Supporting healthful lifestyles during pregnancy: a health coach intervention pilot study

We recruited between July 2015 and January 2016 at one northeastern US clinic. We included participants if they were less than 16 weeks’ gestation at the time of their initial obstetric appointment, overweight or had obesity (pre-pregnancy BMI ≥ 25 kg/m²), 18 years of age or older, English speaking, and planned to remain at the same obstetric clinic for the duration of their pregnancies. The clinic provided us with monthly lists of appointments for potentially eligible women with pre-pregnancy BMI ≥ 25 kg/m² and less than 16 weeks of gestation. Before each scheduled appointment, we notified healthcare providers that their patients could be eligible for the study. At the end of the clinical encounter, healthcare providers asked patients if they would like to meet with research staff to hear about a healthful lifestyle study. Of 37 individuals approached by the healthcare providers, 30 agreed to meet, and all 30 were eligible and gave written informed consent to participate. Two participants later withdrew from the study during the intervention because of concerns about the time commitment (Fig. 1). Our recruitment goal of 30 women was based on recommendations for conducting qualitative in-depth interviews and the need to account for potential loss to follow-up due to pregnancy events or drop-outs. Because our primary aims of this pilot study were to show feasibility and provide qualitative evaluation of the interventions, we did not attempt to show effect on clinical outcomes for which power calculations would be indicated.

We gave women a $25 gift card after completing baseline surveys at enrollment, and mailed participants another $25 gift card after completing follow-up phone interviews. This study was approved by the Harvard Pilgrim Health Care Human Studies Committee. All participants provided written informed consent at enrollment.

Participants completed baseline surveys in person at enrollment, as well as follow-up online surveys and phone interviews at the end of the intervention (36 weeks gestation). Follow-up surveys repeated many of the baseline questions, and additionally asked questions about the intervention. We collected information on demographics, discussions with healthcare providers, attitudes related to weight status and pregnancy, opinions about the intervention, and included general food frequency questions based on surveys used in prior maternal-child health studies conducted at our institution [34]. End of intervention phone interviews followed a guide of open-ended questions to determine insights on what helped achieve goals, motivation, opinions of the health coach intervention, and areas for improvement (Additional file 1: Table S1). A single study staff member (MWS), who had experience with qualitative data collection, conducted all of the individual interviews.

We connected participants at enrollment (median gestation of 12.5 weeks, IQR: 11–15) with a trained health coach who called participants every 2–3 weeks until 36 weeks of gestation. During these phone calls, health coaches helped participants adopt and maintain new healthful lifestyle behaviors that were evidence-based, simple, and easy to track (Additional file 1: Table S2). Goals aimed to promote appropriate gestational weight gain by addressing several behavioral domains including diet, physical activity, screen time, and sleep.

During the first call, health coaches invited participants to prioritize these goals according to their level of self-efficacy, readiness to change, preferences, and values. Specifically, participants were asked to evaluate their lifestyle and identify areas they felt they wanted to focus on to improve their health and wellness. Participants were then asked about barriers to meeting these goals, and this allowed health coaches to create conversations about individual participant’s readiness to make changes and to discuss other resources that may be available to them to establish a prioritization that would be acceptable for each participant. Throughout the intervention, health coaches used principles of motivational interviewing that relied on a patient-centered approach to enhance readiness to change by exploring ambivalence and resistance to change [35]. At each check in, the health coach would discuss goals established and if goals were being met, per the participant report. If the goals were not being met, the health coach would discuss barriers and modify goals to be more realistic to the participants’ individual situation. Health coaches applied several behavioral theories to modify lifestyle including Social Cognitive Theory, Health Belief Model, and Protection Motivation Theory [36‐38]. Together, these theoretical bases emphasized the importance of self-regulation, developing specific behavior change plans, monitoring progress towards goals, attaining skills necessary to reach goals, and receiving support through the behavior change process. In addition to setting personal goals adapted to each participant preference and reality, health coaches also presented optimal goals, including targets for ideal cardiovascular health based on the Life’s Simple 7 health factors from the American Heart Association (Additional file 1: Table S2) [39].

