Effects of e-Health interventions on psychological outcomes among high-risk pregnant women: a systematic review and meta-analysis

Abstract

Background

High-risk pregnant women face not only physical health issues but also substantial psychological challenges. With the growing interest in innovative strategies, e-Health interventions have gained recognition as a supportive approach for high-risk pregnant women. This systematic review and meta-analysis examined the effects of e-Health interventions on psychological outcomes in high-risk pregnant women.

Methods

Interventional studies were identified through searches of PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase, Cochrane Central, and Scopus databases. A total of 15,996 articles were retrieved, and 42 interventional studies published between 2007 and 2024 were included in this systematic review and meta-analysis.

Results

Among 42 studies, 24 (57.2%) were published after 2020, and 32 (76.2%) were randomised controlled trials. The e-Health interventions focused on psychological support (47.6%), physical health management (42.9%), and parenting role education (9.5%). The meta-analysis indicated that e-Health interventions were associated with reductions in anxiety (p = 0.017), depression (p = 0.048), and stress (p = 0.006) among high-risk pregnant women. The heterogeneity was high for the overall anxiety and depression analyses (I² > 80%), and these pooled estimates should be interpreted with caution.

Conclusions

Given their accessibility, e-Health interventions may offer potential benefits for the psychological support of high-risk pregnant women; however, these findings should be interpreted cautiously owing to the high heterogeneity in the anxiety and depression estimates. Further research is required to confirm these effects and to evaluate their optimal design, long-term effects, and cost-effectiveness.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12884-026-09767-x.

Background

High-risk pregnancy is associated with increased physical and psychological challenges for women, significantly affecting both maternal and foetal health [1]. Pregnant women classified as having high-risk pregnancies, due to either physical conditions (e.g., gestational diabetes mellitus [GDM], pre-eclampsia, preterm labour risk) or psychological conditions (e.g., antenatal depression or anxiety disorders), often experience significant levels of anxiety, depression, and stress [2]. Such psychological distress during pregnancy has been linked to adverse perinatal outcomes, including preterm birth, low birth weight, and developmental complications in infants [3, 4]. Additionally, maternal mental health disorders can extend into the postnatal period, affecting mother-infant bonding and further increasing the risk of postnatal depression [5]. Despite the critical need for psychological support, many high-risk pregnant women face barriers in accessing traditional in-person interventions due to time restrictions, geographical limitations, financial constraints, and a shortage of specialised healthcare providers [6].

With advancements in digital healthcare, e-Health interventions—which utilise digital platforms such as mobile applications, websites, telehealth services, and wearable devices—have gained increasing attention as potential solutions for addressing these challenges [7, 8]. E-Health interventions offer several advantages, including convenience, accessibility, safety, cost-effectiveness, and the ability to provide continuous engagement tailored to individual needs [9, 10]. Previous research suggests that technology-based interventions can be effective in managing maternal mental health conditions, improving parental self-efficacy, and enhancing overall well-being [11, 12]. However, although the use of e-Health approaches in maternal healthcare is growing, there is limited evidence that specifically evaluates their effects in high-risk pregnant populations. Furthermore, variations in intervention designs, delivery methods, and outcome measures make it difficult to establish definitive conclusions about their effects.

Given these gaps, this systematic review and meta-analysis aimed to comprehensively evaluate the effects of e-Health interventions on psychological outcomes in high-risk pregnant women. By synthesising findings from interventional studies, this review provides insights into the potential of e-Health interventions as supportive strategies for maternal psychological health. Understanding the impact of e-Health interventions on psychological well-being can inform clinical practice and policy decisions, ultimately enhancing the quality of care for high-risk pregnant women.

Methods

Search strategy

This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines provided by the PRISMA group [13]. Five databases were searched: PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase, Cochrane Central, and Scopus. A list of the retrieved documents was compiled, and duplicates were removed. The bibliographic management software EndNote 21 was used to manage the bibliographic information of all documents uniformly.

