This paper is in the following e-collection/theme issue:

Wearables and MHealth Reviews (255) Innovations and Technology for Healthy Eating Education (395) Innovations and Technology for Physical Activity Education (671) mHealth for Wellness, Behavior Change and Prevention (2681) Behavior Change (643) Pregnancy Information, Education and Lifestyle Interventions (276)

Published on in Vol 8, No 7 (2020): July

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/18255, first published .
Exclusively Digital Health Interventions Targeting Diet, Physical Activity, and Weight Gain in Pregnant Women: Systematic Review and Meta-Analysis

Exclusively Digital Health Interventions Targeting Diet, Physical Activity, and Weight Gain in Pregnant Women: Systematic Review and Meta-Analysis

Exclusively Digital Health Interventions Targeting Diet, Physical Activity, and Weight Gain in Pregnant Women: Systematic Review and Meta-Analysis

Authors of this article:

Alexandra Rhodes1 Author Orcid Image ;   Andrea D. Smith1 Author Orcid Image ;   Paul Chadwick1 Author Orcid Image ;   Helen Croker1 Author Orcid Image ;   Clare H. Llewellyn1 Author Orcid Image
Article Authors Cited by (36) Tweetations (22) Metrics

    Review

    University College London, London, United Kingdom

    *these authors contributed equally

    Corresponding Author:

    Alexandra Rhodes, MSc

    University College London

    1-19 Torrington Place

    London, WC1E 7HB

    United Kingdom

    Phone: 44 207679 1720

    Email: alexandra.rhodes.15@ucl.ac.uk


    Background: Interventions to promote a healthy diet, physical activity, and weight management during pregnancy are increasingly embracing digital technologies. Although some interventions have combined digital with interpersonal (face-to-face or telephone) delivery, others have relied exclusively on digital delivery. Exclusively digital interventions have the advantages of greater cost-effectiveness and broader reach and as such can be a valuable resource for health care providers.

    Objective: This systematic review aims to focus on exclusively digital interventions to determine their effectiveness, identify behavior change techniques (BCTs), and investigate user engagement.

    Methods: A total of 6 databases (Medical Literature Analysis and Retrieval System Online [MEDLINE], Excerpta Medica dataBASE [EMBASE], PsycINFO, Cumulated Index to Nursing and Allied Health Literature [CINAHL] Plus, Web of Science, and ProQuest) were searched for randomized controlled trials or pilot control trials of exclusively digital interventions to encourage healthy eating, physical activity, or appropriate weight gain during pregnancy. The outcome measures were gestational weight gain (GWG) and changes in physical activity and dietary behaviors. Study quality was assessed using the Cochrane Risk of Bias tool 2.0. Where possible, pooled effect sizes were calculated using a random effects meta-analysis.

    Results: In total, 11 studies met the inclusion criteria. The risk of bias was mostly high (n=5) or moderate (n=3). Of the 11 studies, 6 reported on GWG as the primary outcome, 4 of which also measured changes in physical activity and dietary behaviors, and 5 studies focused either on dietary behaviors only (n=2) or physical activity only (n=3). The meta-analyses showed no significant benefit of interventions on total GWG for either intention-to-treat data (−0.28 kg; 95% CI −1.43 to 0.87) or per-protocol data (−0.65 kg; 95% CI −1.98 to 0.67). Substantial heterogeneity in outcome measures of change in dietary behaviors and physical activity precluded further meta-analyses. BCT coding identified 7 BCTs that were common to all effective interventions. Effective interventions averaged over twice as many BCTs from the goals and planning, and feedback and monitoring domains as ineffective interventions. Data from the 6 studies reporting on user engagement indicated a positive association between high engagement with key BCTs and greater intervention effectiveness. Interventions using proactive messaging and feedback appeared to have higher levels of engagement.

    Conclusions: In contrast to interpersonal interventions, there is little evidence of the effectiveness of exclusively digital interventions to encourage a healthy diet, physical activity, or weight management during pregnancy. In this review, effective interventions used proactive messaging, such as reminders to engage in BCTs, feedback on progress, or tips, suggesting that interactivity may drive engagement and lead to greater effectiveness. Given the benefits of cost and reach of digital interventions, further research is needed to understand how to use advancing technologies to enhance user engagement and improve effectiveness.

    JMIR Mhealth Uhealth 2020;8(7):e18255

    doi:10.2196/18255

    Keywords

    gestational weight gain; digital interventions; behavior change techniques; user engagement; smartphone; mobile phone 


    Background

    Poor diet and lack of physical activity are 21st century problems contributing to the obesity crisis. In the United Kingdom, more than 50% of women of childbearing age are estimated to have overweight or obesity [1]. Pregnancy has been identified as a teachable moment [2], when women may be motivated to make lifestyle changes to improve their own health and the health of their unborn baby. Encouraging women to improve their diet and levels of physical activity is beneficial for not only supporting healthy gestational weight gain (GWG) and maternal health [3] but also developing behaviors that may potentially improve the health of the whole family.

    In many countries, GWG is monitored, and women are given guidance for recommended levels of weight gain [4]. The most widely used GWG guidelines are the US Institute of Medicine (IOM) 2009 guidelines, where the recommended range of weight gain is based on a woman’s pre-pregnancy BMI [5]. It is estimated that in many high-income countries, more than 50% of women gain excessive weight during pregnancy [6,7]. This is problematic because excessive GWG is associated with an increased risk of adverse health outcomes, such as gestational diabetes, large for gestational age babies, macrosomia, and cesarean section [6,8]. GWG is also associated with an increased risk of postpartum weight retention, which increases the likelihood of starting subsequent pregnancies with overweight or obesity [9]. The effects of excessive GWG are also believed to have an intergenerational impact, increasing the likelihood of overweight and obesity throughout the life of the baby [10,11]. In the United Kingdom, most women are not routinely weighed during pregnancy nor are they given specific advice on healthy levels of weight gain. However, they are provided with general advice to eat a healthy diet and participate in regular physical activity [12].

    Interventions targeting diet, physical activity, or both to encourage healthier lifestyles during pregnancy and reduce rates of excessive GWG have been shown to be effective, with diet-only interventions leading to greater weight reductions than physical activity alone or combined interventions [13-15]. A recent review of systematic reviews and meta-analyses reported reductions in GWG between 0.7 and 1.8 kg, as well as positive effects on maternal and infant health outcomes [16]. The majority of these lifestyle interventions used interpersonal delivery, either in person or telephone. Several more recent interventions have embraced digital delivery methods, recognizing their advantages of significantly lower costs and broader reach [17]. A self-managed digital intervention that can be delivered anytime and anywhere and at a lower cost to the patient and provider can be a valuable resource for health care providers, provided it can affect positive behavior change.

