Of the included studies, seven were conducted in the United States of America [33-39], two were from Spain [40,41], and one study was from the United Kingdom [42]. The number of participants within each of the studies ranged from 18 to 104, with a total sample size of 525 participants. The dropout/loss to follow-up rate ranged from 2.0% to 39.1% in the intervention groups and 0% to 25.0% in the control groups. The intervention group participants were aged between 24 and 36 years, whereas the participants from the control groups were aged between 24 and 35 years. Several of the included studies [33,35,36,39,40,41] provided measures of prepregnancy body mass index (BMI) with values ranging from 26 to 30 kg/m² for the intervention groups and 25 to 30 kg/m² for the control groups. Participant ethnicity varied between the studies and study arms (white: intervention: 12.5%-100%; control: 13.3%-100%) [33-36,38-42]. Several studies reported on the level of education within their population ranging from 78% of the total sample having a secondary degree [36], to other studies reporting approximately 20% or above in the usual care and 21% or above in the intervention having a postsecondary education [34,37,38,40,41]. Additional details on the characteristics of each of the included studies can be found in Multimedia Appendix 2.

Six studies conducted the intervention during pregnancy [35-37,40-42], whereas four studies focused on the postpartum period [33,34,38,39]. Of the studies that provided interventions during pregnancy, several used common eHealth technological elements such as text messaging or website support. In particular, Pollak et al [35] used a text-based intervention targeting four health behavior goals during pregnancy, including: (1) targeted daily walking to 10,000 steps, (2) avoid sweetened drinks, (3) eat at least 5 fruits and vegetables each day, and (4) eliminate fast food intake. Only the first two goals were implemented during the initial stages of the intervention (approximately 10 weeks) with all four goals utilized for the rest of trial (approximately 6 weeks). Participants received targeted text messages each week with regard to their current goals and monthly text-message reminders on the Institute of Medicine’s (IOM) GWG guidelines. Carral et al [41] used a website specifically designed for monitoring people with diabetes during pregnancy that allowed for remote and bidirectional communication between health care professionals and patients, including relaying of messages and alerts for glucose monitoring. Herring et al [37] used text messaging, along with social media support groups and coaching to support women through nutritional and physical activity goals. The text messages were daily in frequency and personalized to each goal, building on skills and self-efficacy. The social media group was a forum to support and add further behavioral skills training. Perez-Ferre et al [40] used a telemedicine website and mobile phone app to support the transmission of glucose levels and for sending text messages. The website was used to monitor, adjust, and recommend insulin doses and goals. Smith et al [36] used a website that intervention arm participants would log on to review exercise and nutrition information. Specifically, this included recommendations, goal setting, problem-solving modules, a journal, a calendar, and a community forum for women to interact with other intervention arm participants. Finally, Soltani et al [42] used text messaging and self-monitoring diaries to support women through behavior modification for weight management, physical activity, and nutrition.

Of the studies that provided postpartum interventions, several eHealth strategies were used, including websites, biosensors/activity monitors (ie, pedometers), and text messaging. Colleran et al [33] utilized a Web-based dietary intervention to reduce dietary intake by 500 kcal/day below calculated energy requirements and compared results with recommendations provided on a weekly basis, along with providing strategies to assist women in meeting their outlined recommendations. Kim et al [34] employed a structured Web-based physical activity intervention in which participants received a pedometer and access to a Web-based curriculum. Participants were also provided with personalized step count goals, strategies for meeting these goals, as well as the opportunity to anonymously interact with other intervention group participants through a Web-based study-specific forum. Nicklas et al [38] modified the diabetes prevention program (DPP) to 12 core modules that provided women with the opportunity to track goals (ie, walking and weight), to share secure messages with health care professionals, and to view Web-based media files. Finally, Herring et al [39] piloted a Web-based and text messaging intervention that focused on six empirically tested weight-related behavior change strategies and monitored women via text messaging.

Among studies, the intervention content was provided at varying frequencies, including: daily [34,37], 3 times per week [35], weekly [33,38,40,42], every 2 weeks [39,41], or on an individualized basis [36]. All of the eHealth technologies employed a bidirectional communication modality with asynchronous or interactive communication between the health care professionals and participants. The duration of the interventions in the pregnancy interventions ranged from 6 to 26 weeks, whereas the postpartum interventions ranged from 23 to 52 weeks. All comparator or control groups received usual standard of care or a simplified educational version of the technology offered to the intervention group, which provided only general health information.

