Mind Your Heart: exploring feasibility, acceptability, and preliminary effects of phone-delivered mindfulness training on medication adherence in outpatients with chronic heart failure

Abstract

Background.

Mindfulness training (MT) may promote medication adherence in outpatients with heart failure (HF).

Objective.

To study the feasibility and acceptability of MT (primary outcomes) and explore effects on medication adherence; functional capacity; cognitive function; depression; and mindfulness skills (secondary outcomes).

Methods.

In this pre/post design study, participants received a 30-minute phone-delivered MT session weekly for 8 weeks.

Results.

We enrolled 33 outpatients (32% women; 69.7 white; mean age 60.3 years). Retention was 100% and session attendance 91%. Overall, participants (97%) rated MT as enjoyable. Objectively assessed (p<.05) adherence decreased post-intervention, while improvements were noted in functional capacity (p=.05), mindfulness (p<.05) and cognitive function (reaching significance for Flanker scores).

Conclusions.

Phone-delivered MT was feasible and acceptable. While no improvements were noted in medication adherence and depression, cognitive function, functional capacity, and mindfulness levels increased post-intervention, suggesting MT may have beneficial effects in HF outpatients.

Introduction

Heart failure (HF) is the most common diagnosis at hospital discharge in older adults and has an unfavorable prognosis, with 5-year mortality rates of 50%.¹ Pharmacological treatment improves symptoms and clinical outcomes but is effective only if patients take their medications regularly. Unfortunately, the large majority of patients with HF do not adhere to their prescribed regimens.² Poor medication adherence, in turn, significantly affects hospital re-admissions, emergency room visits, and mortality.³

Cognitive impairment affects about 60% of patients with HF and is an important predictor of poor medication adherence.^4,5 Domains that are most affected include executive function, attention, and memory.^4,5 In addition to cognitive impairment, depression is also highly prevalent among HF patients and contributes to the poor medication adherence observed in these patients.⁶

Mindfulness training (MT) is a behavioral intervention that involves learning to notice events arising in the field of consciousness at any given moment and to re-direct attention to a chosen focus of attention (‘attention training’). At the same time, the person learns to cultivate an attitude of acceptance (‘open awareness’), which allows for new ways to avoid absorption in habitual mental patterns, behaviors, and emotional reactions.⁷

A recent rigorous systematic review of experimental studies of MT in adults with HF (n=467) showed that MT interventions could significantly improve depression, anxiety and health-related quality of life after the intervention, while effects on physical symptoms and physical function were inconclusive.⁸ In addition, there is preliminary meta-analytical evidence that MT may improve cognitive function,⁹ which, as noted, is an important predictor of adherence and effective self-care in patients with HF.

Despite the possible benefits of MT in patients with HF, there are several barriers to its delivery to this population, as standard MT programs typically require in person attendance at eight 2½ hour long classes, an all-day intensive session, and intensive individual practice. While web- and app-based options are promising in selected sub-groups,¹⁰ phone-delivery has a particular appeal for older adults who typically are less tech-savvy. Because over 95% of Americans currently own a mobile phone,¹¹ phone delivery is an attractive option for these individuals.

Although potential effects of MT on medication adherence have been purported as a possible mechanism by which MT could improve cardiovascular health,¹² there are no published studies of the role of MT in improving medication adherence among patient with cardiovascular disease and data are scarce in other populations. Thus, in this study, we were interested in a) examining the feasibility and acceptability of phone-delivered MT in outpatients with chronic HF (primary outcomes) and b) exploring possible effects of phone-delivered MT on self-reported and objectively assessed medication adherence; functional capacity; cognitive function; depression; and mindfulness skills (secondary outcomes).

Study sample

Participants were recruited from the Providence, RI area through mass mailings; flyers placed in clinics, community centers, libraries, newspapers and supermarkets; and physician and peer referral. All participants provided written informed consent.

Inclusion criteria were age > = 18; a confirmed diagnosis of HF; self-reported suboptimal adherence to medications; access to a telephone; and ability to understand and speak English. The Voils questionnaire, a validated measure of adherence behaviors, was used to identify patients with suboptimal medication adherence.¹³ Participants marking any item response category that indicated suboptimal adherence were considered eligible. A diagnosis of HF was confirmed via review of the patient chart or, for participants enrolled from the community, by the participant’s physician.