    In nonpregnant populations, digital interventions have been shown to be effective in changing nutritional behaviors [18], encouraging weight management [19], and improving levels of physical activity [20]. However, evidence of their effectiveness in improving lifestyle behaviors during pregnancy is mixed. In the past 3 years, 5 systematic reviews have explored various aspects of digital interventions to improve diet, increase physical activity, or manage weight during pregnancy [14,17,21-23]. Of the 4 including meta-analyses, 2 found no significant effect of the interventions [14,21], whereas one showed a significant result for limiting GWG, increasing physical activity, and reducing dietary energy intake in women with overweight or obesity [17] and another found a moderate effect on managing GWG among women of all BMIs [23]. However, except for Lau et al [17], who conducted a subgroup meta-analysis comprising 2 studies, these systematic reviews did not distinguish between interventions that combined digital with an interpersonal element of coaching or support (the majority of studies) and those that were exclusively digital. Moreover, although one review reported on usability [22], systematic reviews to date have not investigated user engagement, a vital component of self-managed digital interventions [24]. Finally, although previous systematic reviews have explored the behavior change techniques (BCTs) [25] used in this type of intervention [26], none have considered BCTs specifically within the context of digital interventions. As it cannot be assumed that BCTs have equal relevance to and effectiveness across different delivery methods, a review focused specifically on the role of BCTs in digitally delivered interventions for this population is a unique contribution to the literature.

    Objectives

    The aim of this systematic review was three-fold: (1) to determine the effectiveness of exclusively digital diet and physical activity interventions to improve lifestyle behaviors or avoid excessive weight gain during pregnancy; (2) to investigate user engagement with the interventions; and (3) to identify and assess the usage of BCTs within the interventions.


    Review Protocol

    This systematic review and meta-analysis follows the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [27] and is registered with the International Prospective Register of Systematic Reviews (CRD42019124838; see Multimedia Appendix 1 for the PRISMA checklist).

    Search Strategy

    A search of 6 databases (MEDLINE [Medical Literature Analysis and Retrieval System Online], PsycINFO, EMBASE, Cumulated Index to Nursing and Allied Health Literature [CINAHL] Plus, Web of Science, and ProQuest Dissertations and Theses) was conducted in February 2019 to identify relevant studies. Advanced searches of keywords and index terms covered 4 concept areas (pregnancy status, diet or physical activity intervention, digital technology, and study design) and were tailored according to each database (Multimedia Appendix 2). The Cochrane Library was also searched for related systematic reviews. Their reference lists along with those of eligible studies were hand searched. Once duplicates had been removed, 2 authors (AR and PC) independently screened and assessed each study for eligibility based on title and abstract.

    Inclusion and Exclusion Criteria

    Studies were included in the study if they fulfilled the following population, interventions, comparators, and outcomes criteria.

    Population

    Pregnant women over the age of 18 years, of all BMIs, were included in this study. However, pregnant women with physical or mental health issues that would preclude them from participating in a diet- or physical activity–based intervention were excluded.

    Interventions

    Digital interventions targeting dietary behaviors or physical activity in pregnancy, with the aim of improving diet or physical activity during pregnancy or managing GWG, were included. Interventions aimed at increasing GWG were excluded. Interventions were exclusively digital and used text messages, apps or websites. Initial in-person or telephone study briefing sessions were deemed acceptable, as they seemed to reflect real-world situations in which health care professionals would introduce an intervention to pregnant women as a part of an antenatal care program. Interventions using interpersonal coaching or support beyond this were excluded, as were digital interventions delivered in a health care setting.

    Comparators

    Comparators were usual antenatal care, minimal interventions (ie, information only rather than active behavior change), or nondiet or physical activity interventions.

    Outcomes

    The primary outcomes were GWG (measured as total gain in kilos or compliance with IOM GWG guidelines [5]), changes in dietary behaviors, and changes in levels of physical activity. The secondary outcome was engagement, which was measured by intervention attrition rates and usage of the intervention features. BCTs were coded according to the BCT Taxonomy (version 1) [25].

    Study Design

    Only randomized controlled trials (RCTs) and randomized pilot studies were included in this review.

    Data Extraction and Data Synthesis

    Data extracted for the systematic review included author and date of publication, geographical region, study design, behaviors targeted and specific behavioral goals, sample size, participant information, the technology used, intervention features, the theory used, gestational week in which intervention started, length of intervention, nature of control, attrition rate, engagement levels, outcome measures, and outcomes. Data extraction was completed independently by 2 authors (AR and HC). In addition, 2 authors (AR and PC) independently coded the BCTs within each intervention according to the BCT Taxonomy (version 1) [25]. If available and required, study development papers and protocols were retrieved for this purpose. In most instances, the authors were contacted for additional information. BCTs were coded only if there was unequivocal evidence of their existence [25]. Disagreements were discussed to reach consensus.

    Quality Assessment

    Two authors (AR and AS) independently evaluated the risk of bias within studies using the Cochrane Collaboration Risk of Bias (RoB) 2.0 tool for assessing the risk of bias (the Cochrane Collaboration) [28]. The 5 domains evaluated were risk of bias arising from the randomization process, risk of bias because of deviations from the intended interventions, missing outcome data, risk of bias in the measurement of the outcome, and risk of bias in the selection of the reported result. Bias was classified as low risk, high risk, or some concerns according to predetermined criteria set by RoB 2.0. Rating discrepancies among the authors were resolved through discussion. The risk of bias across studies could not be evaluated because of the small number of studies included in the meta-analyses [29].

    Data Analysis

    Given the substantial heterogeneity of reported outcome measures in the identified studies, data could only be quantitatively pooled for meta-analysis from studies measuring GWG. Only 4 studies used intention-to-treat (ITT) analysis; therefore, separate analyses were conducted for ITT data and per-protocol (PP) data. Meta-analysis was used to determine the differences in mean total GWG (in kg) from baseline to postintervention using the inverse variance method. The odds ratio (OR) was meta-analyzed for studies reporting GWG as a dichotomous outcome (proportion of women exceeding IOM guidelines) using the Mantel-Haenszel method. The test for the overall pooled effect estimate was assessed using Z‐statistics at P=.04. Heterogeneity between studies was evaluated using the Cochran Q (chi-square test) and the I2 statistics in the Review Manager 5.3 (the Cochrane Collaboration) [30]. Preplanned subgroup analyses were conducted comparing studies where BCTs could be identified in initial briefing sessions with those where none were apparent.


    Study Selection

    The selection process is illustrated in Figure 1. A systematic search of 6 literature databases identified 623 nonduplicate study records. After the assessment of eligibility in accordance with the inclusion and exclusion criteria, 11 eligible studies were identified, of which 6 studies were included in subsequent meta-analyses.

    ‎
    Figure 1. Study selection process.
    View this figure

    Study Characteristics

    Table 1 summarizes the characteristics of the 11 studies included in this review. All studies were published between 2012 and 2019. Of the 11 studies, 7 were randomized pilot or feasibility studies [31-37] and 4 were RCTs [38-41]. Of the 4 RCTs, 2 reported being adequately powered [39,41], whereas the other 2 reported being underpowered as a result of a small starting sample [40] or low follow-up rate [38]. Overall, 9 studies took place in the United States [31-33,35,37-41] and 2 in Australia [34,36]. The sample sizes varied from 35 to 1689. In addition, 2 studies [31,38] targeted diet only, 4 studies [33,36,39,40] targeted physical activity only, and the remaining 5 studies [32,34,35,37,41] targeted both diet and physical activity. Seven studies reported on only one outcome: GWG [35,41], dietary [31,38], or physical activity [33,36,40] behaviors. The remaining 4 studies [32,34,37,39] reported on GWG, as well as changes in diet and physical activity. Three studies focused specifically on women with overweight or obesity [32,34,35], and 3 studies [33,39,40] focused on inactive or sedentary women.

