Physical activity in pregnancy: a mixed methods process evaluation of the FitMum randomised controlled trial interventions | BMC Public Health

Characteristics of participants

Two hundred and twenty healthy, inactive pregnant women were included in the FitMum trial and 219 with a gestational age of 12.9 (9.4–13.9) (median (IQR)) weeks were randomised (Table 2).

Of the randomised participants 80% had an educational level ≥ bachelor’s degree. A total of 20 interviews were conducted; 10 interviews of participants randomised to EXE or MOT, respectively (Table 2). Maternal baseline characteristics of the subset of 20 interviewees and the 219 randomised participants in the FitMum trial did not seem to differ.

Reach

Of the included participants, 58% (n = 128) reported, that they were introduced to the FitMum trial while booking their first-trimester ultrasonic scan, 20% (n = 45) at the outpatient clinic at Copenhagen University Hospital – North Zealand, 15% (n = 32) via posters at e.g. their general practitioner, 10% (n = 23) via social media, 8% (n = 18) via friends or family, 5% (n = 12) via an online Danish pregnancy platform [31], and 9% (n = 19) via other options. Before randomisation, 91% (n = 201) stated that they wanted to participate in the trial to increase their level of PA, 56% (n = 123) to take part in and contribute to research, 7% (n = 16) to have a closer contact with health professionals, 5% (n = 10) to interact with other pregnant women, and 8% (n = 18) had other reasons. Participants in both intervention groups expressed in the interviews that the desire to become more physically active was mostly for the woman’s own good and arose from various factors; in general, there was an underlying understanding that the body naturally weakens during pregnancy. Hence, a physically active pregnancy was equated to an uncomplicated pregnancy with e.g., less pain and decreased risk of pregnancy complications. In extension, the participants reasoned that an uncomplicated pregnancy would lead to an uncomplicated delivery and emphasised that being in a good physical condition was a prerequisite for an uncomplicated delivery. The women assumed that their PA level would be low and mainly reserved to general everyday activities if not being a part of the interventions. One woman linked a hypothetically low PA level with self-blame and expressed that:

“(If not being a part of the intervention) I could fear that I was still on the couch at home. That I hadn’t gotten my act together. And then I think I would have felt guilty if I then had an awful delivery. I could blame myself a bit for that, actually” (Participant no. 117, EXE).

It appeared that the desire to become more physically active unconsciously resulted in a feeling of responsibility not only for the woman herself, but also in terms of the birth outcome and the well-being of the child. In addition, excessive gestational weight gain was framed as a concern. Some women stated that they had gained more weight than wanted in their previous pregnancies and by being physically active they wanted to limit their weight gain in their present pregnancy. One woman explained that she, because of being overweight, felt a greater responsibility to be physically active during the pregnancy. She expressed a concern about being judged by others if she did not try to improve the health of her unborn child through PA.

Fidelity

The original planned sessions were held for 17.5 months with 120 participants included (CON: n = 24, EXE: n = 48, MOT: n = 48). In this period participants in EXE and MOT received physical interventions only. On March 11th, 2020 COVID-19 restrictions were implemented in Denmark. Thus, the original setup of EXE and MOT with physical attendance was altered into an online design of both interventions with participants attending from home [17]. Overall, the online interventions applied the same conditions as the physical interventions. However, in the online setup, EXE sessions were held virtually with 30 min of individual, offline, and self-selected aerobic PA followed by 30 min online structured aerobic and resistance training in groups (except for three months with pertained authority that allowed swimming pool access). In MOT, the content and distribution of group and individual sessions remained the same, however held online.

The online sessions ran for 14.5 months with 63 participants (CON: n = 14, EXE: n = 25, MOT: n = 24). Participants in EXE and MOT received the online interventions only as they were included and gave birth during the pandemic. Thirty-six participants (CON: n = 7, EXE: n = 14, MOT: n = 15) were included before the COVID-19 restrictions but gave birth during the pandemic. Participants in EXE and MOT received both the physical and online intervention. There were no differences in the lost to follow-up rate between participants who were included before or during COVID-19 restrictions.

Dose delivered

EXE sessions were delivered six days a week and the participants were recommended to choose three of the sessions (Figs. 1 and 2). During the trial period of approximately 32 months, one EXE session was cancelled due to sickness among the intervention providers. Only during few holiday periods were EXE sessions offered less than six days a week and some sessions were rescheduled. No MOT sessions were cancelled by the intervention providers. A few MOT sessions were scheduled out of range due to holidays or sickness. However, providers strived to reschedule the sessions as close to the allocated period as possible (Figs. 1 and 2). During the process of re-designing the interventions into the online setup due to COVID-19 restrictions, six consecutive EXE sessions were cancelled.