Exclusion criteria were unwillingness/inability to provide informed consent; New York Heart Association (NYHA) class IV; severe hearing impairment not allowing phone delivery; severe depressive symptoms (Hospital Anxiety and Depression Scale scores > 14); acute psychosis; severe cognitive impairment (Blessed Orientation Memory and Concentration test scores ≥ 10); impossibility to confirm HF diagnosis; a recent (< 6 weeks) hospitalization; and ongoing mind-body practice (i.e., meditation, yoga or tai chi > once/month). To encourage retention, participants received a $50 stipend at each visit. The study was approved by the Institutional Review Board at our institution and is registered at the Clinicaltrials.gov repository (NCT03571581).

Sample size considerations

Based on a planned sample size of n=50, with two-sided α=.05, we had >80% power to detect a significant post-intervention change in secondary outcomes.

Study design

This was a single arm, open-label, pre-test/post-test study. Study staff and participants were not masked.

Intervention

Participants received one 30-minute phone-delivered MT session weekly for eight weeks. In addition, they were invited to practice at home every day using a 20-minute digitally recorded guided mindfulness practice. The intervention sessions were delivered by three certified instructors with 5+ years’ teaching experience. Each participant was trained by the same instructor throughout the duration of the intervention.

The MT protocol maintained the essential components of the standard Mindfulness-Based Stress reduction program created by Jon Kabat-Zinn at the University of Massachusetts,⁷ streamlined to meet the needs of individuals with chronic conditions. The training included attention-focusing practices (i.e., focusing the attention on one object of attention, e.g., the breath, and learning to re-direct attention to the chosen focus of attention when the attention slipped away) and open-awareness practice (learning to develop an attitude of openness and acceptance to whatever event arises at any given moment). We did not consider one of the many apps available on the market because the effects of app-delivered MT are yet to be demonstrated and user’s compliance is low (13% based on a recent study).¹⁴ A random sample (10%) of MT sessions were recorded and audited for consistency with the intervention protocol using a standard checklist developed in previous studies.¹⁵

Primary outcomes

Feasibility and acceptability.

Metrics of feasibility included retention rates, session attendance, and amount of individual mindfulness practice (recorded using a self-reported form). To measure acceptability, we used a one-item intervention enjoyment scale (1 = did not enjoy to 5 = enjoyed very much).

Secondary outcomes

Self-reported medication adherence was measured using the Voils questionnaire,¹³ a 3-item validated measure of adherence (internal consistency reliability = 0.86).¹³ The item response category consistent with perfect adherence (i.e., “none of the time” or “never”) are assigned a value of 1, and the item response category reflecting most non-adherence is assigned a value of 5 (i.e., “all of the time” or “always”). Higher scores (range: 1–5) indicate worse adherence behaviors.

As an objective measure of adherence, we used Electronic Drug Monitoring (EDM). Medication Event Monitoring Systems (MEMS - AARDEX USA, Inc., Union City, CA) are equipped with a digital microchip recording date and time of each cap opening and are widely used in research settings. We monitored one medication per patient using the algorithm developed by Wu et al.¹⁶ Briefly, if the patient took any HF medication twice a day, this medication was monitored. If all HF medications had the same daily prescription schedule, the β-blocker was monitored. If the patient did not receive a β-blocker, the ACE inhibitor or the angiotensin receptor blocker was used instead. If none of these was prescribed, the aldosterone antagonist, statin, anticoagulant, digoxin, diuretic or aspirin was monitored. Participants received a MEMS bottle at the baseline visit and were instructed to open it only when they actually took their medication. Unscheduled lid openings (i.e. refills, accidental openings) were tracked in a diary and removed prior to analysis. Participants started using the cap on the day following the baseline visit (=time 0), and the intervention started 30 days later; this 30-day period was considered the baseline. Adherence was operationalized as the proportion of medication taken/medication prescribed × 100, averaged over 30 days at each data collection point.