    The delivery method varied across studies with 4 using text messaging [31,32,38,40], 3 using an app [33,35,37], 3 using a website [36,39,41], and 1 combining text messaging with a website [34]. In total, 7 studies included an interpersonal briefing session at the start of the study [31-34,37,39,40]. In 4 studies [31,32,37,40], these sessions were for screening or study measures only, but 3 studies [33,34,39] included discussions with intervention participants about the intervention features. In one study [39], these discussions involved an in-person tutorial on how to use the website and its features and practice tracking physical activity. In another study [34], the researcher discussed individual GWG targets and weight monitoring and asked participants to set a physical activity or dietary goal. In a third study [33], a 30-min in-person session covered physical activity recommendations, goal setting, problem solving, social support, and planning for lapses.

    Table 1. Study characteristics.
    Authors, yearCountryStudy designSample size, nTechnology usedBehaviors targetedOutcomes measuredParticipants
    Evans et al, 2012 [31]United StatesPilot90TextDietDietLow-income, underserved pregnant women
    Pollak et al, 2014 [32]United StatesPilot35TextDiet and physical activityGWGa, diet and physical activityBMI 25-40 kg/m2; gestation 12-21 weeks
    Evans et al, 2015 [38]United StatesRCTb996TextDietDietMilitary women; gestation <14 weeks
    Smith et al, 2016 [39]United StatesRCT51WebsitePhysical activityGWG, diet, and physical activitySedentary women; gestation 10-14 weeks
    Choi et al, 2016 [33]United StatesPilot30AppPhysical activityPhysical activityPhysically inactive women; gestation 10-20 weeks
    Willcox et al, 2017 [34]AustraliaPilot91Text and websiteDiet and physical activityGWG, diet, and physical activityBMI>25 kg/m2; gestation 10-17.6 weeks
    Redman et al, 2017 [35]United StatesPilot54AppDiet and physical activityGWGBMI>25 kg/m2; gestation 10.4-13.6 weeks
    Hayman et al, 2017 [36]AustraliaPilot77WebsitePhysical activityPhysical activityGestation 10-20 weeks
    Huberty et al, 2017 [40]United StatesRCT80TextPhysical activityPhysical activityNot meeting physical activity recommendations; gestation 8-16 weeks
    Olson et al, 2018 [41]United StatesRCT1689WebsiteDiet and physical activityGWGBMI 18.5-35 kg/m2; gestation <20 weeks
    Dahl et al, 2018 [37]United StatesPilot87AppDiet and physical activityGWGBMI≥18.5 kg/m2; gestation <20 weeks

    aGWG: gestational weight gain.

    bRCT: randomized controlled trial.

    Risk of Bias

    Table 2 summarizes the study quality assessment. The overall study quality was variable. Five studies were deemed to have an overall high risk of bias, 3 had a low risk of bias, and 3 were classified as having some concerns.

    Table 2. Risk of bias summary.
    Study, yearDomain 1: risk of bias arising from the randomization processDomain 2: risk of bias because of deviations from the intended interventionsDomain 3: missing outcome dataDomain 4: risk of bias in the measurement of the outcomeDomain 5: risk of bias in selection of the reported resultOverall risk of bias
    Evans et al, 2012 [31]HighHighHighSome concernsLowHigh
    Pollak et al, 2014 [32]HighLowLowSome concernsLowHigh
    Evans et al, 2015 [38]LowLowLowSome concernsLowSome concerns
    Smith et al, 2016 [39]LowHighHighHighLowHigh
    Choi et al, 2016 [33]LowLowLowLowLowLow
    Willcox et al, 2017 [34]LowSome concernsLowSome concernsLowSome concerns
    Redman et al, 2017 [35]LowLowLowLowLowLow
    Hayman et al, 2017 [36]LowHighHighLowLowHigh
    Huberty et al, 2017 [40]Some concernsLowLowLowLowSome concerns
    Olson et al, 2018 [41]LowLowLowLowLowLow
    Dahl et al, 2018 [37]Some concernsHighHighSome concernsLowHigh

    Description of the Interventions

    Multimedia Appendix 3 summarizes the intervention features, outcome measures, effectiveness, attrition, and engagement data. All interventions were theory-based, with social cognitive theory [42] being the most widely used (n=8). All trials started in the first or second trimester of pregnancy. The study duration varied considerably with one trial lasting 4 weeks [36], 2 trials lasting 12 weeks [33,37], one trial lasting 16 weeks [32], and the remaining trials lasting 20 weeks or more, completing at or close to term. Most studies compared interventions with usual care [31,32,34,35,38,39] or access to information-only aspects of the intervention [33,36,40,41]. In one study, the control was an equivalently structured intervention targeting stress reduction [37].

    Effectiveness of Interventions

    Of the 11 studies, 3 reported significant positive effects of their interventions on GWG [34,35] and physical activity [34,36] in comparison with control groups.

    The 6 studies with GWG as the primary outcome varied in their measurement of total GWG. Three studies used the difference between last measured weight before delivery (34-37 weeks) and baseline weight (10-17 weeks) [34,35,41]. Two studies [37,39] used self-reported prepregnancy weight as the starting weight, and one study [32] used a model of estimated mean weights at 16 and 40 weeks. One study showed significantly lower total GWG among intervention participants [34], whereas another showed significantly fewer intervention participants exceeding the IOM guidelines [35]. The remaining studies [32,37,39,41] showed no significant difference between the intervention and control groups on any GWG measures. A meta-analysis of ITT data (n=3) showed a nonsignificant effect of the interventions, with a mean difference in total GWG of −0.28 kg (95% CI −1.43 to 0.87) using the inverse variance method and a fixed effects model (I2=0%; P=.38; Figure 2). The mean difference in total GWG for PP data (n=4) was −0.65 kg (95% CI −1.98 to 0.67; I2=53%; P=.10; Figure 3). The subgroup analyses revealed no significant change in this result (Multimedia Appendix 4). A meta-analysis of studies reporting PP percentages exceeding IOM guidelines showed no effect of interventions relative to comparators (OR 1.02, 95% CI 0.82-1.27; I2=45%; P=.16; Figure 4).

