Acceptability, engagement, outcomes, and dose-response associations of a mindfulness-based meditation app in individuals waiting for psychological services

doi:10.21203/rs.3.rs-4952898/v1

Download PDF

Research Article

Acceptability, engagement, outcomes, and dose-response associations of a mindfulness-based meditation app in individuals waiting for psychological services

https://doi.org/10.21203/rs.3.rs-4952898/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

While mindfulness apps have received growing clinical attention, their integration within health systems has received limited empirical investigation. In this study, we evaluated a mindfulness app as a low-intensity treatment option for adults waiting for psychological services. A non-randomized clinical trial was conducted with a 4-week acute intervention period with an 8-week follow-up. Adults (N = 193) with moderate depression and anxiety symptoms, completed a baseline session and received access to AmDTx, a mobile mindfulness training app. Additional assessments were completed at 2, 4, 8, and 12 weeks. Descriptive statistics of attrition, adoption, acceptability, and engagement were computed. Linear mixed models estimated treatment outcomes for functional disability (primary endpoint), depression, anxiety, stress, rumination, and mindful awareness/acceptance. We also evaluated the dose-response association between app use and functional disability.

Results

Using intent-to-treat analyses, there was a 75% adoption of the app and a 30% attrition rate in the first 4 weeks. In addition, 1.09 hours of meditation time and 9.16 exercises were recorded on average within the first 4 weeks. Participants reported positive credibility, acceptability, and usability ratings on established measures. Treatment effects were observed in the expected direction for all outcomes but one (mindful awareness). Dose-response relationships indicated that increases in app engagement correlated with decreases in functional disability.

Conclusions

The findings reinforce the potential for AmDTx, and mindfulness apps more broadly, to serve as low-intensity tools to alleviate unmet service needs and impart clinically meaningful benefit for a significant subset of those waiting for psychological services.

Clinical Trial Registration: Clinicaltrials.gov, NCT05211960, Registered 2022-01-26.

mindfulness

digital interventions

mobile app

mindfulness meditation

mental health

mHealth

smartphone apps

implementation science

Unmet needs for mental health services are pervasive even in more high-income countries [1, 2]. The COVID-19 pandemic has likely exacerbated these unmet services, with reports of youth and emerging adults being disproportionately impacted [3, 4]. Alleviation of these unmet needs will require expansion of treatment resources, as well as attention to the barriers to accessing and engaging with evidence-based treatments [5]. Documented barriers to psychological services include knowledge gaps, lack of mental health service integration, cultural and language barriers, concerns about stigma, costs of services, and inequalities due to geography or demographics [6]. Of particular concern is the often-lengthy wait times for publicly funded offerings of psychological services [7], and the potential negative impact of this wait on treatment once received [8]. Notably, it is during these waiting periods that individuals typically receive no treatment or are offered low-intensity self-help resources, often with limited empirical support. Clinically validated digital interventions hold the potential to increase access to immediate care, particularly when adequately integrated within healthcare systems.

Mental health apps have been accessible on smartphones for over a decade, with growing evidence for their ability to reduce depression, anxiety, and other psychiatric symptoms [9–12]. Benefits of smartphone apps include convenience, scalability, anonymity, personalisation, and real time monitoring of patients. However, very few studies have examined the integration of mental health apps within health systems, making it difficult for clinicians to judge whether a given app is suitable for use in a real-world healthcare setting [13]. Other barriers to widespread app adoption in healthcare include privacy concerns, integration with electronic records, expectations regarding retention and engagement, and limited capacity for clinicians to view potentially insightful data obtained through the app. Still, given the potential impact, there is burgeoning support for the integration of smartphone apps into complex care models, and thereby better support mental health needs across all populations.

Mindfulness-based meditation [14] has received significant empirical evaluation as an evidence-based treatment, with hundreds of face-to-face randomized controlled trials (RCTs) supporting its efficacy for numerous psychiatric and health outcomes [15]. Mindfulness apps typically include the delivery of audio and/or video exercises delivered on-demand and asynchronously to users. Studies have accumulated significant empirical evidence for their ability to improve a wide range of mental health outcomes, including depression, anxiety, stress, wellbeing, life satisfaction, and burnout across more than 70 RCTs [16–18]. The effects on symptoms tend to be small to medium in size, but they are frequently better than no treatment (e.g., waitlist) and active controls (e.g., non-therapeutic app). Importantly, mindfulness apps have also been found to increase mindfulness-based meditation skills such as awareness, acceptance, attention regulation, decentering, and cognitive defusion [19, 20]. Thus, there is robust evidence for mindfulness apps as low intensity tools for the management of mental health symptoms and credible cultivation of mindfulness practice.

Several variables have been used to understand the acceptability and engagement of mindfulness apps. Many studies report attrition rates, which refer to the failure to complete research protocol components, such as follow-up assessments, after receiving access to the intervention. In a review of mindfulness app RCTs, Linardon [21] found an average attrition rate of 25%, and up to 39% in larger samples (greater than 100 participants) and when more general populations of adults were targeted rather than a specific mental health condition. These rates of attrition are similar when compared to reviews of mental health apps studied more generally [22, 23]. Across these reviews, decreases in attrition were associated with use of reminders, monetary compensation, requiring human interaction during enrollment, and when feedback to users was a component of the app. While these reviews did not find effects for the length of the trial, studies have found that attrition rates during trials of self-guided apps steadily increase after 4 weeks [24–27], suggesting that the first 4 weeks of the trial are especially salient to foster engagement. Importantly, attrition metrics provide a limited evaluation of mental health apps adoption, engagement, and impact. For example, adoption refers to how many people access the intervention (or use it at all) and is the first step towards engagement. Mindfulness app RCTs have reported adoption rates as the number of people who “download”, “register”, or “access” the app [21]. Using supplementary information extracted from Linardon [21], we found that adoption ranged from 45–100%, with a weighted average of 62% across 11 RCTs and 1758 participants.

Following adoption, if someone does not engage with app content, then it is unlikely they will receive any therapeutic benefit. Engagement is also variably reported across RCTs of mental health apps [21, 22] but frequent indices are the average number of minutes (or minutes per day) using the app, average number of days (or times) the app is accessed, and the number of exercises (or activities) completed and/or started within the app. Some of these metrics have limitations, as one can open an app or “use” it without engaging in meditation practice as prescribed for treatment purposes. Studies also commonly report total app usage over the entire trial, or on a weekly or daily basis, which may inflate the actual time spent meditating. Fewer studies have assessed time spent meditating directly, which may be complicated by the fact that researchers do not always have access to third-party app data. In reviewing supplementary information reported in Linardon [21], we found 8 RCTs that reported meditation minutes and a weighted average meditation time of 2.07 hours across 930 participants. These findings highlight important issues surrounding how engagement is measured as there is no agreed upon definition of engagement at the present time.

The evaluation of dose-response relationships between app engagement and treatment outcomes has provided key insights into the effectiveness of mindfulness apps. In previous research, dosage has been operationalized as number of times the mindfulness app was opened or total number of days the app was used [25, 28]. Significant associations between more frequent app use and greater decreases in psychological distress were found in both studies. Goldberg et al. [29] examined multiple dosage operationalizations from a large RCT, including number of meditation minutes, which produced a significant dose-response relationship with psychological distress. In the current study, we were also able to assess minutes spent completing meditating, as well as the number of exercises completed, and used these variables to examine engagement and dose-response associations.

