Due to funding restraints, we aimed to enroll 24 SS-IC dyads into the intervention group and 12 dyads into the control group. Individuals were eligible if a dyad was not possible. Potential participants were approached during an outpatient stroke clinic visit. SS are seen in our stroke clinic approximately 2–6 weeks post-hospitalization, and visits are repeated at 3, 6, and 12 months. SS were included if they were at least 18 years old, could speak and read English, could provide written informed consent, had an ischemic stroke, intracerebral hemorrhagic stroke, or transient ischemic attack (recurrent or first-time stroke event) within the past 12 months, and currently lived at home. SS motivation and readiness to participate in recovery has been shown to be present outside of the immediate acute stroke period . IC were included if they were at least 18 years old, could speak and read English, could provide written informed consent and self-identified as an IC of a SS. To reflect standard practice, SS and IC were eligible to participate regardless of whether they were currently receiving anti-depressant or anxiolytic medications. SS residing outside of the home (e.g., extended rehabilitation facility) were excluded to assist us in limiting the enrollment of survivors with major cognitive and functional disabilities and impairments that may limit their ability to participate. SS and IC were excluded for psychosis, illegal substance use, suicidal ideation, or self-report current engagement in psychotherapy or meditation practices of any kind. A brief Montreal Cognitive Assessment (MoCA)  score of < 9 was used to determine the absence of severe cognitive impairment in assessing SS capability of providing consent. However, it is important to recognize that the MoCA may not fully capture one’s ability to understand key informed consent elements and provide informed consent. If a SS had a brief MoCA score < 9, but appears to understand key informed consent elements (e.g., potential risks) demonstrating adequate consent capacity, written informed consent was obtained. Written informed consent was obtained from all participants prior to enrollment. Approval prior to study onset was obtained from the university’s Institutional Review Board, and all procedures were followed in accordance with guidelines. All research was conducted at a medical center in Southeastern Texas.
The estimate of effect size was based on a study that examined the feasibility and efficacy of meditation to improve depressive symptoms as a secondary outcome measure . Based on that study’s effect size (d = 0.65), a repeated measures analysis for change in instrument score only within the intervention group would have 80% power when the sample size is n = 21. Considering an attrition rate of 12%, the targeted enrollment was 36 dyads. We estimated our attrition rate based on our experiences successfully completing other low-risk clinical trials. Power calculations were performed with G*Power 3.1.
Using a two-group blocked randomization scheme with a 2:1 allocation, dyads or individual participants were assigned to the intervention or control groups in blocks of six. The statistician generated the random allocation sequence in SAS software and was not involved in enrollment or intervention assignment. The randomized list was uploaded to Research Electronic Data Capture (REDCap)  where it was used to provide the group assignment.
The breathing-based meditation intervention occurred in 4-week blocks and comprised four group sessions, one session per week, conducted by a senior meditation teacher with 20 years of experience in mind–body interventions (A.C.) . The meditation group sessions were held at a nursing school in the same medical center as the recruitment site. Each session lasted approximately 1 h and built upon the previous session(s) to allow for repetitive practice. Participants were given pre-recorded audio and written instructions of the first three sessions at the end of each corresponding session and encouraged to practice on their own on days when they did not meet with the meditation teacher and daily thereafter. Participants were asked to document home practices using a daily meditation practice questionnaire. Daily meditation practice adherence was assessed as the proportion of daily meditation practice questionnaires that indicated the participant had meditated.
Session one focused on learning breathing techniques to calm one’s mind. Session two added learning how to still one’s body, silence one’s speech, and open one’s mind. Session three added learning to stay deeper in one’s connection through mindfully maintaining and deepening the connection through breathing with more openness and awareness and connecting to one’s qualities of warmth. Session four included expanding the warmth of one’s heart with loving kindness for others without forgetting oneself and an overview of the previous sessions, practice of the entire program, and an opportunity to ask questions. The meditation intervention was developed from the Tibetan tradition of the three doors which includes the body, speech, and mind . A fidelity checklist was employed to evaluate the provision of the sessions by the specialist. All meditation intervention sessions were audio recorded, and every tenth session was reviewed by a member of the research study not involved in the intervention sessions for intervention fidelity by the senior meditation teacher.
Participants assigned to the control group received four educational handouts on an expressive writing technique via email or mail at times corresponding to each of the four sessions in the intervention group. Handout one highlighted the benefits of writing about life experiences. Handout two focused on exploring emotions and thoughts about traumatic experiences through writing. Handout three outlined ideas for developing a writing practice. Participants were asked to write about anything of personal relevance or importance to them. Handout four focused on writing about experiences and expressing emotions through writing. Additional prompts to engage in writing techniques were not provided. A control group allowed for a realistic examination of enrollment and retention rates and randomization procedures.
