The investigators aim to optimize an intervention which teaches women at risk for cardiovascular disease (CVD) to cope with this risk using self-compassion (an attitude of kindness towards oneself) in order to adopt a physically active lifestyle.

The investigators previously conducted a one-on-one self-compassion intervention for 11 women with CVD risk that successfully increased physical activity. However, meeting one-on-one with each participant was time and resource intensive.

The primary purpose of the present study is to determine if group self-compassion intervention delivery is comparably effective to one-on-one delivery for increasing physical activity among women at risk for CVD. The investigators hypothesize that group delivery should be at least as effective as individual delivery at increasing physical activity.

Furthermore, only women low in self-compassion were included in the previous study. The secondary purpose of the present study is to determine if women higher in self-compassion can also benefit from the intervention. The investigators hypothesize that participants at all levels of self-compassion should benefit similarly from the intervention.

Finally, the tertiary purpose is to determine if delivery mode and self-compassion interact such that one method of delivery (group or one-on-one) is better suited to women at low, moderate, or high self-compassion. This outcome is exploratory only and the researchers present no hypothesis.

The present study follows the methods of a randomized optimization trial. Women at risk for CVD will be randomly assigned to receive four sessions of combined physical activity behaviour change and self-compassion training in groups of six to nine or through one-on-one delivery. The intervention will provide one session to discuss participants' CVD risk and physical activity goals, and then three sessions of self-compassion training.

Outcome measures will be assessed pre- and post-intervention. The effectiveness of the one-on-one and group conditions for improving the study outcomes will be examined. The present research will determine how best to deliver an intervention which teaches women to cope with their CVD risk using self-compassion in order to become more physically active. The results of the present study will inform an eventual efficacy trial.

In preparation for the proposed research, the investigators conducted a one-arm pilot intervention (N=11) to examine the feasibility and acceptability of a self-compassion intervention for women at risk for cardiovascular disease (CVD). This pilot study established proof-of-concept (Czajkowski et al., 2015) for this intervention by producing a clinically significant increase in physical activity (M=837.5 Metabolic Equivalent (MET)*minutes per week; Florido et al., 2018). Further, feasibility criteria for recruitment time (1 week), retention rate (100%), and completion of study measures (97%) successfully exceeded the pre-established standards. Exit surveys found the intervention was 100% acceptable to participants. These results are encouraging yet questions remain about how best to deliver the intervention and with whom.

An optimization trial is the appropriate iterative step that will refine the present feasible and acceptable intervention with established proof of concept prior to conducting an efficacy trial. The present optimization trial will advance a research program informed by The ORBIT Model for Behavioural Treatment Development (Czajkowski et al., 2015). Two identified shortcomings of the pilot intervention will be targeted for optimization.

Primary Optimization Target: The pilot intervention was delivered individually to each participant. For 11 participants to complete the study, 33 individual meetings totalling over 40 hours were required over three weeks. This resource intensive delivery method may not be practical. Delivering the intervention to small groups would be more practical. Therefore, the primary objective of this optimization trial is to determine if a self-compassion intervention delivered in small groups is at least as effective at increasing physical activity among women at risk for CVD compared to a self-compassion intervention delivered individually. The investigators hypothesize that group delivery should be at least as effective as individual delivery at increasing physical activity.

Secondary Optimization Target: The previous pilot intervention only included women with CVD risk who were low on self-compassion. While women low on self-compassion are an obvious intervention target, women with moderate or high self-compassion may also benefit from learning to apply their self-compassion to their CVD risk and physical activity efforts. No research exists to inform the effectiveness of self-compassion interventions when delivered to participants low, moderate, and high on self-compassion. Therefore, the secondary objective of this optimization trial is to determine if baseline self-compassion influences physical activity, behavioural outcomes, or psychological outcomes of the intervention. The investigators hypothesize that participants at all levels of self-compassion should benefit similarly from the intervention. Finally, the tertiary objective is to determine if delivery mode and self-compassion interact such that one method of delivery (group or individual) is better suited to women at low, moderate, or high self-compassion. This outcome is exploratory only and the researchers put forth no hypothesis.

Trial Design: This intervention was developed with the Consolidated Standards of Reporting Trials (CONSORT) guidelines (Schulz et al., 2010). This is a two-armed optimization trial with 1:1 randomization to be conducted in a single centre in Manitoba, Canada. Women at risk for CVD will be randomized to receive a combined physical activity behaviour change plus self-compassion intervention in small groups (intervention condition) or individually with a facilitator (control condition). The individual condition serves as a positive control against which the effectiveness of the group condition at increasing physical activity will be judged. The intervention will be delivered to both conditions via an institution-sponsored videoconferencing platform. Data will be collected at pre- and post-intervention using validated surveys completed online. The investigators often use this data collection approach.

Participants: Participants will be women at risk for CVD identified through the Women's Advanced Risk-assessment in Manitoba (WARM) Hearts Cohort (NCT03938155), a large, ongoing prospective study to improve CVD risk detection in Manitoban women. Participants identified through pre-screening will be contacted by telephone and invited to participate in the study. Interested individuals will be sent an online questionnaire by email to assess the study eligibility criteria. Eligible individuals will be stratified by self-compassion score and sampled to equally enroll participants scoring low, moderate, and high on self-compassion. This recruitment strategy is required for the planned method of statistical analysis.