The Present Study

In the current study, we deployed a commercially available mobile health platform called AmDTx containing a variety of mindfulness-based meditation practices. AmDTx has received support for its acceptability and potential efficacy in samples of university students, cancer survivors, and post-concussion adolescents [30–33]. Single-arm trials, as well as placebo-controlled RCTs of the app have typically focused on an acute intervention period of 3 to 16 weeks with results indicating improved quality of life; increased mindful acceptance and awareness, attentional control; and reduced anxiety and stress as treatment outcomes. The present study extended the evaluation of AmDTx in a large sample of individuals with mental health difficulties seeking psychological treatment. Our overall goals were to understand the attrition and adoption rates, acceptability (e.g., credibility, usability), engagement in an integrated clinical context, as well as its potential efficacy in reducing common mental health symptoms (e.g., functional disability, depression, and anxiety). Funding for the study was directly related to a call for hospitals, clinicians, and Canadian health technology companies to partner and evaluate technologies in a real-world setting and to explore their broad adoption into health systems. The study contributes further to the mindfulness app literature by reporting findings from a highly pragmatic investigation with strong generalizability to other real-world clinical settings.

We hypothesized that participants (H1a) would rate the intervention (AmDTx) as credible and acceptable based on established self-report measures and (H1b) that study attrition (defined as non-completion of follow-up assessments) would be below 39% across the acute and follow-up periods, given this was a large study in a more general patient population (as per Linardon [21]). With regards to engagement, we hypothesized (H2a) that AmDTx would be adopted (defined as completing at least one activity) by at least 62% of participants during the 4-week intervention period and that (H2b) participants would spend at least 2 hours meditating (as per Linardon [21]). Because so few studies report actual meditation practice and engagement over time, we also report how many participants meditated (defined as recording at least one minute of meditation) and how many participants were still active during the follow-up period. In terms of treatment outcomes, we hypothesized that (H3a) symptoms of functional disability (primary outcome), depression, anxiety, stress, and rumination would decrease over the 12-week trial. In contrast, we hypothesized that (H3b) mindful awareness and acceptance would increase over the trial. Finally, we hypothesized (H4) that a dose-response relationship would be found, whereby those who completed more exercises and accumulated more mindfulness minutes would be associated with larger improvements in functional disability using a moderation analysis.

Study Design and Recruitment

The study was a single arm, unblinded clinical trial lasting for 12 weeks. The trial was registered with the ClinicalTrials.gov database (NCT05211960) on January 26, 2022 (revised September 6, 2022) and consisted of a 4-week acute intervention period followed by an 8-week follow-up period. Enrollment ran from February 2022 to December 2023. Participants were recruited from a variety of hospital- and community-based sources (e.g., clinician referrals, waitlist and research registries, poster advertisements near hospitals, hospital and other websites, community postings on Kijiji and other advertisement websites). All advertisements sought individuals who were “waiting for psychological services” and were in the province of Ontario. All prospective participants were informed about the study and pre-screened for eligibility over the phone before enrollment.

Inclusion criteria were: (a) 18 years of age and older; (b) fluency in English; (c) ability to provide consent; (d) waiting for psychological services; and (e) own a device capable running the MBI app, AmDTx. We implemented a broad definition of waiting for psychological services, including waiting for a referral, assessment, or treatment with a clinician (e.g., psychiatrist, primary care physician, mental health professional, and/or waiting for acute treatment. Current low intensity treatment (e.g., support groups, sessions once per month) was permitted. This inclusion criterion thus precluded active participation in weekly or biweekly psychological treatment. Exclusion criteria were any known factors precluding participation (practical, acute psychiatric condition, or serious medical condition), or participation in another treatment or intervention study.

Procedure

Eligible participants attended a 1-hour baseline session via video conference, where each participant completed a brief demographic interview and questionnaires, were connected to AmDTx¹, and were provided a detailed overview of its features, including a two-page tip sheet. Participants were advised to use the app routinely, with a suggested minimum usage of 3–4 times per week for 15–20 minutes. Participants then completed follow-up questionnaires at 2, 4, 8, and 12 weeks. Honoraria was up to $30 CDN, provided at the end of the study in the form of an e-gift card, and prorated according to how many follow-up questionnaires were completed. An in-kind incentive was the free access to AmDTx for one year, which had a value of $99 CDN. Participants received $21.88 CDN (SD = 12.65) on average.

Intervention

AmDTx (previously known as Wildflowers and Am Mindfulness) is a digital health platform developed by Mobio Interactive, Inc. (Toronto, Canada). Its main mindfulness app component has, or is current being, evaluated in at least five clinical trials [30–34]. AmDTx supports personalized mindfulness practice through guided meditations, audio lessons, reminders, a timer to facilitate self-guided meditation, journaling features, and psychobiometric recordings and feedback (see Fig. 1 for sample screenshots). The app is currently available in four languages and can be downloaded from the Android and Apple app stores. Using a unique code enabling one year of free access, participants received access to all the therapeutic content. Participants received a suggested list of several sequential guided journeys within the app that would allow an introduction and gradual learning of mindfulness practice (e.g., psychoeducation, guided meditations) that were developed with mindfulness experts. We encouraged each participant to set regular reminders in the app during our baseline session, supplemented by regular push notifications (with motivational messages) sent by the app itself to promote engagement.

Measures

Demographic Information

A brief interview was used to collect pertinent demographic information for each participant (e.g., age, sex, gender, ethnoracial identity, education level, history of inpatient hospitalization, etc.) at the baseline session.

Credibility and Acceptability Measures

The 6-item Credibility and Expectancy Scale [35] (α = .88) was given at baseline to assess whether participants had favorable opinions of the intervention and its potential effectiveness before starting treatment. The first three items were used to evaluate credibility (using a 9-point Likert scale from 1 to 9), whereas a single item was used to evaluate expectancy of clinical improvement (using an 11-point Likert scale from 0–100%).

The 6-item Treatment Acceptability Questionnaire [36] (Ω = .84) was administered at follow-up to assess perceived acceptability, safety, and trustworthiness using a 7-point Likert scale from 1 to 7. A total score is created between 7 and 42. As additional metrics of acceptability, we assessed how many individuals used the app at all, how many engaged in meditation exercises, and how many users were active (indicated by any completed activities on the app) within 4-week intervals.

Engagement and Usability Measures

To assess app engagement, we used the number of exercises completed, lifetime “points”

(a metric reflecting the number of completed activities), and number of meditation minutes.² To

understand usability, we used the 18-item mHealth App Usability Questionnaire [37] (Ω = .91) to assess ease of use, interface and satisfaction, and usefulness of the mindfulness app, assessed at each follow-up. This scale uses a 7-point Likert scale from 1 (strongly agree) to 7 (strongly disagree), with lower scores suggesting better ratings.

Treatment Outcome Measures

The following self-report measures were administered at baseline and each follow-up. Our original protocol excluded assessments of stress and functional disability at Week 2; however, a revised protocol approved at 9 months added them to Week 2 to permit evaluation of early change.

The World Health Organization Disability Assessment Schedule 2.0 [38] is a 12-item self-report measure (Ω = .96) assessing functional disability over the past month in several domains (cognition, mobility, self-care, getting along with others). This measure served as the primary outcome. Items were rated on a Likert scale of 0 (none or no disability) to 4 (extreme or cannot do) and are summed to create a total score, with higher scores indicating more functional disability.

The Patient Health Questionnaire-9, Depression subscale [39] is a 9-item self-report measure used to assess depressive symptoms over the past two weeks, with excellent internal reliability (Ω = .94) and clinical utility in predicting depression. Items are rated 0 (not at all) to 3 (nearly every day) and summed to create a total score, with higher scores

indicating greater severity in symptoms.

The Generalized Anxiety Disorder-7 Scale [40] is a 7-item self-report measure used to assess generalized anxiety symptoms over the past two weeks, with excellent internal reliability (Ω = .94). Items are rated 0 (not at all) to 3 (nearly every day) and summed to create a total score, with higher scores indicating greater severity in symptoms.

The Perceived Stress Scale [41] is a well-known 10-item scale of perceived stress in reaction to daily events, with excellent internal reliability. Items are rated 0 (never) to 4 (very often) and a total score is created, with higher scores indicating greater severity of perceived stress. One poorly performing item (item #3) was removed from analysis. After doing so, reliability improved from Ω = .39 to .87 in the current study.