    Of the 7 studies reporting physical activity, 3 showed significant positive effects of the intervention on levels of physical activity [34,36,39]. Of these, one study relied on self-reported physical activity and showed significantly smaller reductions in total, light-intensity, and moderate-intensity physical activities in the intervention group compared with the control group [34]. Two other studies used smart technology to provide an objective measure of physical activity (Fitbit [36] and SenseWear Mini Arm Band [39]). One study reported a significant increase in moderate-to-vigorous physical activity in the intervention group compared with the control group, although over a 4-week period only [36]. Another study reported significantly greater levels of sustained physical activity for intervention participants compared with control participants in midpregnancy, but the effect had disappeared by the end of the intervention [39]. Only one of the 6 studies reporting on dietary behaviors was effective in improving diet [37]. Using a self-report measure (the Rapid Eating and Activity Assessment for Participants Short Scale), intervention participants in this study scored significantly higher on healthy eating practices (measuring meal behaviors and serving frequencies) compared with the control participants.

    ‎
    Figure 2. Pooled analysis of digital interventions on total gestational weight gain (kg)—intention-to-treat studies.
    View this figure
    ‎
    Figure 3. Pooled analysis of digital interventions on total gestational weight gain (kg)—per-protocol studies.
    View this figure
    ‎
    Figure 4. Pooled analysis of digital interventions on the percentage of women exceeding Institute of Medicine guidelines—per-protocol studies.
    View this figure

    BCTs

    Multimedia Appendix 5 summarizes the 23 different BCTs identified within the interventions. Only one study specified all the included BCTs [34], and this study reported the highest number of BCTs (n=17). In 2 interventions, only 1 BCT was evident [31,38]. In the remaining interventions, the number of BCTs ranged from 5 to 15. The 3 effective interventions used, on average, twice the number of BCTs compared with other interventions (mean 14, SD 2.9 vs mean 6.8, SD 4.1). Information about health consequences was the only BCT to be used in all interventions. Beyond this, goal setting (behavior) appeared in 8 interventions and problem solving and self-monitoring (behavior and outcome) in 7 interventions. Seven BCTs were common to the 3 interventions, showing a significant effect [34-36]. These were goal setting (behavior), problem solving, review of behavior goals, feedback on behavior, social support, information about health consequences, and information about emotional consequences. Review of the behavior goal was the only BCT used exclusively in the 3 effective interventions. The 3 information-only interventions that included no active or interactive BCTs, such as goal setting, self-monitoring, problem solving, or feedback, were ineffective [31,38,40].

    Engagement With Interventions

    Attrition rates were reported by all the studies, although 2 studies [31,40] only provided figures for all participants rather than separating intervention and control participants (Multimedia Appendix 3). Six studies reported intervention attrition rates of 10% or less [33-35,39-41]. These studies identified the reasons for dropping out of the study by distinguishing between medical and study-related reasons. Three of these studies [34,35,40] incentivized participants and 2 [33,39] introduced an element of self-selection by recruiting women who were motivated or willing to increase their physical activity. In the remaining 5 studies, intervention attrition was more than 30% and lost to follow-up reasons were not explained beyond being unable to recontact participants.

    Six studies [32-36,41] reported intervention engagement levels using a variety of measures, including usage of self-monitoring, goal setting, action planning, and social media features, response to texts, completion of tasks, and website logins (Multimedia Appendix 3). Four studies [32,34,36,41] evaluated participants’ views of the intervention, although only one study [36] explored the user experience of the technology. Four studies [32,34-36] reported engagement levels over 70%, including the 3 interventions with significant effects [34-36]. A further study [33] started with a similarly high level of engagement, although it fell to below 50% over the course of the 12-week intervention. The final study [41] reported an engagement level of 46%. Five studies [32-36] integrated interactive elements to encourage engagement with the intervention. Two interventions [32,34] sent 4 or more text messages per week, encouraging self-monitoring and giving tailored feedback, and 3 studies [33,35,36] used in-app messaging to provide tailored feedback.


    Principal Findings

    The aim of this systematic review was to determine the effectiveness of diet and physical activity interventions during pregnancy delivered using exclusively digital technology, and to provide insight into how BCTs and engagement with intervention features might be driving effectiveness. A total of 11 studies were identified, all of which were published from 2012 onward, with 6 studies published in 2017 and 2018. App and mobile-accessible website interventions appeared only in the last 2 years, reflecting the emergent nature of mobile health interventions to encourage healthy behaviors during pregnancy. Meta-analyses showed no significant benefit of exclusively digital interventions on total GWG. Substantial heterogeneity in outcome measures of change in dietary behaviors and physical activity precluded further meta-analyses. BCT coding identified 7 BCTs that were common to all effective interventions. Effective interventions averaged over twice as many BCTs from the goals and planning and feedback and monitoring domains as ineffective interventions. Six studies reported on user engagement, and their data indicated a positive association between high engagement with key BCTs and greater intervention effectiveness. Interventions using proactive messaging, such as reminding participants to engage in BCTs or providing feedback or tips, appeared to have higher levels of engagement.

    Effectiveness

    Meta-analyses of the digital interventions measuring GWG showed no effect on the total GWG or weight gain within the IOM guidelines. Although the majority of these studies were pilot RCTs and insufficiently powered to detect an effect, these findings indicate that exclusively digital interventions to manage GWG may be less effective than those using interpersonal delivery. Lack of consistency in outcome measures precluded meta-analyses of the effects of digital interventions on dietary behaviors and physical activity. Only 3 of the 7 studies measuring changes in physical activity reported significant effects of the intervention, suggesting that for physical activity interventions during pregnancy, digital delivery may similarly be less effective than interpersonal delivery [43].

    The 11 interventions varied considerably in terms of not only the targeted behaviors but also the technologies, functionalities, and BCTs used. As such, it would be premature to conclude that exclusively digital delivery methods per se are less effective than interpersonal delivery methods for lifestyle interventions during pregnancy. Indeed, one of the included studies made a direct comparison of digital delivery and in-person delivery of the same intervention [35]. It found the intervention to be effective via both delivery methods, with digital delivery showing greater adherence and lower costs (for both participants and clinics) compared with in-person delivery.

    BCTs

    This systematic review aimed to identify the BCTs associated with effective interventions. The number of identifiable BCTs ranged from 1 to 17, with the 2 most effective interventions [34,35] using the highest number (n=17 and n=15). The average number of BCTs was 9 compared with approximately 5 reported in 2 earlier systematic reviews of lifestyle interventions targeting pregnant women [44,45]. It is unclear whether this increase reflects a trend toward greater intervention complexity, reflects the opportunity digital interventions afford to include more components, or is simply a matter of improved reporting of BCTs. Consistent with previous systematic reviews, this review found that effective interventions tended to report a greater number of BCTs [44,46]. A meta-analysis of 122 physical activity and healthy eating interventions (for all adults) showed effectiveness to be a function of not simply the number of BCTs but particular BCTs—self-monitoring and at least one other technique derived from control theory [47]. In this review, self-monitoring appeared in 7 interventions but was notably absent from one of the 3 interventions showing a significant effect [36]. The 3 effective interventions did however average over twice as many goals and planning and feedback and monitoring BCTs as ineffective interventions (mean 7.6, SD 2.1 and mean 3.4, SD 2.9, respectively).

