally registered. De-identified data from this study are not available in a public archive. De-identified data form this study will be made available (as allowable according to IRB standards) by emailing the corresponding author. There is not analytic code associated with this study. Materials used to conduct the study are not publicly available, but the intervention protocol can be made available by contacting the corresponding author.