Eligibility criteria

The search terms were based on the research population (P) of ‘high-risk pregnancy’ AND the intervention variables (I) of ‘e-Health’, ‘m-Health’, ‘mobile health’, ‘telehealth’, ‘telemedicine’, ‘telenursing’, ‘internet-based intervention’, ‘web-based intervention’, ‘online intervention’, ‘application’, ‘app’, ‘social network service’, or ‘SNS’, including various synonyms to compose the search terms. The following selection criteria were applied for the review: (1) studies involving women with high-risk pregnancies due to either physical conditions (e.g., GDM, pre-eclampsia, preterm labour, anaemia, obesity) or psychological conditions (e.g., anxiety, depression, stress); (2) studies employing e-Health strategies, including e-Health, m-Health, mobile health, telehealth interventions, as well as web-based platforms, applications, or SNS; (3) studies that assessed various psychological outcomes among participants such as anxiety, depression, stress, mindfulness, affect, and other related measures; (4) study designs comprising both randomised controlled trials (RCTs) and non-randomised controlled trials (non-RCTs); and (5) studies published in English. As high-risk pregnancy in this review encompassed both physical and psychological conditions, the included population represented a broad range of clinical profiles, which may introduce clinical heterogeneity across the studies.

Study selection

The PRISMA flowchart was utilised to detail the step-by-step process of literature selection based on the study aim and selection criteria (Fig. 1). A total of 15,996 articles were retrieved from PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase, Cochrane Central, and Scopus databases. After initially removing 5,835 duplicate articles, 10,161 articles were reviewed by title. Subsequently, the abstracts of 645 articles were further reviewed, and 418 articles that did not meet the inclusion criteria were removed. A full-text review was conducted on 227 articles, and 185 inappropriate articles were excluded, resulting in a final selection of 42 articles. Two researchers independently evaluated the literature search and selection, and compared the results. Any discrepancies were resolved through re-evaluation of the original articles and collaborative discussions among all contributing authors.

Fig. 1.

PRISMA flow chart of the identification and selection of studies

Quality appraisal

Two authors independently evaluated the methodological quality of the selected articles using the Joanna Briggs Institute’s (JBI) Critical Appraisal Checklist for quasi-experimental and randomised experimental studies. The evaluation tool for quasi-experimental studies comprised 9 items (total score 0–9 points), and the tool for randomised experimental research comprised 13 items (total score 0–13 points). Each item was evaluated using four options (‘Yes’, ‘No’, ‘Unclear’, or ‘Not applicable’), with ‘Yes’ scored as 1 point and all other options scored as 0 points. The total number of ‘Yes’ responses thus determined the total score. Two authors engaged in discussions and reached a consensus on the methodological quality of each study. All 42 studies were included in the review after critical appraisal (Additional file 1).

Data synthesis

In order to synthesise the findings of the selected studies, we created a data extraction form to capture the key components of each study. Using a pre-defined data extraction form, two review authors independently extracted the following data from the included studies: authors, publication year, country, study aim, research design, theoretical framework, population, sample size, intervention method, intervention contents, outcome variables, follow-ups, main findings, and funding sources. We performed a descriptive synthesis of the findings of selected studies to better understand the key elements of the interventional studies.

Further, we performed meta-analyses to determine the intervention effects across studies that implemented comparable interventions, measured similar outcome variables, and reported mean and standard deviation values, restricting RCTs only to ensure methodological consistency. We conducted meta-analyses based on the most commonly measured outcomes of anxiety, depression, and stress scales, as well as subgroup analyses according to the intervention categories. We utilised RevMan 5.4 software for meta-analyses, and used point estimation (p < 0.05) and interval estimation (95% confidence intervals [CI]) for statistical significance. We obtained the standardised mean differences (SMDs) in the outcomes. All the authors engaged in individual and collaborative efforts to reach a consensus on the synthesis of the selected studies. Additionally, a funnel plot was created as a visual aid to detect reporting bias. Although the plot appeared reasonably symmetrical, publication bias cannot be ruled out due to the limited number of studies included in the meta-analysis.

Results

General characteristics and research methodology of the included studies

Forty-two interventional studies [14–55] were included in the systematic review. The general characteristics of the included studies are summarised in Table 1.

Table 1.