    There was considerable variation in the execution and delivery of BCTs. For example, in some studies, participants were invited to set a single goal, whereas in others, they were able to set multiple goals. In some instances, participants were encouraged to choose their own goal, whereas in others, the goal was prescribed. Similarly, some interventions required participants to submit self-monitoring data regularly, whereas others recommended and provided functionality for self-monitoring but did not make it obligatory. Four studies proactively messaged participants to remind them to self-monitor [32-34,37], whereas one messaged participants only if they failed to self-monitor [35]. In 3 of the studies that incorporated an initial in-person session for intervention participants [33,34,39], one or more BCTs were identifiable at this stage, raising the question as to whether the content of these sessions contained sufficient BCTs in their own right to bring about a change. The influence of these variations in the context, execution, and delivery of BCTs on intervention effectiveness could not be quantified by the methods used in this study. Given the interactive and dynamic nature of digital interventions, additional measures may be needed to capture the impact of features, such as the timing of delivery and degree of individual tailoring of BCTs.

    Consistent with other systematic reviews of this type of intervention [44,48,49], information about health consequences was the most widely used BCT, which featured in all interventions. Goal setting (behavior) was the next most widely used BCT, appearing in all but 3 text message–only interventions [31,38,40]. Problem solving,self-monitoring of behavior, self-monitoring of outcomes, and instructions on how to perform a behavior all appeared in 7 interventions. Feedback on behavior was provided in 6 interventions, including the 3 [34-36] reporting significant effects of the interventions. The BCT review behavior goal was only present in the 3 effective interventions [34-36], suggesting that this may be a critical active ingredient in these digital interventions. It is possible that review behavior goal in combination with self-monitoring of behavior and feedback on behavior work together to support the self-regulation of energy balance behaviors during pregnancy.

    Social support was present in 6 interventions [33-37,40], including the 3 effective interventions. Once again, the execution of social support varied, ranging from advice on how to seek support to online group forums for participants. There is no consensus on whether social support or interaction with other participants improves intervention effectiveness, and no clear pattern emerged from this review [14,23,50]. More research is needed to understand the type of social support that is most beneficial to digital interventions encouraging healthy behaviors during pregnancy.

    Insufficient description of intervention components, coupled with a lack of systematic recording of BCTs, compromised the quality of the BCT analysis. Only 1 study provided details of all the BCTs [34] used in the intervention, whereas the presence of BCTs had to be inferred from descriptions of the interventions in all other studies. This raises the possibility that there may be additional but unreported BCTs in other studies. Previous studies that have coded BCTs used in gestational weight management trials have called for greater clarity and accuracy in the reporting of BCTs [26]. Without systematic reporting of active intervention ingredients, it is difficult to precisely determine which BCTs may be driving effectiveness.

    Engagement

    Six studies provided measures of user engagement [32-36,41]. These varied considerably, including the number of replies to texts, frequency of inputting weight monitoring data and logging onto and viewing web pages. Only one study [36] provided feedback on user experience. Given the importance of user engagement to the success or otherwise of self-managed digital interventions, more detailed and standardized measures could facilitate better evaluation and cross-study comparison [51]. Perski et al [52] proposed more comprehensive measures, including both the extent (ie, amount, frequency, duration, depth) of usage and the user experience. Reinforcing the need for a more holistic evaluation of engagement, Yardley et al [24] proposed identifying and reporting on effective engagement rather than simply higher levels of engagement. The combination of web analytics and survey feedback clearly offers the opportunity to develop specific and relevant indices of engagement [53].

    The 3 effective interventions [34-36] all reported engagement levels of over 70% with key BCTs (goal setting [34,36], self-monitoring [34,35], and action planning [36]). Conversely, the study with the lowest engagement level, where only 46% of participants logged onto the website at least once every 45 days, and the use of goal setting and self-monitoring features was 35% and 23% of the participants respectively, reported no effect of the intervention [41]. These findings suggest that ineffectiveness may be partially a function of poor engagement with key BCTs rather than poorly designed interventions per se. Supporting this hypothesis, this study with low levels of engagement [41] conducted secondary analyses investigating whether usage patterns of the intervention features reduced the risk of excessive GWG and found frequent usage patterns were associated with lower total GWG [54,55]. In addition, the use of the dietary tool (goal setting and self-monitoring) was associated with improved GWG management for women with normal BMI, although not for those with high BMI.

    One consistent feature of the interventions reporting the levels of engagement over 70% was regular in-app messaging or text messaging giving encouragement, reminders to self-monitor, or tailored feedback on progress. Prompts and reminders have been shown to promote engagement in digital interventions [56]. Similarly, tailoring messages to the characteristics and usage patterns of the individual has been shown to improve adherence [57]. Notably, the study [41] in which participants were sent a generic weekly email reported particularly low levels of engagement. The frequency and timing of these messages are also important [58]. One study [34] delivering 4 to 5 texts per week found that 79% of participants thought the frequency of messages was about right, although 21% thought it was too high. Another study [40] investigated the dose and timing of messages to promote physical activity by comparing 3 texts per week with daily texts and found daily texts to be less effective, indicating that too much messaging can be counterproductive. None of the studies referred to the use of gamification techniques to promote engagement, although elements of some of the interventions could potentially be classified as gamification, such as team challenges [37,59]. Incorporating gamification features, such as badges and challenges, has been shown to increase regular engagement and immersion in digital health interventions [60,61].

    The final issue regarding engagement concerns who the interventions are reaching. Only one study reported (in follow-up analyses) on high versus low engagers, revealing significant differences according to ethnicity, income group, BMI, and partner status [54]. Often, it is those who would benefit most from behavior change who are least likely to engage in behavior change interventions [62]. Greater insight into who engages with the interventions could enhance learnings from these studies.

    Strengths and Limitations

    A strength of this systematic review is that it is the first to focus on exclusively digital interventions to promote healthy dietary behaviors, physical activity, or weight management during pregnancy. In addition to evaluating their effectiveness, this review conducted a rigorous assessment of BCTs and participant engagement to provide detailed insight into what may be driving effectiveness—a crucial step if the cost and reach advantages of digital interventions are to be leveraged. However, there are several limitations to this systematic review. First, most of the studies included were pilot studies rather than RCTs and, as such, were not adequately powered to show effect sizes. Moreover, there was considerable heterogeneity of intervention features and outcome measures, and several studies reported the results from PP analyses rather than ITT. As such, the results of the meta-analyses should be interpreted with caution. Second, the risk of bias across the studies was moderate to high, with 5 studies scoring overall high and a further 3 scoring as some concerns, as assessed by RoB 2.0. Third, the timing of the interventions within pregnancy varied both in terms of the start point within the gestational window and duration. This, coupled with inconsistent measures of GWG and, in some cases, reliance on self-reported weight measures should be considered when appraising the findings. Finally, limited reporting of intervention features meant that not all BCTs were recorded. Providing more detailed descriptions of the interventions’ design and content (in supplementary files) would augment shared learnings from these studies. Similarly, more detailed and consistent engagement measures would have enhanced the interpretation of user engagement data.