All 10 studies reported on participant weight management in terms of weight gain, loss, or maintenance. During pregnancy, six studies [35-37,40-42] that evaluated eHealth technology for weight management found a nonsignificant reduction in GWG, with a mean difference of −1.62 kg (95% CI −3.57 to 0.33) after exposure to the intervention (Figure 2). Four studies contributed to the pooled analysis for the postpartum eHealth technology weight intervention, showing a statistically significant difference in weight loss, with a mean difference of −2.55 kg (95% CI −3.81 to −1.28) after completing eHealth weight management interventions (Figure 2) [33,34,38,39]. The overall pooled analysis for any eHealth technology intervention in the combined population of interest resulted in a statistically significant reduction in weight, with a mean difference of −2.1 kg (95% CI −3.35 to −0.85; Figure 2). When examining the percentage of women gaining weight above recommendations, two studies [36,37] provided a nonsignificant OR of 0.76 (95% CI 0.13 to −4.59; Figure 3). However, the change in BMI in the pooled postpartum studies [33,34,38] showed a significant reduction with a mean difference of −0.87 kg/m² (95% CI −1.56 to 0.18; Figure 4).

Three studies provided data for changes in glycemic parameters, two studies during pregnancy [40,41], and one study postpartum [34]. The pooled change in glycemic parameters during pregnancy as measured by glycated hemoglobin (HbA_1C) was an increase of 0.10 (95% CI −0.08 to 0.28; Figure 5). One study [34] that reported on glycemic parameters (fasting, 2 hour glucose, log fasting insulin) found that technology raised fasting glucose nonsignificantly by 0.09 mmol/L (95% CI −0.27 to 0.45) and 2-hour postprandial glucose by 0.06 mmol/L (95% CI −0.98 to 1.10). Finally, log fasting insulin decreased by −0.20 (95% CI −0.44 to 0.04; Figure 6). All glycemic changes were not statistically significant.

In addition, one study reported on nutrition status during pregnancy [36]. The study found that after exposure to a Web-based program, women reported a nonsignificant reduction in energy intake from carbohydrates (1.10%, 95% CI −4.24 to 2.04) and from fat (−0.90%, 95% CI −3.37 to 1.57), as well as a nonsignificant increase in energy intake from protein (1.40%, 95% CI 0.11-2.69; Figure 7).

Another study reported on nutrition status changes postpartum after exposure to an eHealth technology intervention for 4 months [33]. This study found a statistically significant reduction in total daily energy intake of 442.0 kcal (95% CI −803.10 to −80.90). The same study found statistically significant reductions (Figure 8) in percentage of total daily intake of fat and added sugars by −4.90% (95% CI −7.84 to −1.96) and −5.70% (95% CI −8.66 to −2.74), respectively. Changes in the percentage of energy intake from carbohydrate significantly increased by 4.60% (95% CI 1.69-7.51), and the percentage of energy intake from protein decreased by −0.80% (95% CI −0.89 to 2.49), although this small change was not statistically significant (Figure 8). When examining daily servings of food groups, statistically significant reductions in the number of daily milk servings (−1.20, 95% CI −1.80 to −0.56) and daily servings of whole grains (−1.20, 95% CI −2.31 to −0.09) were noted [33]. However, daily servings of fruit, vegetables, oils or fat, and sweets were not significantly impacted by the 4-month exposure to the eHealth technology.

Finally, the pregnancy-specific study utilized the self-reported pregnancy physical activity questionnaire (PPAQ) for estimating physical activity levels during pregnancy [35]. No significant differences were found between the two groups in light or moderate physical activity as determined by the PPAQ at 32 weeks of gestation (moderate: 95% CI −3.5 to −0.3, P=.71; light: 95% CI −2.6 to 0.4, P=.08). A postpartum study that employed a Web-based self-report survey on physical activity habits [34] found no significant differences at baseline and follow-up with regard to the proportion of individuals within three activity categories (0 min/week, <60 min/week, and ≥60 min/week) for physical activity levels between the control and intervention groups, including any activity (baseline: P=.61; follow-up: P=.25), mild (baseline: P=.26; follow-up: P=.20), moderate (baseline: P=.81; follow-up: P=.51), and vigorous physical activity (baseline: P=.81; follow-up: P=.65). As a result of the heterogeneity of the measurement protocols between the pregnancy and postpartum studies, a meta-analysis was not conducted for the physical activity parameters.

This review summarizes the most relevant/applicable trial evidence available to assess the effectiveness of eHealth technologies on weight management in pregnant or postpartum women. Notably, all of the included studies were recent (published between 2010 and 2016), which highlights the emerging use of this technology as a novel health care strategy, yet the low number of studies emphasizes the need for further evidence to support its efficacy for weight management in the clinical environment. To that end, the analysis of evidence within this review demonstrated that exposure to eHealth technology was associated with a nonsignificant benefit for weight management during pregnancy and a statistically and clinically significant weight reduction in the postpartum period. The pooled estimate for change in weight (kg) during pregnancy was −1.62 kg (95% CI −3.57 to 0.33), whereas the pooled estimate for change in weight postpartum was −2.55 kg (95% CI −3.81 to −1.28).