General characteristics of included studies

Variables	Categories	n (%)
Publication year	2007–2010	4 (9.5)
	2011–2020	14 (33.3)
	2021–2024	24 (57.2)
Country	USA	9 (21.4)
	China	7 (16.7)
	Iran	5 (11.9)
	South Korea	4 (9.5)
	Netherlands	3 (7.1)
	Australia	2 (4.8)
	Sweden	2 (4.8)
	Switzerland	2 (4.8)
	UK	2 (4.8)
	Others	6 (14.3)
Theoretical framework	Yes	7 (16.7)
	No	35 (83.3)
Funding source	Yes	32 (76.2)
	No	1 (2.4)
	Not reported	9 (21.4)
Study design	Randomised controlled trial	32 (76.2)
	Non- randomised controlled trial	10 (23.8)
Sample size	≤ 60	11 (26.2)
	60 <	31 (73.8)
Mean age of women (year)	< 35	31 (73.8)
	35 ≤	2 (4.8)
	Not reported	9 (21.4)
High-risk pregnancy diagnosis
Physical conditions	Gestational diabetes mellitus	13 (30.9)
	Preterm labour	5 (11.9)
	Pre-eclampsia	4 (9.5)
	Anaemia	1 (2.4)
	Obesity	1 (2.4)
	Various physical conditions	1 (2.4)
Psychological conditions	Depression or anxiety	10 (23.8)
	Stress	1 (2.4)
Combined conditions		6 (14.3)

The earliest study included in this review was published in 2007 [55], and a total of 24 studies (57.2%) were published after 2020 [14–37]. The included studies originated from a diverse range of countries. The majority were conducted in the USA (n = 9, 21.4%) and China (n = 7, 16.7%), followed by Iran (n = 5, 11.9%), South Korea (n = 4, 9.5%), and the Netherlands (n = 3, 7.1%). Additionally, two studies were conducted in each of Australia, Sweden, Switzerland, and the UK, while one study was reported from each of Italy, Norway, Saudi Arabia, Taiwan, Thailand, and Turkey.

A total of 35 studies (83.3%) did not use a theoretical framework, with only 7 studies explicitly applying one [15, 19, 22, 28, 41, 51, 52]. Thirty-two studies (76.2%) disclosed their funding sources, one reported no funding, and 9 studies (21.4%) did not provide this information. Regarding study design, 32 studies (76.2%) were RCTs [14, 17, 20–27, 29–42, 45–50, 53, 55], and 10 studies (23.8%) were non-RCTs [15, 16, 18, 19, 28, 43, 44, 51, 52, 54].

In terms of sample size, 31 studies (73.8%) enrolled more than 60 participants [14, 16, 21–26, 29, 31, 32, 34–43, 45, 46, 48–55]. The mean age of participants was reported as 34 years or younger in 31 studies (73.8%) [14, 16–21, 24–26, 28, 29, 31–41, 43, 45–48, 50, 54, 55]; and 35 years or older in 2 studies (4.8%) [44, 51]. High-risk pregnancy diagnoses included physical conditions such as GDM (n = 13, 30.9%) [15, 19, 26, 27, 35, 36, 42–44, 49, 52, 54, 55], preterm labour (n = 5, 11.9%) [23, 28, 48, 50, 53], and pre-eclampsia (n = 4, 9.5%) [14, 17, 20, 34], as well as psychological conditions such as depression or anxiety (n = 10, 23.8%) [16, 24, 33, 37, 39–41, 45–47]. Additionally, 6 studies included combined conditions involving both physical and psychological factors [18, 29, 30, 32, 38, 51].

Table 2 presents the research methodologies used in the included studies. E-Health-only interventions (n = 35, 83.3%) [14–18, 20–30, 32–35, 37–40, 42–48, 50, 51, 53, 54] were more commonly reported than hybrid interventions (n = 7, 16.7%) [19, 31, 36, 41, 49, 52, 55], which combined e-Health and offline methods. Delivery formats for e-Health interventions included websites (n = 26, 61.9%) [15, 16, 18, 21, 23, 24, 28, 29, 32–36, 38–41, 44, 45, 47–51, 54, 55], mobile applications (n = 22, 52.4%) [14, 15, 17–19, 22, 25–28, 30–32, 34, 36–38, 42–44, 46, 53], phone or text messages (n = 17, 40.5%) [16, 18, 20–23, 26–29, 33, 40, 44, 45, 51, 52, 54], email (n = 10, 23.8%) [23, 24, 28, 29, 32, 38, 40, 45, 47, 48], and wearable devices (n = 2, 4.8%) [14, 22], with multiple formats often used in combination.