The Ruminative Response Scale [42] is a 10-item scale (revised from a 22-item version; Ω = .88) that assesses the tendency to engage in ruminative thinking, often associated with depression. Items are rated from 1 (almost never) to 4 (almost always) and are summed to create a total score, with higher scores indicating greater severity of rumination.

The Philadelphia Mindfulness Scale [43] is a 20-item scale that measures two components of mindfulness: the tendency to be highly aware of one’s experiences (i.e., awareness; Ω = .90) and taking an accepting, nonjudgmental stance (i.e., acceptance; Ω = .90). Items are rated from 1 (never) to 5 (very often), with odd items summed for the awareness subscale and even items reversed and summed for the acceptance subscale. Higher scores on each subscale indicate higher dispositional levels of mindful awareness and acceptance.

Data Preparation

Out of a maximum of 965 baseline and follow-up surveys, a total of 730 (75.6%) were available and used for analyses. After the baseline assessment, participants completed or attempted an average of 2.80 (SD = 1.49) follow-up assessments (Week 2: 79.3%, Week 4: 69.9%, Week 8: 58.5%, and Week 12: 68.4%). Imputation was not used at the scale level as all data was complete. The number of missing summary scores varied depending on the measure (0 to 20 for depression, anxiety, rumination; mindfulness; 44 for stress, and 59 for functional disability) and as noted above, was substantially influenced by lack of administration of Week 2 stress and disability measures for the first 55 participants. For descriptive analyses, summary scores were calculated using listwise deletion. For longitudinal analyses, we utilized an intent-to-treatment approach and used restricted maximum likelihood (REML) estimation which utilizes all available data. For all tests, we planned two-tailed statistical tests with an alpha significance level was p < .05.

Statistical Analyses

Descriptive statistics were used to summarize demographic, attrition, acceptability, and engagement data. Because we were especially interested in understanding how many participants not only used the app but also engaged in meditation, we used the “points” variable to calculate how participants completed any activities at all (indicated by a score > 0). Then we used the meditation minutes variable to calculate how many participants completed meditation (indicated by a score > 0 minutes) and was also later used to subgroup those who meditated from the overall sample. For ease of interpretation, we converted meditation minutes to hours by dividing by 60. Cumulative engagement data (including the number of exercises) for each participant was summarized at the 4-, 8-, and 12-week time points to align with study follow-up assessments. Number of exercises, points, and mindfulness minutes were all highly correlated with each other at Week 4 (rs > .77, p < .0001).

All other statistics were run in the statistical program R, Version 4.2.1 (2022-06-23). To evaluate the internal consistency, we calculated Cronbach alpha [44] for single-use measures and between-person omega reliability for repeated measures [45], the latter of which was calculated using “omegaSEM” from the MultilevelTools package (Version 0.1.1). For treatment outcomes, we ran a series of linear mixed models with the lme4 package (Version 1.1–26) [46] with each outcome variable serving as a dependent variable in separate models. As recommended in clinical trials [47], we adjusted each model by incorporating the baseline dependent variable recorded for each person, except when the variable was not measured at baseline (e.g., acceptability). Covariates were added to assess possible differences in treatment response based on demographic variables and included: age, sex (1 = female, 2 = male), gender³ (1 = woman, 2 = man, 3 = gender diverse), history of inpatient hospitalization (1 = yes, 0 = no), and current antidepressant medication (1 = yes, 0 = no). To support interpretation of results and protect participant privacy, answers such as “prefer not to answer” or “do not know” or another low frequency responses (< 3%) were recoded as missing. Time was entered as a continuous variable measured in days and divided by 7 to calculate Weeks (rounded to nearest one) since baseline, allowing us to account for variability in survey completion dates between participants on specific weeks.

All linear mixed models included a random intercept for person and assessed whether the inclusion of random slopes for Week (linear and quadratic) improved model fit (tested one by one) using a diagonal variance-covariance structure. However, no final model included the quadratic random intercept due to lack of convergence or lack of improved model fit. All model comparisons were evaluated using maximum likelihood estimation with the lmerTest (Version 3.1–3) [48] package, which uses Satterthwaite's degrees of freedom method. An optimizer called optimx (Version 2022–4.30) set to the “nlminb” method (i.e., Nonlinear Minimization subject to Box Constraints) was used in all models. The outputs of each linear mixed model are reported in Supplementary Materials for transparency and contain unstandardized effects. To examine dose-response effects, we again ran linear mixed models with our primary outcome serving as the dependent variable predicted by fixed effects of time in Weeks (linear and quadratic), covariates for the outcomes at baseline that were significant, and interactions for each engagement variable with time in Weeks. Engagement variables were tested as moderators (grand mean centered) and included the maximum number of exercises and total meditation hours achieved by each participant up to 12 weeks.

Power

We estimated that a small effect size (f = 0.10) could be detected with power of .80 and α of 0.5 with a sample size of 125 using G*Power 3.1.9.2 [49]. With up to 45% attrition possible, we proposed a sample size recruitment of 200 a priori, so that this study would be well-powered for primary analyses reported here as well as secondary mediation analyses.

Enrollment Sample

As seen in Fig. 2, 357 participants were screened for eligibility and 193 were deemed eligible and completed all aspects of the baseline session with the research team. As seen in Table 1, the intent-to-treat (ITT) sample was 34.3 years old on average (SD = 11.5) and was predominantly female (70.5%), woman (66.8%), heterosexual (61.1%), and either White North American (37.3%) and/or White European (35.2%). At baseline, participants reported moderate severity symptoms of depression and anxiety (see Table 2).

Table 1

Demographic Characteristics of the Intent-to-Treat Sample (N = 193)
Characteristic	M (SD) / N (%)	Characteristic	N (%)
Age	34.3 (11.5)	Level of Education
		High school or GED	20 (15.0%)
Sex		Some college/university	32 (16.6%)
Male	55 (28.5%)	Associate degree/technical certificate	5 (2.6%)
Female	136 (70.5%)	College diploma	29 (15.0%)
Other response	2 (1.0%)	Bachelor’s degree	66 (34.2%)
		Master’s/professional degree	24 (12.4%)
Gender¹		Other response	8 (4.1%)
Man	50 (25.9%)
Woman	129 (66.8%)	Ethnicity¹
Non-binary	7 (3.6%)	East Asian	22 (11.4%)
Other response	11 (5.7%)	South East Asian	9 (4.7%)
		South Asian	17 (8.8%)
Sexual Orientation¹		Black African	6 (3.1%)
Heterosexual/straight	118 (61.1%)	Black North American	6 (3.1%)
Gay	7 (3.6%)	Black Caribbean	5 (2.6%)
Bisexual	25 (13.0%)	Indian-Caribbean	5 (2.6%)
Queer	9 (4.7%)	Latin American	8 (4.1%)
Pansexual	10 (5.2%)	Middle Eastern	10 (5.2%)
Not sure or questioning	6 (3.1%)	White European	68 (35.2%)
Prefer not to answer	10 (5.2%)	White North American	72 (37.3%)
Other response	10 (5.2%)	Mixed Heritage	7 (3.6%)
		Other response	9 (4.7%)
Inpatient Hospitalization History
Yes	62 (30.2%)	Current Psychotropic Medication
No	143 (69.8%)	None	71 (34.6%)
		1 Class	73 (35.6%)
Employed		2 Classes	35 (17.1%)
Yes (Part-time or Full-time)	131 (63.9%)	3 Classes	19 (9.3%
No	72 (35.1%)	4 or more Classes	7 (3.4%)
Prefer not to answer	2 (1.0%)	Antidepressant Class Currently	102 (49.8%)
Note: M = Mean; N = Number; SD = Standard Deviation. ¹ Participants could select multiple options for Gender, Sexual Orientation, and Ethnicity. Other responses options were: Sex (Intersex, Prefer not to answer), Gender (Trans girl/woman, Genderqueer, Genderfluid, Not listed, and Prefer not to answer), Sexual orientation (Lesbian, Asexual, Not listed), Ethnicity (First Nations, Indigenous, Metis, Prefer not to answer, Not listed), Level of education (Grade 8 but not high school, Doctoral degree).