Functional capacity was measured with the 6-min walk test (6MWT), a simple, safe and reliable first-line assessment of aerobic capacity that correlates with peak oxygen uptake and predicted survival in pharmacological trials.¹⁷

Cognitive function was assessed using the Fluid Cognition battery from the NIH Toolbox and included: a) The Flanker Inhibitory Control and Attention Test (a measure of attention and inhibitory control). During the test, participants are instructed to focus on a stimulus while inhibiting attention to another stimulus (arrows). Sometimes the middle stimulus points in the same direction as the “flankers” (congruent) and sometimes in the opposite direction (incongruent) over 20 trials; b) The Oral Symbol Digit Test (a measure of working memory and processing speed). A key with nine abstract symbols is presented - each paired with a number between 1 and 9. Participants are asked to verbally indicate which numbers go with the symbols that are presented in a long string on the computer screen. The participant is given 120 seconds to call out as many numbers that go with the corresponding symbols as he/she can – in order, without skipping any; c) The List Sorting Working Memory Test (a measure of working memory). Pictures of foods and animals are displayed with accompanying audio recording and written text (e.g., “elephant”), and the participant is asked to say the items back in size order from smallest to largest, first within a single dimension (either animals or foods) and then on two dimensions (foods, then animals); d) The Picture Sequence Memory Test (a measure of episodic memory). Sequences of pictured objects and activities are presented in a particular order. The participants are asked to reproduce the sequence of pictures that is shown on the screen. Participants respond by dragging pictures from the yellow box on the screen into the gray boxes on the screen; e) The Dimensional Change Card Sort Test (a measure of cognitive flexibility and attention). Two target pictures are presented that vary along two dimensions (e.g., shape and color). Participants are asked to match a series of bivalent test pictures (e.g., yellow balls and blue trucks) to the target pictures, first according to one dimension (e.g., color) and then, after a number of trials, according to the other dimension (e.g., shape). The relevant dimension for sorting is indicated by a cue word (e.g., “shape” or “color”) that appears on the screen for all participants. Performance on all these measures correlates significantly with scores on “gold standard” neuropsychological tests of each of the cognitive domains of interest and have high test-retest reliability, internal consistency, and convergent and divergent validity.^18–20 Each test was independently completed by study participants (3–5 minutes each) using a study tablet after a baseline practice trial. For all measures, higher scores indicate better cognitive performance.

Depression was assessed with the Hospital Anxiety and Depression Scale (HADS), a self-administered measure commonly used in clinical settings, that has been validated in cardiac patients. Scores range between 0 and 21; a score of eight or more indicates borderline caseness for anxiety or depression, and a score greater than 14 indicates severe depressive symptoms.²¹

Mindfulness skills were assessed using the Multidimensional Assessment of Interoceptive Awareness (MAIA), a 32-item, 8-subscale self-report measure with higher scores indicating higher levels of mindful awareness. This instrument has adequate to excellent internal-consistency reliability, with Cronbach’s alphas between 0.66 and 0.87, and > 0.70 for five of the eight scales.²² Items are rated based on a 6-point Likert scale (0–5) with ‘0’ indicating ‘Never’ and ‘5’ indicating ‘Always’.

Descriptive variables

At baseline, we collected information on age, sex, ethnicity, education, and marital status via self-reported standardized surveys. Medical history was collected from medical records using standardized abstraction forms. Social support was assessed with the Multidimensional Scale of Perceived Social Support, a 12-item, unidimensional tool to measure how one perceives their social support system.²³ Body-mass index (BMI) [mass (kg) / height²(m)] was calculated from baseline weight and height measured using standard procedures.

Adverse events

An adverse event (AE) was defined as any untoward medical occurrence in a participant temporally associated with participation in the study. A Serious Adverse Event (SAE) was defined as any adverse event that results in death, a life-threatening event, and hospitalization or prolongation of an existing hospitalization.

Statistical analysis

The primary outcomes of this study were feasibility and acceptability. Retention and session attendance rates, amount of individual mindfulness practice, and program satisfaction scores were calculated post-intervention. The secondary outcomes were self-reported and objectively assessed medication adherence; functional capacity; cognitive function; depression; and mindfulness. To examine pre/post intervention changes in secondary outcomes we used a series of generalized linear models (GLMs), in which each secondary outcome measure was regressed on time, baseline value of the outcome and potential confounders. For medication adherence, these were selected a priori based on the literature and included sex and social support. For the remaining secondary outcomes, no baseline characteristic was associated with the outcomes of interest in univariate analyses and thus were not included in the analysis.

RESULTS

Recruitment started in October 2018 and was prematurely terminated in March 2020 due to the Covid-19 lockdown in Rhode Island. Considering our high-risk sample, follow-up assessments were since then conducted remotely. Data collection was completed in August 2020.