Table 2.

Research methodology of included studies

Variables	Categories	n (%)
Intervention method	E-Health	35 (83.3)
	Hybrid (e-Health & offline)	7 (16.7)
Delivery format for intervention*	Website	26 (61.9)
	Mobile application	22 (52.4)
	Phone or text message	17 (40.5)
	Email	10 (23.8)
	Wearable device	2 (4.8)
Intervention category	Psychological intervention	20 (47.6)
	Physical health management intervention	18 (42.9)
	Parenting role education	4 (9.5)
Intervention period	< 1 month	6 (14.3)
	1–2 months	18 (42.9)
	2 months <	8 (19.0)
	Not specified	10 (23.8)
Follow-up period*	Antepartum	36 (85.7)
	Postpartum	13 (31.0)

^*Duplicated

The interventions were grouped into three categories according to similar characteristics described in the studies, with categorization determined by the dominant component reported: (1) Psychological interventions (n = 20, 47.6%) [16–18, 20, 24, 25, 28, 30, 32, 37–41, 45–48, 50, 53], (2) Physical health management interventions (n = 18, 42.9%) [14, 19, 21, 22, 26, 27, 31, 34–36, 42–44, 49, 51, 52, 54, 55], and (3) Parenting role education (n = 4, 9.5%) [15, 23, 29, 33]. Psychological interventions included mindfulness-based training, guided imagery, hypnotherapy, stress management programs, self-compassion programs, and mood tracking interventions. Physical health management interventions comprised exercise programs, WeChat-based support, home telemonitoring, healthy weight programs, health-promotive lifestyle interventions, and self-care programs. Parenting role education encompassed childbirth education, breastfeeding support programs, and preparatory education for prospective parents.

The intervention period was less than one month in 6 studies (14.3%) [17, 26, 30, 45, 51, 52], one to two months in 18 studies (42.9%) [15, 16, 18, 20, 24, 25, 28, 32, 35, 37–41, 43, 46, 48, 50], and more than two months in 8 studies (19.0%) [23, 27, 29, 31, 33, 44, 47, 53]. Other studies did not specify the exact period of intervention.

Follow-up measurements were performed during the antenatal period in 36 studies (85.7%) [14, 16–18, 20, 22–31, 33–38, 40, 41, 43–55], and 13 studies measured during the postnatal period (n = 13, 31.0%) [15, 19, 21, 25, 29, 32–34, 37, 39, 40, 42, 54].

Diverse dependent psychological variables were measured during the antepartum period, including anxiety (n = 21, 50.0%) [16–18, 23, 24, 29, 30, 33–38, 40, 44–48, 50, 53], depression (n = 18, 42.9%) [16, 18, 24, 25, 30, 33, 35–38, 40, 41, 44–48, 52], stress (n = 9, 21.4%) [18, 25, 28, 30, 37, 41, 45, 48, 50], sleep dysfunction (n = 5, 11.9%) [14, 20, 37, 38, 47], mindfulness (n = 3, 7.1%) [24, 25, 41], positive affect (n = 3, 7.1%) [25, 37, 41], self-efficacy (n = 3, 7.1%) [22, 51, 55], and so on. In addition, during the postpartum period, psychological variables including depression (n = 9, 21.3%) [21, 25, 29, 32, 33, 37, 39, 40, 54], anxiety (n = 6, 14.3%) [21, 32–34, 37, 40], stress (n = 4, 9.5%) [25, 37, 39, 54], and so on, were measured (Table 3).

Table 3.