Table 2

Descriptive Statistics of Repeated Measure Outcomes Over the Trial Period
	Baseline		Week 2		Week 4		Week 8		Week 12
	N	M (SD)	N	M (SD)	N	M (SD)	N	M (SD)	N	M (SD)
Treatment Acceptability and Usability
Acceptability	--	--	146	32.65 (5.76)	128	32.77 (6.21)	90	33.11 (6.39)	96	32.76 (6.67)
Ease of Use	--	--	146	2.66 (1.12)	128	2.82 (1.29)	107	2.69 (1.37)	129	2.66 (1.38)
Interface and Satisfaction	--	--	146	2.86 (1.09)	128	3.02 (1.28)	107	2.96 (1.33)	129	2.85 (1.33)
Usefulness	--	--	146	3.11 (1.04)	128	3.09 (1.12)	107	3.02 (1.20)	129	3.00 (1.31)
Primary Outcome
Functional Disability	191	18.03 (10.10)	112	15.79 (10.69)	129	13.95 (9.25)	109	15.06 (10.53)	130	13.78 (9.48)
Secondary Outcomes
Depression	193	12.43 (6.67)	153	10.15 (5.92)	137	9.39 (5.24)	114	9.31 (6.05)	133	8.48 (5.61)
Anxiety	193	10.58 (5.42)	153	8.80 (5.07)	135	8.04 (5.11)	113	7.56 (5.35)	132	7.26 (5.53)
Perceived Stress	193	20.60 (6.44)	116	18.94 (6.03)	133	17.94 (5.79)	112	17.86 (6.04)	132	17.28 (6.49)
Rumination	192	25.32 (6.20)	150	24.15 (6.12)	133	23.08 (5.78)	112	22.69 (5.77)	132	22.64 (5.60)
Mindfulness Awareness	191	37.24 (7.72)	146	36.75 (7.15)	130	37.12 (7.07)	112	37.46 (7.54)	131	38.22 (7.19)
Mindfulness Acceptance	191	24.18 (7.10)	146	25.40 (6.95)	130	26.90 (6.64)	112	27.47 (6.73)	131	27.22 (6.54)

Note: M = Mean; N = Number; SD = Standard Deviation. There was fluctuation in completion of certain measures, largely due to protocol changes during the study that impacted the frequency of those assessments. Depression was the first measure presented in the questionnaire package and therefore represents the maximum number of observations (completed or partially completed) captured in each linear mixed model.

Credibility and Acceptability

At baseline, credibility of the intervention was rated positively (M = 6.43; SD = 1.60) suggesting that participants thought the intervention was logical, was likely to raise their quality of functioning, and were confident in recommending it to another person. Participants expected a 47.13% (SD = 22.47) improvement in symptoms on average at baseline. As seen in Table 2, follow-up ratings of treatment acceptability were relatively high throughout the trial, ranging from 32.65 to 33.11 (SDs = 5.76 to 6.39) and did not change over time (ps > .51) while controlling for covariates (see Supplementary Materials, Table S1). Participants rated ease of use the most positively over the course of the trial, followed by the interface and satisfaction, and usefulness. Ease of use and usefulness did not change over time while controlling for covariates (ps > .10; see Supplementary Materials, Tables S2-S4); however, interface and satisfaction decreased over the trial (indicated by increasing scores; b_linear = .10, p = .005, 95% CI [.03, .17]; b_quadratic = − .01, p = .003, 95% CI [-.01, − .002]).

Engagement

Of the 193 ITT individuals who registered themselves on AmDTx, 24.9% did not complete any activities (producing 0 “points” overall) within the first 4 weeks, indicating they either did not open the app, or opened it but did not interact in potentially therapeutic ways. Meanwhile, 42.0% did not complete any mindfulness activities (producing 0 “mindfulness minutes”) within the same period. Table 3 summarizes cumulative engagement on AmDTx at Weeks 4, 8, and 12 for the ITT sample and the subgroup of users who completed at least one minute of mindfulness (i.e., meditators). Within the first 4 weeks, the ITT sample spent an average 1.09 hours meditating (Median = .18 hours) and completed about 9.16 exercises (Median = 2) within the first 4 weeks. When we restricted the analyse to those who recorded at least one minute of meditation, these users recorded 1.87 hours of meditation (Median = .83) and 15.55 exercises (Median = 8) within the same period. Participants who stayed active following 4 weeks (32.1% of ITT) and 8 weeks (20.2% of ITT) reported substantial increases in meditation hours and number of exercises.

Table 3

Descriptive Engagement Statistics for Intent-to-Treat Sample and Subsample of Meditators
	ITT			Meditators
	N	M (SD)	Median (Range)	N	M (SD)	Median (Range)
Cumulative Data at 4 weeks
Meditation Hours	193	1.09 (2.07)	.18 (0-13.47)	112	1.87 (2.43)	.83 (1-13.47)
Number of Exercises	193	9.16 (15.63)	2.00 (0–92)	112	15.55 (17.98)	8.00 (1–92)
Cumulative Data at 8 weeks*
Meditation Hours	68	4.14 (4.67)	2.38 (0-26.58)	62	4.54 (4.71)	2.96 (.05-26.58)
Number of Exercises	68	32.28 (34.18)	20.50 (0-187)	62	35.21 (34.40)	26.00 (1-187)
Cumulative Data at 12 weeks**
Meditation Hours	40	7.42 (7.35)	6.54 (0-33.18)	39	7.61 (7.34)	7.33 (.18-38.18)
Number of Exercises	40	53.63 (50.62)	45.00 (0-233)	39	55.00 (50.52)	46.00 (4-233)

Note: ITT = Intent-to-treat. M = Mean; N = Number; SD = Standard Deviation. *Based on users reporting any completed activity after 4 weeks. **Based on users reporting any completed activity after 8 weeks.

Changes in Outcomes Over Time

Throughout the analyses reported here, there were few significant effects of any demographic or clinical covariates other than baseline scores (see Supplementary Materials, Tables S5-S11). Controlling for baseline symptoms and covariates, our primary outcome of functional disability decreased from baseline to Week 12 (b_linear = -2.84, p < .001, 95% CI [-1.03, − .52]; b_quadratic = .34, p = .001, 95% CI [.02, .06]). Current antidepressant use was associated with higher functional disability throughout the trial (b = .94, p = .04, 95% CI [.04, 1.84]). Similarly, secondary mental health outcomes decreased over the course of the trial, including depression (b_linear = − .59, p < .0001, 95% CI [-.77, − .42]; b_quadratic = .02, p < .0001, 95% CI [.02, .04]), anxiety (b_linear = − .54, p < .0001, 95% CI [-.70, − .38]; b_quadratic = .02, p < .0001, 95% CI [.01, .04]), and perceived stress (b_linear = − .53, p < .0001, 95% CI [-.72, − .34]; b_quadratic = .03, p = .0002, 95% CI [.02, .04]). Finally, when controlling for baseline scores and covariates, rumination decreased (b_linear = − .48, p < .0001, 95% CI [-.68, − .28]; b_quadratic = .02, p = .001, 95% CI [.01, .04]), whereas mindful acceptance increased (b_linear = .68, p < .0001, 95% CI [.47, .90]; b_quadratic = − .04, p < .0001, 95% CI [-.05, − .02]) from baseline to Week 12. This effect was not supported for mindful awareness over the course of the trial (b_linear = − .12, p = .29, 95% CI [-.68, − .28]; b_quadratic = .02, p = .05, 95% CI [.01, .04]). All significant time effects were steeper in the early phase of the trial, followed by plateauing over time.