Out of 85 screened patients (Figure 1), a total of 33 subjects consented to participate in the study and provided baseline data. Of the 33 participants originally enrolled, 2 never started the intervention and were excluded from outcome analyses; thus, the final analytical sample was 31. All participants had a confirmed diagnosis of New York Heart Association functional class I-III HF. As shown in Supplementary Table 1, participants were on average 60.3 years old. Fifty-eight percent had at least some college education, 30% were women, and 70% were white. Approximately 45% had diabetes and 64% had hypertension; the average BMI was 35. No participant met HADS criteria for depression at baseline; 27% of participants received anti-depressants. The most commonly prescribed HF medications were beta-blockers (82%).

Figure 1.

Consort diagram

Note: Because all subjects were screened to completion, multiple ineligibility reasons are possible

Primary outcomes

Feasibility

The retention rate was 100% at 3 months. Participants attended, on average, 7.3 (SD =1.5) out of the 8 planned classes with an overall 91% attendance (classes attended/classes planned). Participants’ individual mindfulness practice averaged 80 minutes/week.

Acceptability

Overall, 97% of participants rated MT as enjoyable or very enjoyable, and 77% rated it as very enjoyable. Eight adverse events and 3 serious adverse events occurred during the study period (all were unrelated to the study intervention).

Secondary outcomes

Medication adherence

As shown in Table 1, we did not observe changes in self-reported medication adherence. EDM assessed medication adherence decreased from 73% at baseline to 58% post-intervention (p <.05).

Table 1.

Medication adherence and other secondary outcomes at baseline and post-intervention^a

	Baselineb n = 31	Post-intervention n = 31	P-value
Self-reported medication adherence (Voils)	2.8 (.7)	2.6 (.6)	.11
Objective medication adherence (EDM), %	72.7 (27.6)	57.9 (34.9)	.04
Functional capacity (6MWT, meters)	350 (93)	378 (97)	.05
Cognitive function scores (age-adjusted)
Dimensional Change Card Sort Test	101 (19.7)	100.7 (19.0)	.81
Picture Sequence Memory Test	98.2 (15.2)	102.1 (18.8)	.49
Flanker Inhibitory Control and Attention Test	87.9 (14.5)	93.2 (16)	.02
List Sorting Working Memory	98.2 (13.7)	101.5 (14.9)	.29
Oral Symbol Digit Test	58.8 (18.0)	65.7 (22).0	.33
HADS Depression	3.6 (3.3)	4.3 (3.2)	.13
MAIA
Noticing score	12.7 (4)	14.5 (4.1)	.03
Not distracting score	6.6 (3.2)	7.2 (3.3)	.33
Not worrying score	8.2 (2.1)	8.3 (2.4)	.77
Attention regulation score	20.1 (7.9)	24.2 (6.5)	.004
Emotional awareness score	15.9 (5.2)	18.3 (5.3)	.01
Self-regulation score	11.7 (5.0)	14.7 (3.6)	.002
Body listening score	7.4 (4.2)	8.9 (3.4)	.02
Trusting score	9.8 (3.9)	11.1 (2.8)	.05

^aValues are means (SD)

^bTwo subjects withdrew from the study before the MT intervention started and were not included in final outcome analyses. Thus, the analytical sample was 31.

EDM = Electronic drug monitoring

HADS = Hospital Anxiety and Depression Scale

PSQI = Pittsburgh Sleep Quality Index

MAIA = Multidimensional Assessment of Interoceptive Awareness

Functional capacity

Functional capacity (6MWT distance) increased from 350 (±92) at baseline to 378 (±97) meters at intervention completion (p=.05, Table 1).

Cognitive function

Baseline cognitive function was within normative ranges. With the exception of the Dimensional Change Card Sort Test, improvements were noted in all cognitive measures post-intervention, reaching significance for Flanker scores (Table 1).

Mindfulness and depressive symptoms

Mindfulness scores on most MAIA sub-scales significantly improved at the end of the intervention, while no changes were observed in depressive symptoms.

DISCUSSION

This proof-of-concept study confirms that phone-delivered MT is a feasible and acceptable option for patients with chronic conditions and is a highly scalable intervention for HF patients.