    Conclusions

    Meta-analyses show that the mean impact on GWG of exclusively digital interventions targeting dietary behaviors, physical activity, and weight management during pregnancy to be nonsignificant, meaning that the current exclusively digital interventions are less effective than interpersonal interventions in this field. There was considerable variation in intervention effectiveness across the 11 studies, with 3 studies from 2017 reporting significantly positive effects of their interventions. Limited data precluded confident identification of the ingredients of successful interventions, although this review suggests that variation in effectiveness could be partially explained by the BCTs used and levels of interactivity to encourage engagement with the intervention features. Effective interventions used more BCTs (particularly BCTs from goals and planning and feedback and monitoring domains) and reported higher levels of engagement with key BCTs. Effective interventions also used interactivity, in the form of messages of encouragement, personalized feedback, and prompts to remind participants to use key BCTs, such as goal setting and self-monitoring, to promote engagement.

    There are several compelling reasons for considering using digital interventions to promote healthy energy balance behaviors during pregnancy: smartphone ownership is over 90% among women of childbearing age [63] and usage of pregnancy apps is pervasive [64]; digital interventions have broader reach and lower costs than interpersonal interventions [35,65]; and apps have been shown to be particularly successful in reaching those who may be less likely to engage with traditional antenatal health care [66]. Meanwhile, midwives frequently report that they have neither the time nor expertise to advise pregnant women on physical activity or healthy eating [67]. Future research needs to consider how to seize the opportunities presented by new technologies to enhance interactivity, improve user engagement, and bring greater effectiveness to these digital interventions.

    Acknowledgments

    This work was supported by the Economic and Social Research Fund (grant number ES/P000592/1) and UK charity, Best Beginnings. The authors would like to thank the authors of the included studies for the additional information they provided to help this study’s analyses.

    Conflicts of Interest

    CL reports grants from the Economic and Social Research Fund, grants from Best Beginnings (UK charity), during the conduct of the study; personal fees from Yellow Kite; personal fees from the experiment; and personal fees from Diamond Inc, outside the submitted work. The remaining authors have no conflicts of interest to declare.

    ‎
    Multimedia Appendix 1

    PRISMA Checklist.

    PDF File (Adobe PDF File), 52 KB
    ‎
    Multimedia Appendix 2

    Subject headings and key words for search.

    PDF File (Adobe PDF File), 59 KB
    ‎
    Multimedia Appendix 3

    Intervention features, engagement and effectiveness.

    PDF File (Adobe PDF File), 63 KB
    ‎
    Multimedia Appendix 4

    Subgroup analyses.

    DOCX File , 88 KB
    ‎
    Multimedia Appendix 5

    Summary of BCTs.