Participants were formally assessed at four time-points: pre-intervention or baseline and post-intervention (0 [immediately], 4, and 8 weeks after intervention completion) using REDCap electronic questionnaires [16, 19]. Baseline demographic data were obtained. The Center for Epidemiologic Studies Depression (CES-D) scale was used to assess depressive symptom severity with a cutoff score of 16 or higher to identify risk for clinical depression and has been validated in community-dwelling adults with stroke [20, 21]. Diagnostic Interview for Genetic Studies (DIGS)  questionnaires were administered via telephone by the PI or a trained research staff member at baseline. The DIGS  questionnaires were reviewed by a psychiatrist (A.P.) and the PI (JB), using algorithms to determine the presence or history of psychiatric disorders and conditions. Anxiety was measured using the State-Trait Anxiety Inventory for Adults (STAI) . Range of scores for each State and Trait subscale is 20–80, with higher scores indicating greater levels of anxiety. A cutoff score of 40 on the State-Anxiety subscale is routinely used to indicate clinical levels of anxiety . The Brief Pain Inventory-Short Form (BPI-sf) was used to assess pain intensity on a 0 to 10 scale with 0 indicating “no pain” and 10 indicating “pain as bad as you can imagine” and interference on a 0 to 10 scale with 0 indicating “does not interfere” and 10 indicating “completely interferes” in the past 24 h . While there are no published reports of the sensitivity and specificity of the STAI or BPI-sf in screening for post-stroke anxiety or pain, our experience is that these instruments are commonly used with the post-stroke community [25, 26]. Participants assigned to the intervention group completed daily meditation practice questionnaires during the 4-week intervention and 8-week post-intervention phases. The questionnaire included five questions evaluating the daily total time in minutes and frequency of the day’s practice as number of times per day, whether the practice was alone as a yes or no response, and how the participant felt (e.g., happy, anxious, sad, or depressed) before and after the practice. An open-ended question was posed to obtain any additional thoughts, notes, or comments about the practice.
In order to understand the feasibility of obtaining and storing blood samples over time, samples were collected at the four data collection time-points. Samples were centrifuged to obtain 500 µl plasma aliquots and stored at − 80 °C for future analysis of biomarkers (e.g., pro-inflammatory cytokines). Research of the effects of mind–body interventions suggests multiple biological pathways (e.g., inflammatory and neuroendocrine) that may benefit SS .
Adherence to the intervention protocol
Intervention fidelity strategies were employed, including a fidelity checklist. All in-person meditation intervention sessions were recorded, and every tenth session was reviewed for intervention fidelity. Daily participant meditation practice adherence was assessed as the proportion of daily meditation practice questionnaires that indicated the participant had meditated.
Acceptability of the intervention and trial design
Immediately after each final (fourth) meditation session, the PI and senior meditation teacher informally interviewed participants assigned to the intervention group to assess the acceptability of the intervention and trial design. These unstructured group interviews were audio recorded and manually transcribed by a member of the research team. Reasons for non-participation were obtained from eligible individuals who declined to participate [16, 19]. Participant field notes were documented in REDCap [16, 19] by the PI and research staff.
Participant scores for the instruments were calculated as the mean of the answered items multiplied by the total number of items in the scale (or just the mean of the answered items for the BPI-sf). This method addresses missing data by using the scale items that were answered by a participant to calculate their scale score . Descriptive statistics in the form of frequencies and percentages were used to examine demographic and clinical characteristics. Based on our experiences completing other low-risk, longitudinal RCTs with community-dwelling SS, pre-determined criteria for determining the success of the feasibility study were based on 80% or more of eligible individuals enrolling. Additionally, 80% or more needed to complete at least one questionnaire at baseline and one questionnaire at 8 weeks post-intervention. Participants needed to complete 75% or more of the questionnaires during the visits. Intervention participants needed to attend 75% or more of meditation sessions.
Quantitative data analysis was based on intention-to-treat principle. The analysis tested for changes in CES-D, STAI, and BPI-sf scores for the intervention group only. These within-group comparisons between time points were conducted with paired t tests. Although the study had adequate statistical power to analyze change only within the intervention group, additional analyses were conducted using repeated measures analysis with linear mixed models to compare the intervention and control groups for changes over time in the CES-D, STAI and BPI-sf questionnaires. Separate models were conducted for the SS and IC participants as individuals and not as dyads. Qualitative data analyses included two independent members of the research team reviewing the transcriptions of the unstructured group interviews and the field notes and grouping the information into categories. As a group, a third member of the team worked with the independent members to review and determine the final categories. Analysis included reviewing the reasons for declining study participation and grouping the reasons into categories.