The eHealth interventions that were effective in minimizing excessive weight gain (kg) during pregnancy comprised multiple components, including individualized text messaging and the use of social media [35,37]. Specific components that were effective in one study focused on a multimodal approach to eHealth, utilizing individualized text messaging for skills training and self-monitoring, private social media chat group for support, and individualized health coaching telephone calls [37]. Another study that was effective in reducing weight in pregnancy employed frequent text messaging (3 times per week), which focused on nutrition and physical activity by providing concise suggestions for modifying nutritional behavior (ie, avoid sweetened drinks) and increasing physical activity (ie, goal of 10,000 steps/day) [35].

Three studies demonstrated effectiveness in reducing postpartum weight with eHealth interventions. One study that included an eHealth solution (Web-based MyPyramid Menu Planner) with additional in-person counseling/support [33] had a greater reduction in weight and BMI as compared with the intervention that included only eHealth components (Web-based information, online forum, text messaging, and email) [34] implemented during the same period. Furthermore, another study found significant weight reduction when the eHealth intervention focused on both nutrition and physical activity (Healthy4Baby) [39]. Finally, one study modified the lifestyle-intensive DPP for postpartum women and also achieved a significant reduction in weight and BMI [38]. Overall, the multifaceted interventions (ie, targeting both physical activity and nutrition) with multiple and different modalities may be more effective than an eHealth-focused intervention targeting physical activity alone. More importantly, none of the studies performed an evaluation to separate the effects of personal contact with a health professional from the effects of the eHealth intervention alone. This information could help determine predictors of participant engagement or adherence with the eHealth technology. Further research is needed to determine the independent effects of these technologies on weight management for studies employing multimodal intervention methods.

The growth in eHealth apps is related to the underlying presumption that their use will be associated with lower health care costs and improvements in health outcomes, particularly when focusing on the prevention of diseases and promotion of healthy lifestyles. Although eHealth technologies have the potential to improve prenatal health care delivery by providing frequent, interactive, and personalized information to broad populations in real time, there is a risk that the app may not be effective or could potentially result in harm to the mother and her unborn child. Thus, it is critical that eHealth intervention technologies be designed using an evidence-based approach and tested/evaluated with the addition of appropriate safeguards to ensure safety of the participants before being implemented into widespread use among the general population. This may include the performance of clinical trials that use a data safety and monitoring committee who will intervene in the occurrence of increased adverse events within a study.

Participant engagement is also critical to the success of any eHealth intervention. To date, technology-based weight management approaches have been well accepted in nonpregnant populations [7] and postpartum populations with up to 74% of postpartum women accessing and reviewing weight management materials immediately after receiving the resources in one study [12]. Studies have reported significant variability in the number of intervention participants that read and respond to study-based text messages, similar to the postpartum participants receiving eHealth interventions. Although the findings from this review suggest that multicomponent interventions (ie, combined focus on both nutrition and physical activity) resulted in more favorable weight management during pregnancy and postpartum, it is difficult to ascertain which component attributed to the observed effect or whether it is related to the entire “bundle” of interventions. Moreover, not all eHealth components are considered as useful or desired by participants within a weight management intervention. For example, only 14% of postpartum women utilized an online forum for interacting with other participants for peer support [34]. Consequently, before implementing an intervention of this type, investigators must carefully consider the design and features of the eHealth intervention for their target population. This includes ensuring the use of both effective and appropriate strategies and frameworks to provide reasonable engagement and adherence both in the short-term with long-term follow-up to determine whether these behaviors that are the targets of such interventions have lasting effects.

Recognizing the importance and value of patient engagement, the Canadian Institutes of Health Research (CIHR) examined a strategy for patient-oriented research (SPOR) where patients, researchers, health care providers, and decision makers work together to build a sustainable, accessible, and equitable health care system [43]. Applying this principle of patient engagement to the development of new eHealth apps is beneficial and necessary [44,45]. Including pregnant and postpartum women, whether during the initial app developmental process or through preliminary focus group trials, would provide tangible feedback during this critical period in areas such as GWG, physical activity, sleep, and nutrition. Ensuring that embedded tools and features are clear and easily accessible for various levels of literacy and digital experience is also a requirement [46]. For eHealth apps to gain traction and thus reach the widest audiences, endorsement from clinical stakeholders and health care providers will likely also be necessary [47].