    PDF File (Adobe PDF File), 57 KB
    1. Baker C. Obesity Statistics. House of Commons Library - UK Parliament. 2019.   URL: https://researchbriefings.parliament.uk/ResearchBriefing/Summary/SN03336 [accessed 2019-07-12]
    2. Phelan S. Pregnancy: a 'teachable moment' for weight control and obesity prevention. Am J Obstet Gynecol 2010 Feb;202(2):135.e1-135.e8 [FREE Full text] [CrossRef] [Medline]
    3. Raghavan R, Dreibelbis C, Kingshipp B, Wong Y, Abrams B, Gernand A, et al. Dietary patterns before and during pregnancy and maternal outcomes: a systematic review. Am J Clin Nutr 2019 Mar 1;109(Suppl_7):705S-728S. [CrossRef] [Medline]
    4. Rasmussen K, Yaktine A. Weight Gain During Pregnancy: Reexamining the Guidelines. New York, USA: National Academies Press; 2009.
    5. Scott C, Andersen CT, Valdez N, Mardones F, Nohr EA, Poston L, et al. No global consensus: a cross-sectional survey of maternal weight policies. BMC Pregnancy Childbirth 2014 May 15;14:167 [FREE Full text] [CrossRef] [Medline]
    6. Goldstein R, Abell S, Ranasinha S, Misso M, Boyle J, Harrison C, et al. Gestational weight gain across continents and ethnicity: systematic review and meta-analysis of maternal and infant outcomes in more than one million women. BMC Med 2018 Aug 31;16(1):153 [FREE Full text] [CrossRef] [Medline]
    7. Crozier S, Inskip H, Godfrey K, Cooper C, Harvey N, Cole Z, Southampton Women's Survey Study Group. Weight gain in pregnancy and childhood body composition: findings from the Southampton women's survey. Am J Clin Nutr 2010 Jun;91(6):1745-1751 [FREE Full text] [CrossRef] [Medline]
    8. Hedderson MM, Gunderson EP, Ferrara A. Gestational weight gain and risk of gestational diabetes mellitus. Obstet Gynecol 2010 Mar;115(3):597-604 [FREE Full text] [CrossRef] [Medline]
    9. Gilmore LA, Klempel-Donchenko M, Redman LM. Pregnancy as a window to future health: excessive gestational weight gain and obesity. Semin Perinatol 2015 Jun;39(4):296-303 [FREE Full text] [CrossRef] [Medline]
    10. Mamun AA, Mannan M, Doi SA. Gestational weight gain in relation to offspring obesity over the life course: a systematic review and bias-adjusted meta-analysis. Obes Rev 2014 Apr;15(4):338-347. [CrossRef] [Medline]
    11. Josey MJ, McCullough LE, Hoyo C, Williams-DeVane C. Overall gestational weight gain mediates the relationship between maternal and child obesity. BMC Public Health 2019 Aug 7;19(1):1062 [FREE Full text] [CrossRef] [Medline]
    12. Recommendation 2 Pregnant Women. The National Institute for Health and Care Excellence.   URL: https://www.nice.org.uk/guidance/ph27/chapter/1-Recommendations#recommendation-2-pregnant-women [accessed 2019-08-22]
    13. Craemer K, Sampene E, Safdar N, Antony K, Wautlet C. Nutrition and exercise strategies to prevent excessive pregnancy weight gain: a meta-analysis. AJP Rep 2019 Jan;9(1):e92-120 [FREE Full text] [CrossRef] [Medline]
    14. Walker R, Bennett C, Blumfield M, Gwini S, Ma J, Wang F, et al. Attenuating pregnancy weight gain-what works and why: a systematic review and meta-analysis. Nutrients 2018 Jul 22;10(7):- [FREE Full text] [CrossRef] [Medline]
    15. Shieh C, Cullen D, Pike C, Pressler S. Intervention strategies for preventing excessive gestational weight gain: systematic review and meta-analysis. Obes Rev 2018 Aug;19(8):1093-1109. [CrossRef] [Medline]
    16. Farpour-Lambert N, Ells L, de Tejada BM, Scott C. Obesity and weight gain in pregnancy and postpartum: an evidence review of lifestyle interventions to inform maternal and child health policies. Front Endocrinol (Lausanne) 2018;9:546 [FREE Full text] [CrossRef] [Medline]
    17. Lau Y, Klainin-Yobas P, Htun T, Wong S, Tan K, Ho-Lim T, et al. Electronic-based lifestyle interventions in overweight or obese perinatal women: a systematic review and meta-analysis. Obes Rev 2017 Sep;18(9):1071-1087. [CrossRef] [Medline]
    18. Villinger K, Wahl D, Boeing H, Schupp H, Renner B. The effectiveness of app-based mobile interventions on nutrition behaviours and nutrition-related health outcomes: a systematic review and meta-analysis. Obes Rev 2019 Oct;20(10):1465-1484 [FREE Full text] [CrossRef] [Medline]
    19. Hutchesson MJ, Rollo ME, Krukowski R, Ells L, Harvey J, Morgan PJ, et al. eHealth interventions for the prevention and treatment of overweight and obesity in adults: a systematic review with meta-analysis. Obes Rev 2015 May;16(5):376-392. [CrossRef] [Medline]
    20. Stockwell S, Schofield P, Fisher A, Firth J, Jackson SE, Stubbs B, et al. Digital behavior change interventions to promote physical activity and/or reduce sedentary behavior in older adults: a systematic review and meta-analysis. Exp Gerontol 2019 Jun;120:68-87. [CrossRef] [Medline]
    21. Sherifali D, Nerenberg K, Wilson S, Semeniuk K, Ali M, Redman L, et al. The effectiveness of ehealth technologies on weight management in pregnant and postpartum women: systematic review and meta-analysis. J Med Internet Res 2017 Oct 13;19(10):e337 [FREE Full text] [CrossRef] [Medline]
    22. Overdijkink SB, Velu AV, Rosman AN, van Beukering MD, Kok M, Steegers-Theunissen RP. The usability and effectiveness of mobile health technology-based lifestyle and medical intervention apps supporting health care during pregnancy: systematic review. JMIR Mhealth Uhealth 2018 Apr 24;6(4):e109 [FREE Full text] [CrossRef] [Medline]
    23. Chan KL, Chen M. Effects of social media and mobile health apps on pregnancy care: meta-analysis. JMIR Mhealth Uhealth 2019 Jan 30;7(1):e11836 [FREE Full text] [CrossRef] [Medline]
    24. Yardley L, Spring BJ, Riper H, Morrison LG, Crane DH, Curtis K, et al. Understanding and promoting effective engagement with digital behavior change interventions. Am J Prev Med 2016 Nov;51(5):833-842. [CrossRef] [Medline]
    25. Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013 Aug;46(1):81-95. [CrossRef] [Medline]
    26. Soltani H, Arden MA, Duxbury AM, Fair FJ. An analysis of behaviour change techniques used in a sample of gestational weight management trials. J Pregnancy 2016;2016:1085916 [FREE Full text] [CrossRef] [Medline]
    27. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009 Jul 21;6(7):e1000097 [FREE Full text] [CrossRef] [Medline]
    28. RoB 2: A Revised Cochrane Risk-of-Bias Tool for Randomized Trials. Cochrane Methods.   URL: https://methods.cochrane.org/bias/resources/rob-2-revised-cochrane-risk-bias-tool-randomized-trials [accessed 2019-08-14]
    29. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Training. 2019.   URL: https://training.cochrane.org/handbook#how-to-cite [accessed 2020-01-10]
    30. Cochrane RevMan. Cochrane Training.   URL: https://community.cochrane.org/help/tools-and-software/revman-5 [accessed 2019-08-14]
    31. Evans WD, Wallace JL, Snider J. Pilot evaluation of the text4baby mobile health program. BMC Public Health 2012 Nov 26;12:1031 [FREE Full text] [CrossRef] [Medline]
    32. Pollak K, Alexander S, Bennett G, Lyna P, Coffman C, Bilheimer A, et al. Weight-related SMS texts promoting appropriate pregnancy weight gain: a pilot study. Patient Educ Couns 2014 Nov;97(2):256-260 [FREE Full text] [CrossRef] [Medline]
    33. Choi J, Lee JH, Vittinghoff E, Fukuoka Y. mHealth physical activity intervention: a randomized pilot study in physically inactive pregnant women. Matern Child Health J 2016 May;20(5):1091-1101 [FREE Full text] [CrossRef] [Medline]
    34. Willcox J, Wilkinson S, Lappas M, Ball K, Crawford D, McCarthy E, et al. A mobile health intervention promoting healthy gestational weight gain for women entering pregnancy at a high body mass index: the txt4two pilot randomised controlled trial. BJOG 2017 Oct;124(11):1718-1728 [FREE Full text] [CrossRef] [Medline]
    35. Redman LM, Gilmore LA, Breaux J, Thomas DM, Elkind-Hirsch K, Stewart T, et al. Effectiveness of SmartMoms, a novel ehealth intervention for management of gestational weight gain: randomized controlled pilot trial. JMIR Mhealth Uhealth 2017 Sep 13;5(9):e133 [FREE Full text] [CrossRef] [Medline]
    36. Hayman M, Reaburn P, Browne M, Vandelanotte C, Alley S, Short CM. Feasibility, acceptability and efficacy of a web-based computer-tailored physical activity intervention for pregnant women - the Fit4Two randomised controlled trial. BMC Pregnancy Childbirth 2017 Mar 23;17(1):96 [FREE Full text] [CrossRef] [Medline]