Participants in both intervention groups expressed in the interviews that the intervention accessibility was high. All participants in EXE expressed that the accessibility of the sessions was important to fit the exercise training into their daily lives. Some expressed that the scheduled sessions resulted in a regular exercise routine in which they preferred to attend sessions on the same weekdays. Some participants even scheduled the sessions into their work calendar to indicate to colleagues that they were occupied. For others, the timing of the EXE sessions was a barrier to participation, as it was difficult to fit in to their everyday life and commitments. They were dependent on the frequently offered EXE sessions to devise a more flexible schedule. A woman in EXE mentioned that:

“It (attending exercise sessions) has been a bit difficult to juggle, but being employed as I am, I have quite flexible working hours, and as the sessions were offered on so many different days, I could sort of choose the days when I didn’t have to show physically for work” (Participant no. 73, EXE).

Dose received

Throughout the trial period, participants randomised to EXE attended on average 1.3 [95% confidence interval, 1.1; 1.5] sessions/week of the recommended 3 sessions/week from randomisation to birth. The attendance rate in the online setup of the EXE intervention was 45% higher compared to the attendance rate in the physical setup (online: 1.6 [1.3; 2.0] sessions/week; physical: 1.1 [0.9; 1.4] sessions/week, p = 0.027) [18].

During the trial period 28% (n = 24) of the 87 participants in EXE participated on average in 2 or more sessions/week, 32% (n = 28) participated on average in 1-1.9 sessions/week, and 40% (n = 35) participated on average in less than 1 session/week. Among the 48 participants who received the physical EXE intervention only, 19% (n = 9) attended 2 or more sessions/week, 35% (n = 7) attended 1-1.9 session/week, and 46% (n = 22) attended less than 1 session/week. Among the 25 women who received the online EXE intervention only, 52% (n = 13) attended 2 or more sessions/week, 24% (n = 6) attended 1-1.9 sessions/week and 24% (n = 6) attended less than 1 session/week. The attendance rate in EXE in relation to gestational age is presented in Fig. 2.

Dose received among participants in EXE who were still included and not lost to follow-up did not differ from dose received among all participants randomised to EXE. Throughout the trial period, morning and afternoon sessions seemed to be equally attractive, whereas the Saturday session (a morning gym session) was the most attended during the week (Table 3).

Throughout the trial period, participants randomised to MOT attended 5.2 [4.7; 5.7] out of 7 counselling sessions (74%) during their pregnancies. The number of MOT sessions attended did not differ between participants offered physical or online sessions (physical: 5.3 [4.6; 6.0]; online: 5.6 [4.8; 6.4], p = 0.970) [18]. 64% of the 87 participants in MOT (n = 56) attended six or seven sessions, 13% (n = 11) attended four or five sessions and 23% (n = 20) attended up to three sessions. More than 80% of participants randomised to MOT attended the first group and the first individual session whereas 57% attended the last group session (Table 4).

The average percentage of attendance in group and individual sessions were 67% and 79%, respectively. Among participants not lost to follow-up in MOT, more than 80% attended the first group and all individual sessions and approximately 70% attended group session 2 and 3. The average percentage of attendance in group and individual sessions was 76% and 89%, respectively.

In both intervention groups, participants expressed in the interviews that being part of a group was valued, but that it was seen only as a fun and enjoyable factor and not to network or build new relationships. To some degree, participants in EXE expressed that it was difficult to participate in the sessions due to work, logistics, and family commitments in their everyday life, which they to a larger extent than usual needed to organise. They experienced that in relation to some family activities they were less present than they used to be and wanted to be. In addition, they were more dependent than usual on their partner, for example, to pick up and drop off their children at day care or school etc. and accompany them to leisure activities due to scheduled EXE sessions. A participant in EXE, aged 30 years and with a three-year old child, described how she and her husband organised everyday activities:

“Well, we need to do some planning. For example, I often attend [training sessions] on Wednesday afternoons, and my daughter has also started gymnastics – so they (husband and child) also come home late, and we will eat leftovers that day” (Participant no. 117, EXE).

Furthermore, it was difficult for participants in EXE to take part in family routines such as evening meals or preparation of these on days with an afternoon EXE session. For some participants, this led to a sense of guilt for not being present in family matters. However, participating in EXE sessions was perceived as a good opportunity to focus on oneself and, despite spending less time with the family, the women experienced increased energy to take care of older children and everyday chores at other times. On a purely practical level, participants in EXE expressed that they needed a car to be able to reach the gym or swimming pool. A 31-year-old woman explained how attendance was hindered:

“I didn’t really think transportation would matter, but it did, because we only have one car … I had to drop off my child beforehand, it just didn’t add up. I actually invested in a travel card for the train, but it was so much easier when the car was available” (Participant no. 87, EXE).