During follow-up calls, health coaches monitored progress and helped adjust goals when necessary (e.g. too many goals, or the goal was too ambitious). Health coaches also addressed barriers and potential solutions with participants, and helped them target higher goal settings or select new goals when participants attained them. Health coaches sent emails or text messages depending on participant preferences to check-in about progress toward goals or clinical appointments between calls. Research staff (including an MD) met weekly with health coaches to review their conversations with participants, and to address any medical issues to ensure that it would be reported to the primary care provider obstetric team if appropriate. Although this procedure of notifying the medical team was in place throughout the study, no major concerns were raised to health coaches so there was no direct interaction between the health coaches and the medical team.

We accessed electronic medical records (EMR) and calculated GWG based on pre-pregnancy reported weight and the last weight recorded during pregnancy (> 34 weeks). We defined excessive GWG based on IOM definition per categories (> 11.3 kg of GWG in overweight category, > 9.1 kg of GWG in obesity category).

We measured feasibility via the ease of recruitment based on the ratio of enrolled women to the number of women approached about the study, and via maintenance of contact with a health coach based on the study participant attrition and follow-up survey completion rates (retention). We measured acceptability via survey items about components of the intervention and through qualitative comments from a semi-structured interview.

We performed all analyses using SAS software version 9.4 (SAS Institute Inc., Cary, NC). For most questions, participants chose between four categories: strongly agree, somewhat agree, somewhat disagree, and strongly disagree. For the group of questions asking about attitudes related to weight status and pregnancy, we pooled answers into “agree” and “disagree” for simplicity (Table 3). We conducted McNemar-Bowker symmetry tests to examine differences between baseline and follow-up survey responses.

We transcribed phone interviews verbatim during the calls, and completed content analysis of the transcripts using principles of the immersion-crystallization method [40]. This qualitative technique includes multiple rounds of “immersion” through close readings of transcripts, followed by reflection and the “crystallization” of emerging themes. Three investigators (MWS, DS, and MFH) read the transcripts, and one investigator (MWS) coded the transcripts. The three investigators (MWS, DS, and MFH) had to arrive at a unanimous consensus concerning the emerging themes after reading the verbatim transcripts. We reported the 5 themes that emerged from the analysis with representative quotes.

Participants at baseline (N = 30) had a median age of 31.5 years (IQR: 28.3–33), a median self-reported pre-pregnancy BMI of 27.3 kg/m² (IQR: 25.7–31.1), and 57% (17 of 30) were white, 30% (9 of 30) African-American, and 10% (3 of 30) Asian. Three-quarters (75%, 22 of 29) of participants completed a college degree or higher level of education. Two-thirds (67%, 20 of 30) of participants were married, 23% (7 of 30) were single and living with a partner or a significant other, and 10% (3 of 30) were single. Thirteen percent (4 of 30) reported an annual household income of $40,000 or less, 20% (6 of 30) reported incomes between $40,001 and $100,000, and 60% (18 of 30) reported incomes above $100,000 (Table 1).

Most (81%, 30 of 37) of the women potentially eligible to join the study and 100% (30 out of 30) of women approached by research staff about the study enrolled as participants. Participants completed baseline (N = 30) surveys in person at enrollment, as well as follow-up online surveys (N = 26) and phone interviews (N = 18) at the end of the intervention (36 weeks gestation) (Fig. 1).

We conducted 18 individual follow-up phone interviews for a mean of 12 min each (range: 8–20). Several important themes emerged about the quality of the intervention including motivation for personal health and the health of the baby, as well as accountability and support from health coaches. Participants also suggested ways to improve the intervention such as integration with standard clinical care and expressed desire for more visual resources such as mobile apps and tracking tools (Table 6).