    37. Dahl A. Healthy Motivations for Moms-To-Be (Healthy MoM2B) Study: A Mobile Health Intervention Targeting Gestational Weight Gain among US Women. Semantic Scholar. 2018.   URL: https:/​/www.​semanticscholar.org/​paper/​Healthy-Motivations-for-Moms-To-Be-(Healthy-MoM2B)-Dahl/​828d334dde24173f90099c03457c0674e0050740 [accessed 2020-06-16]
    38. Evans W, Nielsen P, Szekely D, Bihm J, Murray E, Snider J, et al. Dose-response effects of the text4baby mobile health program: randomized controlled trial. JMIR Mhealth Uhealth 2015 Jan 28;3(1):e12 [FREE Full text] [CrossRef] [Medline]
    39. Smith K, Lanningham-Foster L, Welch A, Campbell CK. Web-based behavioral intervention increases maternal exercise but does not prevent excessive gestational weight gain in previously sedentary women. J Phys Act Health 2016 Jun;13(6):587-593. [CrossRef] [Medline]
    40. Huberty J, Buman M, Leiferman J, Bushar J, Hekler E, Adams MJ. Dose and timing of text messages for increasing physical activity among pregnant women: a randomized controlled trial. Transl Behav Med 2017 Jun;7(2):212-223 [FREE Full text] [CrossRef] [Medline]
    41. Olson C, Groth S, Graham M, Reschke J, Strawderman M, Fernandez I. The effectiveness of an online intervention in preventing excessive gestational weight gain: the e-moms ROC randomized controlled trial. BMC Pregnancy Childbirth 2018 May 9;18(1):148 [FREE Full text] [CrossRef] [Medline]
    42. Bandura A. Social cognitive theory: an agentic perspective. Annu Rev Psychol 2001;52:1-26. [CrossRef] [Medline]
    43. Chan C, Yeung EA, Law B. Effectiveness of physical activity interventions on pregnancy-related outcomes among pregnant women: a systematic review. Int J Environ Res Public Health 2019 May 23;16(10):- [FREE Full text] [CrossRef] [Medline]
    44. Hill B, Skouteris H, Fuller-Tyszkiewicz M. Interventions designed to limit gestational weight gain: a systematic review of theory and meta-analysis of intervention components. Obes Rev 2013 Jun;14(6):435-450. [CrossRef] [Medline]
    45. Gardner B, Croker H, Barr S, Briley A, Poston L, Wardle J, UPBEAT Trial. Psychological predictors of dietary intentions in pregnancy. J Hum Nutr Diet 2012 Aug;25(4):345-353. [CrossRef] [Medline]
    46. Webb TL, Joseph J, Yardley L, Michie S. Using the internet to promote health behavior change: a systematic review and meta-analysis of the impact of theoretical basis, use of behavior change techniques, and mode of delivery on efficacy. J Med Internet Res 2010 Feb 17;12(1):e4 [FREE Full text] [CrossRef] [Medline]
    47. Michie S, Abraham C, Whittington C, McAteer J, Gupta S. Effective techniques in healthy eating and physical activity interventions: a meta-regression. Health Psychol 2009 Nov;28(6):690-701. [CrossRef] [Medline]
    48. Katzmarzyk P, Barlow S, Bouchard P, Catalano D, Hsia DS, Inge TH, et al. An evolving scientific basis for the prevention and treatment of pediatric obesity. Int J Obes (Lond) 2014 Jul;38(7):887-905 [FREE Full text] [CrossRef] [Medline]
    49. Brown MJ, Sinclair M, Liddle D, Hill AJ, Madden E, Stockdale J. A systematic review investigating healthy lifestyle interventions incorporating goal setting strategies for preventing excess gestational weight gain. PLoS One 2012;7(7):e39503 [FREE Full text] [CrossRef] [Medline]
    50. Morrison LG. Theory-based strategies for enhancing the impact and usage of digital health behaviour change interventions: a review. Digit Health 2015;1:2055207615595335 [FREE Full text] [CrossRef] [Medline]
    51. Perski O, Blandford A, Garnett C, Crane D, West R, Michie S. A self-report measure of engagement with digital behavior change interventions (DBCIs): development and psychometric evaluation of the DBCI engagement scale. Transl Behav Med Internet 2019:30. [CrossRef]
    52. Perski O, Blandford A, West R, Michie S. Conceptualising engagement with digital behaviour change interventions: a systematic review using principles from critical interpretive synthesis. Transl Behav Med 2017 Jun;7(2):254-267 [FREE Full text] [CrossRef] [Medline]
    53. Taki S, Lymer S, Russell CG, Campbell K, Laws R, Ong K, et al. Assessing user engagement of an mhealth intervention: development and implementation of the growing healthy app engagement index. JMIR Mhealth Uhealth 2017 Jun 29;5(6):e89 [FREE Full text] [CrossRef] [Medline]
    54. Graham ML, Strawderman MS, Demment M, Olson CM. Does usage of an ehealth intervention reduce the risk of excessive gestational weight gain? Secondary analysis from a randomized controlled trial. J Med Internet Res 2017 Jan 9;19(1):e6 [FREE Full text] [CrossRef] [Medline]
    55. Olson CM, Strawderman MS, Graham ML. Use of an online diet goal-setting tool: relationships with gestational weight gain. J Nutr Educ Behav 2019 Apr;51(4):391-399. [CrossRef] [Medline]
    56. Alkhaldi G, Hamilton FL, Lau R, Webster R, Michie S, Murray E. The effectiveness of prompts to promote engagement with digital interventions: a systematic review. J Med Internet Res 2016 Jan 8;18(1):e6 [FREE Full text] [CrossRef] [Medline]
    57. Bidargaddi N, Pituch T, Maaieh H, Short C, Strecher V. Predicting which type of push notification content motivates users to engage in a self-monitoring app. Prev Med Rep 2018 Sep;11:267-273 [FREE Full text] [CrossRef] [Medline]
    58. Muench F, Baumel A. More than a text message: dismantling digital triggers to curate behavior change in patient-centered health interventions. J Med Internet Res 2017 May 26;19(5):e147 [FREE Full text] [CrossRef] [Medline]
    59. Hamari J, Koivisto J, Sarsa H. Does Gamification Work? - A Literature Review of Empirical Studies on Gamification. In: Proceedings of the 47th Hawaii International Conference on System Sciences. 2014 Presented at: HICSS'14; January 6-9, 2014; Waikoloa, HI, USA. [CrossRef]
    60. Cugelman B. Gamification: what it is and why it matters to digital health behavior change developers. JMIR Serious Games 2013 Dec 12;1(1):e3 [FREE Full text] [CrossRef] [Medline]
    61. Sardi L, Idri A, Fernández-Alemán JL. A systematic review of gamification in e-health. J Biomed Inform 2017 Jul;71:31-48 [FREE Full text] [CrossRef] [Medline]
    62. Dalton JA, Rodger D, Wilmore M, Humphreys S, Skuse A, Roberts CT, et al. The health-e babies app for antenatal education: feasibility for socially disadvantaged women. PLoS One 2018;13(5):e0194337 [FREE Full text] [CrossRef] [Medline]
    63. Smartphone Ownership Penetration in the United Kingdom (UK) in 2012-2020, by Age. Statista.   URL: https://www.statista.com/statistics/271851/smartphone-owners-in-the-united-kingdom-uk-by-age/ [accessed 2019-07-12]
    64. Hughson JP, Daly JO, Woodward-Kron R, Hajek J, Story D. The rise of pregnancy apps and the implications for culturally and linguistically diverse women: narrative review. JMIR Mhealth Uhealth 2018 Nov 16;6(11):e189 [FREE Full text] [CrossRef] [Medline]
    65. Iribarren SJ, Cato K, Falzon L, Stone PW. What is the economic evidence for mhealth? A systematic review of economic evaluations of mhealth solutions. PLoS One 2017;12(2):e0170581 [FREE Full text] [CrossRef] [Medline]
    66. Impact and Evaluation of Baby Buddy App. Best Beginnings.   URL: https://www.bestbeginnings.org.uk/evidence-impact-and-evaluation [accessed 2019-07-12]
    67. de Vivo M, Mills H. 'They turn to you first for everything': insights into midwives' perspectives of providing physical activity advice and guidance to pregnant women. BMC Pregnancy Childbirth 2019 Dec 3;19(1):462 [FREE Full text] [CrossRef] [Medline]

    ‎
    BCT: behavior change technique
    GWG: gestational weight gain
    IOM: Institute of Medicine
    ITT: intention-to-treat
    OR: odds ratio
    PP: per-protocol
    PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analysis
    RCT: randomized controlled trial

    Edited by G Eysenbach; submitted 15.02.20; peer-reviewed by A Merkx, M Bestek, L König; comments to author 23.04.20; revised version received 30.04.20; accepted 14.05.20; published 10.07.20

    Copyright

    ©Alexandra Rhodes, Andrea D. Smith, Paul Chadwick, Helen Croker, Clare H. Llewellyn. Originally published in JMIR mHealth and uHealth (http://mhealth.jmir.org), 10.07.2020.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mHealth and uHealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.