Dose-Response Associations between Engagement and Primary Outcome

As seen in Table 4, when the entire ITT sample was included, there was a significant Weeks (and Weeks²) × Meditation Hours interaction in the prediction of functional disability, over and above baseline covariates and main effects for time in Weeks. The interaction indicated that a higher number of hours spent meditating was significantly associated with greater reductions in functional disability over time (see Fig. 3a). This association was also found in the subgroup of meditators and was larger in magnitude. As a sensitivity analysis we also tested the number of fully completed exercises using an interaction term (Weeks [and Weeks²] × Number of Exercises) predicting functional disability over and above baseline covariates and main effects for time in Weeks. As with meditation hours, a greater number of completed exercises was associated with greater reductions in functional disability over the course of the trial (Fig. 3b). These associations were stronger when we examined the same effect in our subgroup of meditators.

Table 4

Dose-Response Models Predicting Functional Disability over Time for Intent-to-Treat and Subsample of Meditators
	ITT				Meditators
	b	SE	p	95% CI	b	SE	p	95% CI
Meditation Hours
Intercept	17.05	.35	< .0001	16.37, 17.73	17.17	.45	< .0001	16.30, 18.04
Weeks (linear)	− .73	.13	< .0001	− .98, − .47	− .53	.14	.0003	− .81, − .25
Hours	− .002	.06	.97	− .11, .11	− .001	.06	.98	− .12, .12
Weeks² (quadratic)	.04	.009	< .0001	.02, .06	.02	.01	.02	.004, .04
Baseline functional disability	.90	.02	< .0001	.86, .94	.91	.03	.19	.86, .96
Current Antidepressant	.92	.44	.04	.06, 1.77	.70	.53	.19	− .34, 1.75
Weeks × Hours	− .06	.03	.04	− .11, − .003	− .09	.03	.0009	− .14, − .04
Weeks² × Hours	.004	.002	.06	− .0002, .007	.006	.002	.002	.002, .01
Number of Exercises
Intercept	17.10	.35	< .0001	16.37, 17.74	17.20	.44	< .0001	16.34, 18.06
Weeks (linear)	− .72	.13	< .0001	− .98, − .47	− .53	.14	.0002	− .80, − .25
Exercises	− .001	.008	.87	− .02, .01	− .001	.009	.87	− .02, .02
Weeks² (quadratic)	.04	.009	< .0001	.02, .06	.02	.01	.02	.004, .04
Baseline functional disability	.90	.02	< .0001	.86, .94	.91	.03	< .0001	.86, .97
Current Antidepressant	.91	.44	.04	.05, 1.76	.66	.53	.21	− .37, 1.69
Weeks × Exercises	− .009	.004	.02	− .02, − .001	− .14	.04	.0002	− .02, − .007
Weeks² × Exercises	.0006	.0003	.04	.00004, .001	− .001	.0003	.0004	.0004, .002

Note: ITT = Intent-to-treat; SE = standard error.

Ancillary Analyses

We ran analyses stratified by sex to explore potential differences in variables of interest and/or treatment outcome (see Supplementary Materials, Tables S12-S13). At baseline, the only significant difference was that males were significantly older than females. Both females and males reported significant decreases in functional disability during the trial, but there was a significant association between current antidepressant use and higher functional disability in females and not males. There were no sex differences with respect to maximum meditation hours and maximum number of exercises completed over the course of the trial.

The aim of the present study was to understand how people waiting for psychological services would interact with a third-party mobile health platform when made available at no expense. This study was the first large scale study of AmDTx integrated into a health system, as a low-intensity tool for treatment-seeking adults with mental health concerns. The present analyses report on the acceptability, engagement, and treatment outcomes.

Our first hypothesis was largely supported. Participants rated AmDTx as credible and acceptable, expecting a 47% symptom improvement on average. While usability ratings varied depending on the subscale, the average rating was consistently on the positive versus negative end of the rating scale. These findings extend the available literature on the credibility and acceptability of mindfulness apps as a method of low-intensity treatment for mental health symptoms [10–12, 18]. Attrition was roughly 30% during the acute 4-week intervention period and varied during later timepoints but generally stayed below the expected average reported in a review of mindfulness app RCTs (i.e., upwards of 39%) in studies where larger samples, like ours (n > 100), and more general population targets are recruited. Our protocol incorporated strategies to reduce attrition, such as providing some monetary incentives, requiring human interaction during enrollment, and sending reminders to complete exercises, which were all associated with less attrition in previous research [21–23].

The second hypothesis was partially supported. The adoption rate was 75.1% of the ITT sample, which was above an average of results reported by 11 mindfulness app RCTs [21]. In fact, our adoption rate is even more conservative given that we examined how many participants completed at least one meaningful activity on the app, whereas most studies reported the adoption rate as those who downloaded or accessed/opened the app, regardless if used for the intended purpose [50–52]. Thus, given the opportunity to use AmDTx, most people waiting for treatment did in fact use it within a 4-week period (and used it as intended). With respect to engagement hours, our ITT sample spent 1.09 hours on average meditating in the first 4 weeks, which increased to 1.87 hours on average when we looked at those who recorded at least one minute of meditation. These findings are well within the range of previous mindfulness app trials that reported time spent meditating, from .45 hours [50] to 3.08 hours [53]. Larger trials (n > 75) of various lengths reported similar engagement rates as the current study, such as 1.70 hours in the general population [28], 2.13 hours in school employees [54], and 1.50 hours in university students [25, 55]. Number of exercises completed was also comparable with studies that reported this information [28, 54], although it is notable that for our trial with AmDTx, we only tracked the number of fully completed exercises, excluding mindfulness exercises initiated but not completed. This stringent tracking protocol may differ from other apps. Importantly, time spent meditating and the number of exercises completed increased as participants continued to use AmDTx, with those who reported using the app longer than 8 weeks recording roughly 7.5 hours on average. Thus, for a not-insignificant portion of our sample, engagement was very high and could be comparable to that of a standard meditation course (e.g., 30 min in a weekly face-to-face treatment lasting 14 weeks).

It was clear from our pattern of adoption and attrition that the greatest dropout occurred within the first two weeks of the trial, which may be a critical period to target when addressing strategies to further improve engagement and retention. Nevertheless, 32.1% were still active after 4 weeks, and 20.2% were still active after 8 weeks, which is comparable to a study by Flett et al. [25] who reported that 32.3% of undergraduate students assigned to the intervention arm were still using a popular mindfulness app after 4 weeks and 25.0% after 6 weeks. Theories about adoption determinants tend to revolve around usability, user-centric design, privacy concerns, and poor integration with other healthcare services [13, 56]. Adjustments to the study procedures in future research could include additional check-ins and better educating patients on the potential impact of using the tool regularly for their mental health and ultimate prognosis. While we did not conduct exit interviews with our participants or collect qualitative feedback, previous mixed methods work of that nature with the same app has been positive [30, 31]. In addition, reasons for using or not using a mindfulness app may be confounded with treatment outcomes [29]; some participants may use it more because they are having difficulties and use it less when their difficulties cease, consistent with the psychotherapy where participants tend to discontinue treatment as their symptoms decrease [57]. Therefore, adoption and engagement on an app can be impacted by both positive and negative treatment outcomes.

Our third hypothesis was also largely supported, extending previous findings from mindfulness app trials, with the novel observation that use of AmDTx in this clinical context is associated with improvements in patients’ functional disability and decreases in their tendency to ruminate. We also confirmed decreases in depression, anxiety, and stress over the 12-week trial, results consistent with a growing literature concerning mindfulness apps [18, 20, 25, 52, 58], including those that used AmDTx in different target samples [30, 33]. Most covariates assessed were not significant, suggesting that treatment outcomes were consistent among different demographic features. Nevertheless, while mindful acceptance increased over the trial period, there was no significant change for mindful awareness. One review that assessed changes in psychological processes within RCTs of mindfulness apps suggests that significant mindful awareness increases are common but not always found [20]. These studies had control groups, however, and most used different measures compared to the present study.

