LIVING - Physical Activity in a Self-management Program Among Persons With Type 2 Diabetes

Trial design The LIVING project is designed as a parallel-group, 2-arm pragmatic RCT. The trial sites are to deliver either sub study 1: Lev Livet versus a five month waiting list or sub study 2: Lev Livet + HIIT versus a five month waiting list. At the clarifying interview with a health professional in the municipality, the participants will be randomly allocated to either Lev Livet (+HIIT) or a waiting list (control group). Within each sub study, the participants will be randomly allocated (2:1) to intervention or a control group.

Participants All people living with T2D being referred from their general practitioner or a hospital to the municipality can be included in the project.

Intervention Lev Livet is a group-based program where the aim for the participant is to acquire skills to master their daily life with T2D. The program consists of 9 sessions with a duration of two and a half to three hours for each session. In the first four sessions coping skills are taught across different chronic diseases. In this part of the program citizens suffering from different chronic diseases can participate. Participants living with T2D then move on to four diabetes-specific sessions, followed by a ninth follow up session. The eight first sessions are held consecutively over eight weeks. The ninth session is held three weeks after the eighth session.

During the 8 weeks, a supervised HIIT-protocol comprising blocks of 5 x 1 min high-intensity intervals interrupted by 1 min active recovery will be added. Between each blocks, a 4 min brake will be planned where the participant can shift from an upper body exercise to a lower body exercise or vice versa. The HIIT intervention will progress continuously, from two weekly exercise sessions with two exercise blocks (22 minutes of exercise) in the first two weeks, to two weekly exercise sessions with four exercise blocks in the last four weeks of the intervention. The exercise intensity will be monitored using the Borg 15 scale. The intensity of exercise will be set to Borg 14-16 corresponding to 77-93% of maximal heart rate. In addition to the two exercise therapy sessions, the participants will be instructed to independently complete an exercise session equivalent to the exercise therapy session for that week.

Statistical analyses The primary analyses of both the primary and secondary outcomes will be based on the mean difference between individual changes from baseline to 3 months when comparing the Lev Livet intervention and control group.

Analysis will be done according to the intention-to-treat principle, meaning that all participants allocated to an active treatment group will be analysed as a member of that group regardless of their compliance with the intervention. Missing data will be imputed in the primary analysis using a multiple imputation procedure29.

The analysis will be performed using a mixed model with repeated measures. Specifically, time as a categorical covariate, randomisation status, and their interaction will be included. This will allow assessment of difference at each time point of follow-up between groups, as well as of the overall difference. Dependence of repeated observations for participants will be handled with a random effect for participant. In addition, random effects for group and municipality will be included, assuming independence between random effects at different levels (participant, group, municipality). To explore the effect of adding HIIT to the intervention, the investigators will use a mixed model as described above, but with an added interaction term between sub study 2, sub study 1 and follow-up status. Further, the effect of adding HIIT to the intervention will be explored on other outcomes conducting analysis of mediation.

Analyses of baseline equivalence between groups of intervention and control group will be performed. In order to increase precision of the outcome estimate, covariates can be included in the analyses if relevant.

Power calculation and sample size In a realistic real-world set-up, each municipality will be able to run two rounds of intervention, e.g. randomize participants into two intervention groups and two control groups. The intervention to control group ratio is estimated to be 2:1. Number of participants in each intervention group is expected to be 12 and 6 in each control group, meaning that each municipality has to randomise 36 participants. A simulation procedure was applied to estimate the power assuming a difference in change between groups of 5 and that individual changes had an SD (standard deviation) of 4. Including four municipalities in each sub study, 8 in total (288 participants in total) would yield a power of 80%.

Qualitative study on the health care professionals' perspective To gain knowledge on the implementation process of the intervention Lev Livet, and Lev Livet + HIIT, the perspectives of the providing health care professionals will be investigated13.

During the delivery of the intervention, the experiences of twenty to thirty healthcare professionals from various sites will be examined. Data collection will include participant observations30 and focus group interviews31. The following research questions will guide data collection:

• What is at stake in clinical practice when delivering the intervention?
• How do healthcare professionals perceive their relation to the patients during the intervention?
• What characterizes the health care support given to patients during the intervention? Thematic analysis32 will lead to new insight into the healthcare professional perspective during the intervention, relevant for further refinements and implementation.

Qualitative approach to explore participants' experiences Inspired by realistic evaluation33, the aim is to explore what parts of the disease management program work for whom and why. This includes an exploration of the barriers of implementing new knowledge and habits from Lev Livet into the participants' daily lives as well as the elements they have successfully implemented.

After completing Lev Livet, participants from sub study 2 in two different municipalities (n=24), will be invited for an interview. Using a semi-structured, open-ended interview methodology34, the interviews will touch upon topics such as; their daily life with diabetes, expectations before participating in Lev Livet, their experienced outcomes, motivation to participate and carry on with new habits, and being part of a group with a disease in common.

Participants in the first round of interviews will be invited for follow-up interviews four to six months later. These will focus on how the participants have managed their diabetes since Lev Livet, which habits they have or have not implemented into their daily lives and which factors play into what works in their daily setting.

Health economic evaluation As there are expectedly multiple effects of varying character, a cost consequence analysis will be performed to analyse the health economic consequences of the intervention. The costs of the Lev Livet intervention as well as the HIIT component is estimated and these costs are related to the primary and secondary outcomes collectively. When estimating the costs, a rigorous approach will be applied. This includes identifying all relevant cost components in collaboration with project participants, measuring the resources consumed using the most appropriate method for each cost component, and finally calculating the monetary value by using individual unit prices35.