Commuting back and forth to the EXE sessions was by some of the participants in EXE not living near the training facilities, perceived as time heavy and as a barrier towards participation. In contrast, commuting was expressed as one of the most significant changes in the everyday life among a large part of participants in MOT. Instead of driving between their workplace and home as they normally would, they incorporated physical active commuting like biking. In addition, the participants in MOT incorporated more PA into already existing activities and added new activities that also involved family members. Participants in MOT expressed that it was important for them not to let their PA level limit their presence in family matters. A woman who was unemployed tried to schedule her exercise routines by separating them from family time:

“I wanted to be physically active while my boyfriend was at work and my daughter was at day care, so in that way I don’t think it (her being physically active) had any impact on our daily lives” (Participant no. 71, MOT).

In contrast, a woman with two older children combined family time with her being physically active:

“My children do gymnastics twice a week, and instead of them biking alone, I bike with them. They find it very nice. Additionally, my husband and I have had a few more evening walks together just the two of us while the kids were at home. It was really nice because I’ve also needed to “achieve” some more steps (on the tracker). My husband just said: “Okay, then I’ll come with you”” (Participant no. 109, MOT).

Notably, it seemed like participation in PA in MOT was perceived as easier to fit into everyday life, and that it caused less conflicts in planning everyday life than what was perceived among participants in EXE.

Mechanisms of impact

In general, participants in both intervention groups expressed in the interviews that they valued the interventions and appreciated being part of a research trial. Some of the participants expressed that they were able to plan their own working hours, which allowed them to participate in the interventions. A mechanism of impact was that the scheduled EXE sessions represented a commitment that participants in EXE felt responsible for keeping. It resulted in participants not having to continually “renegotiate”, either with their families or with themselves, to prioritise time for PA in their daily lives. Participants in EXE expressed that having intervention providers and other EXE participants waiting for them influenced highly on their commitment to the intervention and was a motivator for being physically active:

“I’m a very dutiful person, so when something is in my calendar and I’ve said it’s a deal, well, it’s a deal. I’m not so dutiful when it comes to my own obligations to myself. But when I say I’m going to show up, I show up.“ (Participant no. 73, EXE).

In contrast, participants in MOT expressed that they felt self-determined towards PA and how to structure and organise PA on their own while supervised and supported by the intervention providers. A perception of empowerment was one of the most motivating and important mechanisms of impact for participation in MOT and for their PA level and intervention maintenance. As participants in EXE, they expressed a great ability to independently structure their everyday life, which was essential for participation:

“I have a job where I have a lot of flexibility, so when I had to go in for a counselling session, I’ve just taken time off and worked at another time” (Participant no. 124, MOT).

Another mechanism of impact was the perceptions of PA which were notably different between the two groups. Participants in EXE considered PA to be an event that took place at a specific time point. Once they had participated in an EXE session, PA was not considered integrated in the remaining day:

“Well, I think (when attending an exercise session), I can tick that one off. Then I have kind of been active today. It was like one of those things that I had on my agenda” (Participant no. 103, EXE).

It appeared that participants in EXE separated everyday activities from what they perceived as actual exercise and distinguished between PA intensities. They found the sessions to be fruitful and valuable, but at the same time they noted the low degree of autonomy regarding the specific content of the sessions. For example, some of the participants in EXE found the 30 min session on the stationary bike (the first part of the 1-hour session in the original setup) to be monotonous and bland. However, their motivation was that stationary biking was the best activity to increase the heart rate to the required level when they felt heavier which made them continue. As oppose to the understanding of PA as an event in the EXE group, MOT participants seemed to have integrated PA more in daily activities. Participants in MOT expressed that PA of all kinds was considered valid, regardless of intensity. A woman expressed it like this:

“Exercise doesn’t have to be me going to the gym three times a week or me going for that run like everybody else does. This thing about exercise, it can be many things. It can also be that I just take the stairs 10 times or that I just walk faster with the pram now that I’m out walking anyway. So, I think it (PA) just became more simplified. That it doesn’t have to be so difficult” (Participant no. 133, MOT).

The different perceptions of PA were also expressed in the way that the participants referred to the mental and physical reactions and changes they experienced. Participants in EXE focused particularly on bodily capacities, changes, and appearance:

“You can tell from my body that I’ve been training hard … I can see it in my posture and just things like thighs and glutes and arms and stuff. They were maybe just a bit more untrained [before]” (Participant no. 81, EXE).

and

“I think it has been amazing to feel that I have become stronger” (Participant no. 86, EXE).

Participants in MOT expressed a somewhat broader perception of PA effects as they found themselves with greater insight and understanding of themselves being pregnant and with increased mental health and well-being:

“I’ve really felt good about my body in this pregnancy, and I think that’s so great. I think it’s largely because I’ve gotten to know my body and I’m in such good shape” (Participant no. 124, MOT)

and

“I think it (being a part of the intervention) had an impact on my well-being in general, including my mental well-being. Because I can feel my mood gets better, when I exercise (Participant no. 74, MOT).