Finally, in support of our fourth hypothesis, we found significant dose-response associations in our ITT sample, consistent with other studies using different operationalizations of dosage [25, 28, 29]. Participants who recorded a greater number of exercises or a greater number of meditation hours also reported larger decreases in functional disability over the course of the trial. These effects were larger in sensitivity analyses using a subsample who recorded at least one minute of meditation. The dose-response findings suggest that improvement in functional disability was driven moreso by engagement with therapeutic aspects of the app as opposed to other factors, such as additional health care provider appointments or other low-intensity treatments.

Limitations

While our results demonstrate potential utility and clinical benefit from integrating clinically validated mobile mindfulness app into a health system waitlist context, this study was an uncontrolled, unblinded, single-arm clinical trial. Therefore, the treatment outcomes may overestimate (or underestimate) the true effect of AmDTx to inactive (i.e., waitlist) or more active control conditions (e.g., non-therapeutic app or placebo control). Similarly, our dose-response analysis was self-selected. An improved design might be an RCT that directly manipulates dosage of mindfulness practice without choice from the participant [29]. In addition, we experienced some attrition and non-completion of outcome measures (including a problematic item on our stress measure), which may have also contributed to bias in understanding changes over time and dose-response analyses. There were some technical and administration issues which led to some participants having incomplete data and/or app accounts at baseline; however, we avoided substantial data loss. While we excluded those who were participating in other psychological treatments at baseline (e.g., cognitive behavioral therapy), it is possible that some participants received access to unreported medical services during the intervention or follow-up period. That is, we did not restrict access to services that became available to patients. We also did not restrict access to other low-intensity services, such as support groups, psychiatrist appointments, or self-help resources. By not restricting or more controlling for these factors, additional bias and confounds may have been introduced.

Additionally, while our recruitment procedures were as inclusive and diverse as possible, the final sample may have differed from the broader population, impacting generalizability. For example, we had a higher proportion of female participants than males, and relatively young sample compared to the national average. There are also relationships between demographic variables and attitudes towards mental health services, which may have impacted who presented for our study in the context of treatment-seeking behavior [59]. Additional trials of AmDTx with strategic over-sampling of local minorities and underserved populations are underway. Finally, AmDTx is not just a mindfulness app, but a broader mobile health platform that performs remote monitoring via momentary assessments, questionnaires, and an objective measure of psychological stress [60]. Our analyses did not consider these data, neither as direct measures of patient wellbeing, nor for their potential relevance within clinical decision making. These limitations are mitigated by the context of delivery, reflecting the real-world expectations for integration of digital tools like AmDTx into health systems.

Collectively, the findings provide support for the integration of clinically validated low-intensity digital interventions to alleviate unmet service needs caused by waitlists. Significant portions of participants in this study adopted the mindfulness app, engaged with it meaningfully, and achieved therapeutic benefit, with positive associations between app use and reduced functional disability. Results support the use of mindfulness apps, including AmDTx, as effective interventions for those waiting (or not receiving) psychological services and perceptions of them are positive, with high ratings on credibility, acceptability, usability, and adoption. Engagement and retention may be bolstered through broader integration within the health care system, including system-wide implementation procedures that reinforce the use of mindfulness apps while incorporating the obtained date remotely into clinical decision making.

Ethical Approval and Consent to Participate: 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. The study was approved by the Research Ethics Board the Centre for Addiction and Mental Health (#088/2021).

Availability of Data and Materials: Data for the current study can be requested from the first author.

Competing Interests: BJS is the Chief Scientist and CEO of Mobio Interactive Inc., and he owned approximately 23% of the company at the time of this study. SS is a data scientist employed by Mobio Interactive Inc., with fewer that 0.1% in company stock options. BJS and SS served as technical liaisons for the study and did not contribute to study design, selection of primary outcome, data collection, or analysis. No other authors have connections to Mobio Interactive Inc. None of the authors of this study received financial compensation or any other form of compensation for the research undertaken herein. Mobio Interactive Inc. did, however, contribute in-kind funds (e.g., one-year licenses, staff, technical resources) to help fund the study as mandated by the funding agency.

Funding: The study was funded by the Ontario Bioscience Innovation Organization as part of the Early Adoption Health Network program, Project number: E010 (https://eahn.obio.ca/).

Authors’ Contributions: Conceptualization: ARD, AP, LCQ. Methodology: ARD, AP, OO. Formal analysis and investigation: ARD, AP, LCQ. Writing - original draft preparation: ARD. Writing - review and editing: ARD, AP, OO, SS, BJS, LCQ. Funding acquisition: ARD, BJS, LCQ. Resources: LCQ, SS, BJS. Supervision: ARD, LCQ.

Acknowledgements: The authors thank the participants who took part in this study. We also acknowledge the many trainees who helped oversee this study over the course of two years: Sinan Shariff, Elizabeth Alex, Raga Sivasothy, Louie Paolo Leynes, Bea Calahong, Mahak Dhawan, Samantha Lucchetta, Megan Brndjar, Leanne Lecap, and Aaron Boughen.