The majority (11 of 18) cited primarily their own personal health as motivation to achieve lifestyle goals. Most of these women listed “general wellbeing” or “staying healthy and fit” as important drivers. Many participants also mentioned their bodies as motivation: “I want to be able to get my body back once I have the baby,” or “I definitely did not want to come out of pregnancy sloppy.” One woman said she “didn’t want high blood pressure,” and another said she is a “happier person when I’m moving and active.”

A few (3 of 18) cited principally the health of the baby as motivation: “I just want to have a healthy baby.” Another woman stated: “Making sure the baby is healthy to be honest with you. It was my number one motivation. I don’t want to deprive him of anything, taking prenatal vitamins every day, and yeah number one was ‘what does he need?’” Others (4 of 18) mentioned both personal and maternal sources of motivation: “I wanted to stay healthy for me and the baby.”

Several women discussed how the health coach intervention made them take responsibility for their lifestyle goals: “It was nice…to have someone that made me accountable…I didn’t have to do it on my own.” One woman particularly liked how the health coach filled a unique role as an independent source of feedback and described how “it was really good to have a third party outside of family and friends... [The health coach] helped me set goals and helped me keep track of where I stood with those goals.” Most women agreed that the health coaches provided helpful check-ins: “It was good having someone there that was sort of checking in on you so that you had that urge to please someone.”

Many participants welcomed the help they received from health coaches and thought they were “very supportive.” Here is how one woman described her experience with the health coach:

“I liked the fact that I had someone on my team. She never said I was doing something wrong so it was good positive reinforcement. It was great to bounce stuff off...My coach helped me be able to better grasp the need to do these things and I think at this point, the fruit has become a habit that I will keep after the baby. Now I’m loving the fruit. [My husband] comes in with tangerines and now I’m tearing the fruit up!”

One participant described her preference for phone communication over texts or emails from health coaches, she noted that the phone calls allowed the health coach to show “she had a supportive nature and genuinely was concerned.” Another woman agreed that this personal touch was important: “I always thought she was like a friend and she would go above and beyond to find answers. We talked about certain goals. I felt like I could go to her even if it wasn’t for a call. If had a question I could reach out to her.”

When asked about how the intervention could be improved, some women wished the health coach interactions were “more closely tied with the doctor’s appointments… It was too removed from the medical side.” Another woman agreed that incorporating the intervention into the clinic would be more convenient: “more integration with the doctors in general. If you’re already going in to see the doc once a month it would be great to see someone [a health coach] there already.” Besides convenience, another participant explained that more integration would also help keep herself accountable: “If the health coach had access to medical records she could ask what has happened over the last two weeks if she sees you gained 2 pounds. Sometimes you don’t want to admit you gained weight to a health coach, so it would be better if she already knew.”

Although the majority of participants thought the intervention was useful, several women said they “want more resources, not just a check in to talk.” Another participant explained that “all the goals are just discussed verbally. Maybe having more concrete goals that are maybe a worksheet or something.” One woman suggested “there’s no way to track your goal other than a conversation, so if there was some study app that made it a little more automated.” Another woman proposed front-loading these additional resources: “Maybe offering some guidelines in the beginning…starting off with ‘try these meal planners’ or ‘try this calorie tracking app.’”

We conducted this study at one clinic with a high socioeconomic sample of women, so results may vary for other women depending on the clinic and location; however, the ethnicities and ages of study participants were similar to the overall group medical practice. Because this study was a one-arm pilot without a control group, we were unable to determine if differences between baseline and follow-up surveys were responses to the intervention or the results of a progressing pregnancy. A few participants in interviews expressed dissatisfaction with the intervention after some lag time occurred when switching a few participants’ health coach due to an unexpected event unrelated to the study. Future interventions could use multiple health coaches to avoid unanticipated break of continuity in health coach support. Another limitation of the study is that we did not measure GWG. Yet, based on electronic medical record data, we observed that 35.7% (10 of 28) had excessive GWG, which is less than the 62% of overweight women and 56% of women with obesity who had excessive GWG in the national report based on CDC data from 2012 to 2013 [3].