Randomization: Participants will be randomly allocated to the two intervention conditions with a 1:1 ratio. The random number list will not be accessible to research staff involved in recruitment and intervention delivery. The results of allocation will not be revealed until participants begin the intervention sessions in their respective conditions.

Intervention: Participants will engage in four, weekly one-hour sessions delivered by research assistants trained in self-compassion (six-week course). Both facilitators will deliver interventions to a similar number of participants in both conditions by alternating between conditions every four weeks. Intervention delivery will be guided by a script to prevent drift. To ensure facilitator fidelity, 20% of all intervention sessions will be recorded and shared with a self-compassion expert for feedback. This intervention draws from the Mindful Self-Compassion Program (Neff & Germer, 2013) and was developed by self-compassion experts.

Participants in the group condition will complete the four weekly sessions with six to nine other participants plus a facilitator; those in the individual condition will meet individually with a facilitator. Both conditions will be exposed to identical intervention content.

In Session One participants will discuss risk factors for CVD, be advised of the Canadian 24-Hour Movement Guidelines (Ross et al., 2020) and guided to set a physical activity goal using the SMART goals framework (American College of Sports Medicine, 2017).

In Session Two, participants will apply the three components of self-compassion (self-kindness, mindfulness, and common humanity) (Neff, 2003) to their CVD risk and physical activity efforts.

In Session Three, participants will learn to use the Yin (gentle; Neff, 2003) side of self-compassion to cope with difficult emotions, and the Yang (fierce, Neff, 2003) side of self-compassion to improve their health through physical activity.

In Session Four, participants will use the mindful practices of savouring and gratitude (Neff, 2003) to find enjoyment in physical activity.

Identical self-compassion writing exercises will be assigned to both conditions as home practice after Sessions 2-4.

Outcomes: Reliable and valid measures will be distributed using SurveyMonkey (www.momentive.ai) at pre-intervention and post-intervention.

Statistics: The sample required to achieve statistical power is 82 participants. The power analysis was conducted based on data showing that 10-minute increases in daily moderate to vigorous physical activity significantly lower CVD risk (Glazer et al., 2013). To protect against approximately 15% attrition (Ferrari et al., 2019) 96 participants will be enrolled in the trial.

Analyses will employ a 2 by 3 factorial ANOVA to address the three study objectives. Each factor in a factorial ANOVA is controlled for within the combinations of factors tested. The primary objective of the study is to determine if the group condition (intervention condition) is at least as effective as the individual condition (control condition) at increasing physical activity. The factorial ANOVA will test for main effects of the group condition on post-intervention physical activity across all levels of self-compassion, controlling for effects of the individual condition. The secondary and tertiary objectives of this study are to determine if baseline self-compassion influences how women respond to the intervention, and if delivery method and self-compassion interact such that one method of delivery is better suited to women higher or lower in self-compassion. The factorial ANOVA will explore main and interaction effects between baseline self-compassion (low, moderate, high) and intervention condition (group or individual) on physical activity, behavioural outcomes, and psychological outcomes.

Inclusion Criteria:

• Age 55 years or older
• Flagged response on CANHEART Index (Maclagan et al., 2014) (response of "yes" to having high blood pressure)
• Insufficient self-reported physical activity relative to guidelines of 150 minutes of moderate to vigorous physical activity per week.
• No current or planned enrolment in programming similar to the intervention
• No current diagnosis of cardiovascular disease
• No medical conditions that make it unsafe to engage in physical activity
• Available for all intervention sessions
• Comfortable using videoconferencing
• Comfortable reading and communicating in English

Score on the self-compassion scale will be monitored in order to ensure equal enrolment of participants scoring low, moderate, and high on the self-compassion scale (Neff, 2021). Eligibility criteria may be adjusted according to self-compassion score in order to represent the full range of possible self-compassion scores.

Exclusion Criteria:

• Being under 50 years old
• No flagged response on CANHEART Index (Maclagan et al., 2014) (response of "no" to having high blood pressure)
• Self-report meeting or exceeding guidelines of 150 minutes of moderate to vigorous physical activity per week.
• Current or planned enrolment in programming similar to the intervention
• Current diagnosis of cardiovascular disease
• Medical conditions that make it unsafe to engage in physical activity
• Not available for all intervention sessions
• Not comfortable using videoconferencing
• No comfortable reading and communicating in English

A third-party with no involvement in intervention delivery or data analysis will use a random numbers list generated by the IBM Statistical Package for the Social Sciences (SPSS) 28© (statistical software) to allocate participants to the two intervention conditions. The random number list will not be accessible to research staff involved in recruitment and intervention delivery. The results of allocation will not be revealed until participants begin the intervention sessions in their respective conditions.

Complete blinding of facilitators and participants is not possible; facilitators and participants will know if they are in the individual or group condition but will be blind to participants' pre-intervention self-compassion score.