We did not observe improvements in self-reported adherence; objectively assessed adherence decreased following MT. It is possible that objectively assessed baseline adherence was inflated due to reactivity²⁴ and in fact, it was relatively high (73%) considering we recruited only participants who reported to be non-adherent. These findings are consistent with those of a pilot RCT of phone-delivered MT conducted in HIV+ patients,²⁵ which also found a decrease in anti-retroviral adherence from 85% at baseline to 74% post-intervention. Adherence typically decreases over time even in patients who are initially adherent,² and because this single-arm, pre/post design study did not have a control group, we do not know whether an even worse decrease in adherence would have been observed in the comparison group. Alternatively, it is possible that MT was not effective in promoting adherence, especially considering that the intervention did not focus on medication adherence. Further, medication adherence is a complex behavior which is influenced by different socioeconomic, health care system-, and patient-related factors which were not targeted by the MT intervention.²⁶

Functional capacity significantly improved at intervention completion. Although a 28-meter change may seem modest, the magnitude of the change in 6MWT distance in studies of pharmacological treatments for HF is around 50 meters.²⁷ A first explanation for this finding is encouragement from the study staff during the performance of the test, but this is unlikely given that the same staff member assisted study participants during the 6MWT at both study visits. Second, because of a lack of a control group, we cannot exclude that the observed change was due to unmeasured factors (and not to the intervention). We note, however, that 6MWT distance decreased or did not change at post-intervention in the control groups of several RCTs of behavioral interventions for patients with HF.²⁸ A possible mechanism underlying the improvement in functional capacity is that by inducing the relaxation response (a common effect of meditation practices) MT may counteract the effects of the prolonged sympathetic stimulation observed in chronic HF. Interestingly, similar improvements in 6MWT distance were reported in a small study using diaphragmatic breathing.²⁹

We found improvements in cognitive domains that are typically affected in patients with HF, namely attention, memory, and executive function. Although most changes were non-significant, changes of this magnitude are clinically meaningful. Further, due to the Covid-19 pandemic, we were not able to meet recruitment goals (n=50), and as a consequence, the study was underpowered to detect significant changes in secondary outcomes. Also, it should be noted that the focus of this proof of concept study was to obtain indications of the direction and magnitude of MT effect on the proposed measures, rather than establish efficacy. Interestingly, our findings are consistent with those reported in a recent literature review suggesting that mind-body interventions (including MT) may have positive effects on cognitive function in older adults with mild cognitive impairment.³⁰ Given the high prevalence of cognitive impairment among patient with HF,^4,5 the observed improvements in cognitive function are a notable finding of this study and could have an important clinical impact if confirmed in a large, fully powered randomized controlled study (RCT). While it is yet to be determined whether gains in cognitive abilities following MT training are associated with increased neuroplasticity in older populations, a recent small RCT conducted in adults with mild cognitive impairment has shown, along with improvements in memory and attention, changes in gray matter cortical thickness following a 9-month mindfulness-based training.³¹

Finally, we found significant post-intervention changes in mindfulness skills. This is an important finding because the literature has shown that changes in mindfulness mediate the effects of mindfulness-based interventions on psychological outcomes.³² Notably, our intervention improved mindfulness using a much less intense protocol compared to that used in other studies (typically, for in-person programs, 8 sessions lasting 2.5 hours and all-day intensive session, plus 45 minutes of individual practice per day).⁷ It is possible that a less intense training may result in better compliance with the MT intervention and ultimately yield greater effects. For example, in our study, participant retention was 100% vs. 70% for standard MT programs.³³

This study has several notable strengths, including a good representation of minorities and women, the use of rigorous assessments of cognitive function, and the employment of both self-reported and objective assessments of adherence. This study also has important limitations. Because of the lack of a control group we cannot exclude that the observed changes are due to bias (e.g., regression to the mean; unmeasured confounding). Second, because our sample had high education and normal depression scores, findings are not generalizable to less educated patients and to those with depression. Finally, we were unable to reach our planned recruitment goal due to the onset of the Covid-19 pandemic, and as a consequence, the study was underpowered.

CONCLUSIONS

In sum, phone-delivered MT is a feasible and acceptable option for patients with HF. While we did not observe effects of MT on medication adherence, improvements were noted in most measures of cognitive function, functional capacity, and mindfulness levels. These findings suggest MT may have a beneficial effect in these patients and if confirmed in a fully powered RCT, they could have important clinical implications.

What’s New?

• In this proof-of-concept study, we found phone-delivered MT was a feasible and acceptable option for patients with HF.

• While no changes were observed in medication adherence, MT improved cognitive function, mindfulness, and functional capacity compared to baseline.

• These preliminary findings, if confirmed in a fully powered RCT, could have a beneficial effect in patients with HF.