Jaeschke K, Hanna F, Ali S, Chowdhary N, Dua T, Charlson F. Global estimates of service coverage for severe mental disorders: Findings from the WHO Mental Health Atlas 2017. Glob Ment Health. 2021;8:e27. https://doi.org/10.1017/gmh.2021.19
Wang PS, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Borges G, Bromet EJ, Bruffaerts R, De Girolamo G, De Graaf R, Gureje O, Haro JM. Use of mental health services for anxiety, mood, and substance disorders in 17 countries in the WHO world mental health surveys. Lancet. 2007;370(9590):841–850. https://doi.org/10.1016/S0140-6736(07)61414-7
Coulaud PJ, Jesson J, Bolduc N, Ferlatte O, Jenkins E, Bertrand K, Salway T, Jauffret-Roustide M, Knight R. Young adults’ mental health and unmet service needs in the context of the COVID-19 pandemic across Canada and France. Community Ment Health J. 2023;59(2):222–232. https://doi.org/10.1007/s10597-022-01000-1
Khattar J, Griffith LE, Jones A, De Rubeis V, de Groh M, Jiang Y, Basta NE, Kirkland S, Wolfson C, Raina P, Anderson LN. Symptoms of depression and anxiety, and unmet healthcare needs in adults during the COVID-19 pandemic: A cross-sectional study from the Canadian Longitudinal Study on Aging. BMC Public Health. 2022;22(1):2242. https://doi.org/10.1186/s12889-022-14633-4
McGorry PD, Mei C, Chanen A, Hodges C, Alvarez-Jimenez M, Killackey E. Designing and scaling up integrated youth mental health care. World Psychiatry. 2022;21(1):61–76. https://doi.org/10.1002/wps.20938
Moroz N, Moroz I, D’Angelo MS. Mental health services in Canada: barriers and cost-effective solutions to increase access. Healthc Manage Forum 2020;33(6):282–287. https://doi.org/10.1177/0840470420933911
Kowalewski K, McLennan JD, McGrath PJ. A preliminary investigation of wait times for child and adolescent mental health services in Canada. J Can Acad Child Adolesc Psychiatry. 2011;20(2):112–119. PMID: 21541100
Furukawa TA, Noma H, Caldwell DM, Honyashiki M, Shinohara K, Imai H, Chen P, Hunot V, Churchill R. Waiting list may be a nocebo condition in psychotherapy trials: A contribution from network meta-analysis. Acta Psychiatrica Scandinavica. 2014;130(3):181–92. https://doi.org/10.1111/acps.12275
Moshe I, Terhorst Y, Philippi P, Domhardt M, Cuijpers P, Cristea I, Pulkki-Råback L, Baumeister H, Sander LB. Digital interventions for the treatment of depression: A meta-analytic review. Psychol Bull. 2021;147(8):749–786. https://doi.org/10.1037/bul0000334
Linardon J, Cuijpers P, Carlbring P, Messer M, Fuller-Tyszkiewicz M. The efficacy of app‐supported smartphone interventions for mental health problems: A meta‐analysis of randomized controlled trials. World Psychiatry. 2019;18(3):325–36. https://doi.org/10.1002/wps.20673
Pauley D, Cuijpers P, Papola D, Miguel C, Karyotaki E. Two decades of digital interventions for anxiety disorders: a systematic review and meta-analysis of treatment effectiveness. Psychol Med. 2023;53(2):567–79. https://doi.org/10.1017/S0033291721001999
Weisel KK, Fuhrmann LM, Berking M, Baumeister H, Cuijpers P, Ebert DD. Standalone smartphone apps for mental health—a systematic review and meta-analysis. NPJ Digit Med. 2019;2(1):118. https://doi.org/10.1038/s41746-019-0188-8
Lim CT, Fuchs C, Torous J. Integrated digital mental health care: a vision for addressing population mental health needs. Int J Gen Med. 2024 Dec 31:359–65. https://doi.org/10.2147/IJGM.S449474
Kabat-Zinn J. Mindfulness-based interventions in context: Past, present, and future. Clin Psychol. 2003;10(2):144–156. https://doi.org/10.1093/clipsy.bpg016
Goldberg SB, Riordan KM, Sun S, Davidson RJ. The empirical status of mindfulness-based interventions: A systematic review of 44 meta-analyses of randomized controlled trials. Perspect Psychol Sci. 2022;17(1):108–30. https://doi.org/10.1177/1745691620968771
Gál É, Ștefan S, Cristea IA. The efficacy of mindfulness meditation apps in enhancing users’ well-being and mental health related outcomes: a meta-analysis of randomized controlled trials. J Affect Disord. 2021;279:131–42. https://doi.org/10.1016/j.jad.2020.09.134
Tan ZY, Wong SH, Cheng LJ, Lau ST. Effectiveness of mobile-based mindfulness interventions in improving mindfulness skills and psychological outcomes for adults: a systematic review and meta-regression. Mindfulness. 2022;13(10):2379–95. https://doi.org/10.1007/s12671-022-01962-z
Linardon J, Messer M, Goldberg SB, Fuller-Tyszkiewicz M. The efficacy of mindfulness apps on symptoms of depression and anxiety: An updated meta-analysis of randomized controlled trials. Clin Psychol Rev. 2023 Dec 3:102370.
Linardon J. Can acceptance, mindfulness, and self-compassion be learned by smartphone apps? A systematic and meta-analytic review of randomized controlled trials. Behav Ther. 2020;51(4):646–58. https://doi.org/10.1016/j.beth.2019.10.002
Macrynikola N, Mir Z, Gopal T, Rodriguez E, Li S, Cox M, Yeh G, Torous J. The impact of mindfulness apps on psychological processes of change: a systematic review. npj Ment Health Res. 2024;3(1):14. https://doi.org/10.1038/s44184-023-00048-5
Linardon J. Rates of attrition and engagement in randomized controlled trials of mindfulness apps: Systematic review and meta-analysis. Behav Res Ther. 2023 Oct 14:104421. https://doi.org/10.1016/j.brat.2023.104421
Linardon J, Fuller-Tyszkiewicz M. Attrition and adherence in smartphone-delivered interventions for mental health problems: A systematic and meta-analytic review. J Consult Clin Psychol. 2020;88(1):1.https://doi.org/10.1037/ccp0000459
Torous J, Lipschitz J, Ng M, Firth J. Dropout rates in clinical trials of smartphone apps for depressive symptoms: a systematic review and meta-analysis. J Affect Disord. 2020;263:413–9.. https://doi.org/10.1016/j.jad.2019.11.167
Daros AR, Guimond TH, Yager C, Palermo EH, Wilks CR, Quilty LC. Feasibility, acceptability, and potential efficacy of a self-guided internet-delivered dialectical behavior therapy intervention for substance use disorders: Randomized controlled trial. JMIR Ment Health. 2024;11:e50399. https://doi.org/10.2196/50399
Flett JA, Conner TS, Riordan BC, Patterson T, Hayne H. App-based mindfulness meditation for psychological distress and adjustment to college in incoming university students: a pragmatic, randomised, waitlist-controlled trial. Psychol Health. 2020;35(9):1049–74. https://doi.org/10.1080/08870446.2019.1711089
Lattie EG, Cohen KA, Hersch E, Williams KD, Kruzan KP, MacIver C, Hermes J, Maddi K, Kwasny M, Mohr DC. Uptake and effectiveness of a self-guided mobile app platform for college student mental health. Internet Interv. 2022;27:100493.b https://doi.org/10.1016/j.invent.2021.100493
Mohr DC, Schueller SM, Tomasino KN, Kaiser SM, Alam N, Karr C, Vergara JL, Gray EL, Kwasny MJ, Lattie EG. Comparison of the effects of coaching and receipt of app recommendations on depression, anxiety, and engagement in the IntelliCare platform: factorial randomized controlled trial. J Med Internet Res. 2019;21(8):e13609.https://doi.org/10.2196/13609
Goldberg SB, Imhoff-Smith T, Bolt DM, Wilson-Mendenhall CD, Dahl CJ, Davidson RJ, Rosenkranz MA. Testing the efficacy of a multicomponent, self-guided, smartphone-based meditation app: Three-armed randomized controlled trial. JMIR Ment Health. 2020;7(11):e23825. https://doi.org/10.2196/23825
Goldberg SB, Kendall AD, Hirshberg MJ, Dahl CJ, Nahum-Shani I, Davidson RJ, Bray BC. Is dosage of a meditation app associated with changes in psychological distress? It depends on how you ask. Clin Psychol Sci. (in press)
Bergerot CD, Malhotra J, Bergerot PG, Philip EJ, Castro DV, Govindarajan A, Molina LN, Fuzita WH, Azeredo AC, Mota AC, Franca MV. Improved health-related quality of life (HR-QOL) with use of an online mindfulness tool in patients with metastatic renal cell carcinoma (mRCC) receiving immunotherapy (IO). J Clin Oncol. 2023; 41(6):66. https://doi.org/10.1200/JCO.2023.41.6_suppl.660
Sicard V, O’Kane K, Brown O, Butterfield L, Kardish R, Choi E, Healey K, Silverberg N, Smith AM, Goldfield G, Saab BJ. Acceptability, usability, and credibility of a mindfulness-based digital therapeutic for pediatric concussion: a mixed-method study. Digit Health. 2024;10. https://doi.org/10.1177/20552076241248296
Subnis UB, Farb NA, Piedalue KA, Speca M, Lupichuk S, Tang PA, Faris P, Thoburn M, Saab BJ, Carlson LE. A smartphone app–based mindfulness intervention for cancer survivors: Protocol for a randomized controlled trial. JMIR Res Protoc. 2020;9(5):e15178. https://doi.org/10.2196/15178
Walsh KM, Saab BJ, Farb NA. Effects of a mindfulness meditation app on subjective well-being: active randomized controlled trial and experience sampling study. JMIR Ment Health. 2019;6(1):e10844. https://doi.org/10.2196/10844
Ledoux AA, Zemek R, Cairncross M, Silverberg N, Sicard V, Barrowman N, Goldfield G, Gray C, Harris AD, Jaworska N, Reed N. Smartphone app–delivered mindfulness-based intervention for mild traumatic brain injury in adolescents: Protocol for a feasibility randomized controlled trial. JMIR Res Protoc. 2024;13(1):e57226. https://doi.org/10.2196/57226
Devilly GJ, Borkovec TD. Psychometric properties of the credibility/expectancy questionnaire. J Behav Ther Exp Psychiatry. 2000;31(2):73–86. https://doi.org/10.1016/S0005-7916(00)00012-4
Hunsley J. Development of the treatment acceptability questionnaire. J Psychopathol Behav Assess. 1992;14:55–64. https://doi.org/10.1007/BF00960091
Zhou L, Bao J, Setiawan IM, Saptono A, Parmanto B. The mHealth app usability questionnaire (MAUQ): Development and validation study. JMIR Mhealth Uhealth. 2019;7(4):e11500. https://doi.org/10.2196/11500
Ustun TB, Kostanjesek N, Chatterji S, Rehm J. Measuring health and disability: Manual for WHO Disability Assessment Schedule (WHODAS 2.0). World Health Organization; 2010.
Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606 – 13.https://doi.org/10.1046/j.1525-1497.2001.016009606.x
Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092–7. https://doi.org/10.1001/archinte.166.10.1092
Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983 Dec 1:385 – 96. https://doi.org/10.2307/2136404
Treynor W, Gonzalez R, Nolen-Hoeksema S. Rumination reconsidered: A psychometric analysis. Cog Ther Res. 2003;27:247 – 59.https://doi.org/10.1023/A:1023910315561
Cardaciotto L, Herbert JD, Forman EM, Moitra E, Farrow V. The assessment of present-moment awareness and acceptance: The Philadelphia Mindfulness Scale. Assessment. 2008;15(2):204–23. https://doi.org/10.1177/10731911073114
Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16(3):297–334. https://doi.org/10.1007/BF02310555
McDonald RP. Test Theory: A Unified Treatment. Taylor and Francis; 1999. https://doi.org/10.4324/9781410601087
Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Soft. 2014;67(1):1–48. https://doi.org/10.18637/jss.v067.i01.
Twisk J, Bosman L, Hoekstra T, Rijnhart J, Welten M, Heymans M. Different ways to estimate treatment effects in randomised controlled trials. Contemp Clin Trials Commun. 2018;10:80–5. https://doi.org/10.1016/j.conctc.2018.03.008
Kuznetsova A, Brockhoff PB, Christensen RH. lmerTest package: Tests in linear mixed effects models. J Stat Soft. 2017;82(13):1–26. https://doi.org/10.18637/jss.v082.i13
Faul F, Erdfelder E, Lang AG, Buchner A. G* Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175 – 91.https://doi.org/10.3758/bf03193146
Bartlett L, Martin AJ, Kilpatrick M, Otahal P, Sanderson K, Neil AL. Effects of a mindfulness app on employee stress in an Australian public sector workforce: randomized controlled trial. JMIR Mhealth Uhealth. 2022;10(2):e30272. https://doi.org/10.2196/30272
Kubo A, Kurtovich E, McGinnis M, Aghaee S, Altschuler A, Quesenberry Jr C, Kolevska T, Avins AL. A randomized controlled trial of mHealth mindfulness intervention for cancer patients and informal cancer caregivers: A feasibility study within an integrated health care delivery system. Integr Cancer Ther. 2019;18. https://doi.org/10.1177/1534735419850634
Huberty J, Green J, Glissmann C, Larkey L, Puzia M, Lee C. Efficacy of the mindfulness meditation mobile app “calm” to reduce stress among college students: Randomized controlled trial. JMIR Mhealth Uhealth. 2019;7(6):e14273. https://doi.org/10.2196/14273
Keng SL, Chin JW, Mammadova M, Teo I. Effects of mobile app-based mindfulness practice on healthcare workers: a randomized active controlled trial. Mindfulness. 2022;13(11):2691–704. https://doi.org/10.1007/s12671-022-01975-8
Hirshberg MJ, Frye C, Dahl CJ, Riordan KM, Vack NJ, Sachs J, Goldman R, Davidson RJ, Goldberg SB. A randomized controlled trial of a smartphone-based well-being training in public school system employees during the COVID-19 pandemic. J Educ Psychol. 2022;114(8):1895. https://doi.org/10.1037/edu0000739
Haliwa I, Ford CG, Wilson JM, Shook NJ. A mixed-method assessment of a 10-day mobile mindfulness intervention. Front Psychol. 2021;12:722995. https://doi.org/10.3389/fpsyg.2021.722995
Torous J, Nicholas J, Larsen ME, Firth J, Christensen H. Clinical review of user engagement with mental health smartphone apps: Evidence, theory and improvements. BMJ Ment Health. 2018;21(3):116–9. https://doi.org/10.1136/eb-2018-102891
Barkham M, Connell J, Stiles WB, Miles JN, Margison F, Evans C, Mellor-Clark J. Dose-effect relations and responsive regulation of treatment duration: the good enough level. J Consult Clin Psychol. 2006;74(1):160–7. https://doi.org/10.1037/0022-006X.74.1.160
Laird B, Puzia M, Larkey L, Ehlers D, Huberty J. A mobile app for stress management in middle-aged men and women (Calm): Feasibility randomized controlled trial. JMIR Form Res. 2022;6(5):e30294. https://doi.org/10.2196/30294
Mackenzie CS, Reynolds K, Cairney J, Streiner DL, Sareen J. Disorder-specific mental health service use for mood and anxiety disorders: Associations with age, sex, and psychiatric comorbidity. Depress Anxiety. 2012;29(3):234–42. https://doi.org/10.1002/da.20911
Al-Jebrni AH, Chwyl B, Wang XY, Wong A, Saab BJ. AI-enabled remote and objective quantification of stress at scale. Biomed Signal Process Control. 2020;59:101929. https://doi.org/10.1016/j.bspc.2020.101929

Due to a mix of technical and administrative errors we were unable to locate app account records for 6 individuals who attended the baseline session in the first year of the study. After revising our protocol, no further issues occurred. However, we ultimately decided to exclude these individuals since were unable to verify that they had created an account.
These engagement variables are existing metrics tracked by the developers of AmDTx and were made available in a partner report that was downloaded independently by the research team. The developers only track fully completed activities and meditations which may differ from other mindfulness apps. Participants could have started additional activities but exited prematurely and therefore, the metrics presented here may represent a more conservative estimate of engagement.
Participants were allowed to select multiple options for gender, but this was a rare occurrence (n = 4). For analysis, anyone who identified as a man and woman were coded as such, but anyone who selected a gender diverse option (e.g., non-binary, gender fluid, genderqueer) was coded into this third category for the purpose of understanding whether sex and/or gender impacted treatment outcomes.

Competing interest reported. BJS is the Chief Scientist and CEO of Mobio Interactive Inc., and he owned approximately 23% of the company at the time of this study. SS is a data scientist employed by Mobio Interactive Inc., with fewer that 0.1% in company stock options. BJS and SS served as technical liaisons for the study and did not contribute to study design, selection of primary outcome, data collection, or analysis. No other authors have connections to Mobio Interactive Inc. None of the authors of this study received financial compensation or any other form of compensation for the research undertaken herein. Mobio Interactive Inc. did, however, contribute in-kind funds (e.g., one-year licenses, staff, technical resources) to help fund the study as mandated by the funding agency.

MOBIOAcceptEngageOutcomeSupplementaryMaterialsV2.0.docx

Download PDF

Reviewers invited by journal
18 Nov, 2024
Editor invited by journal
13 Nov, 2024
Editor assigned by journal
23 Aug, 2024
Submission checks completed at journal
22 Aug, 2024
First submitted to journal
21 Aug, 2024

You are reading this latest preprint version

Acceptability, engagement, outcomes, and dose-response associations of a mindfulness-based meditation app in individuals waiting for psychological services

Status:

Version 1

Abstract

Figures

Introduction

The Present Study

Methods

Study Design and Recruitment

Procedure

Intervention

Measures

Demographic Information

Credibility and Acceptability Measures

Engagement and Usability Measures

Treatment Outcome Measures

Data Preparation

Statistical Analyses

Power

Results

Enrollment Sample

Credibility and Acceptability

Engagement

Changes in Outcomes Over Time

Dose-Response Associations between Engagement and Primary Outcome

Ancillary Analyses

Discussion

Limitations

Conclusions

Declarations

References

Footnotes

Additional Declarations

Supplementary Files

Status:

Version 1