Psychological outcome measurements of included studies

Measurements*	n (%)
During antepartum period
Anxiety	21 (50.0)
Depression	18 (42.9)
Stress	9 (21.4)
Sleep dysfunction	5 (11.9)
Mindfulness	3 (7.1)
Positive affect	3 (7.1)
Self-efficacy	3 (7.1)
Self-compassion	2 (4.8)
Somatization	2 (4.8)
Treatment credibility	2 (4.8)
During postpartum period
Depression	9 (21.3)
Anxiety	6 (14.3)
Stress	4 (9.5)
Mindfulness	2 (4.8)
Positive affect	2 (4.8)
Self-compassion	2 (4.8)
Sleep dysfunction	2 (4.8)

^*Duplicated

Anxiety was measured using the Generalized Anxiety Disorder questionnaire-7 [56] in 7 studies, the State-Trait Anxiety Inventory [57] in 6 studies, the Brief Symptom Inventory-18 [58] in 2 studies, and the Self-rating Anxiety Scale [59] in 2 studies. Depression was measured using the Edinburgh Postnatal Depression Scale [60] in 15 studies, the Self-rating Depression Scale [61] in 4 studies, the Patient Health Questionnaire-9 [62] in 3 studies, the Brief Symptom Inventory-18 [58] in 2 studies, and the Center for Epidemiological Studies Depression scale [63] in 2 studies. Stress was measured using the Perceived Stress Scale [64] in 7 studies, and sleep dysfunction was measured using the Pittsburgh Sleep Quality Index [65] in 5 studies (Additional file 2).

Effects of e-Health interventions for improving psychological outcomes

Figures 2, 3 and 4 present the effects of e-Health interventions on psychological aspects- the primary outcome variables of interest in this systematic review- in high-risk pregnant women. For the most commonly measured variables of the anxiety, depression, and stress scales in the antepartum period, we conducted meta-analyses of RCTs that measured comparable dependent variables at similar time points and provided sufficient data for the analyses. Furthermore, subgroup analyses were performed based on the intervention categories.

Fig. 2.

Effects of e-Health interventions on anxiety among high-risk pregnant women

Fig. 3.

Effects of e-Health interventions on depression among high-risk pregnant women

Fig. 4.

Effects of e-Health interventions on stress among high-risk pregnant women

According to the meta-analysis, overall e-Health interventions were associated with a reduction in anxiety among high-risk pregnant women (with physical [36, 50, 53], psychological [24, 33, 37, 40, 47], and combined [29, 30] conditions) (SMD − 1.72, 95% CI [-3.12, -0.31], p = 0.017); however, given the high heterogeneity (I² = 83%), this pooled estimate should be interpreted with caution. In the subgroup analysis, psychological interventions were associated with a more modest reduction in anxiety (with physical [50, 53], psychological [24, 37, 40, 47], and combined [30] conditions), with low heterogeneity (SMD − 0.59, 95% CI [-1.15, -0.02], I² = 2%, p = 0.042) (Fig. 2).

Overall e-Health interventions were associated with a reduction in depression among high-risk pregnant women (with physical [36], psychological [24, 33, 37, 40, 41, 47], and combined [30] conditions) (SMD − 1.78, 95% CI [-3.54, -0.02], p = 0.048); given the high heterogeneity (I² = 85%), this estimate should also be interpreted with caution. In the subgroup analysis, psychological interventions were associated with a reduction in depression among pregnant women (with psychological [24, 37, 40, 41, 47] and combined [30] conditions), with moderate heterogeneity (SMD − 1.46, 95% CI [-2.87, -0.05], I² = 66%, p = 0.042) (Fig. 3).

For stress, the meta-analysis showed a reduction following e-Health interventions, all of which were psychological interventions, among high-risk pregnant women (with physical [50], psychological [37, 41], and combined [30] conditions), with no observed heterogeneity (SMD − 1.21, 95% CI [-2.07, -0.35], I² = 0%, p = 0.006) (Fig. 4).

Furthermore, the results of the meta-analysis indicated that intervention effects on psychological variables assessed during the postpartum period were not statistically significant.

Discussion

This systematic review and meta-analysis examined the effects of e-Health interventions on psychological outcomes in high-risk pregnant women. These findings suggest that e-Health interventions are associated with improvements in anxiety, depression, and stress, indicating their potential as a supportive approach for this population. Among the intervention categories, psychological interventions appeared potentially beneficial for these outcomes. These results highlight the value of integrating digital health solutions into maternal healthcare to support mental well-being in high-risk pregnancies.

The prevalence of e-Health interventions has increased notably, with 24 of the 42 studies included in this review (57.2%) published after 2020. This trend reflects the growing recognition of digital health technologies as accessible and scalable solutions to address perinatal mental health challenges. However, despite their widespread adoption, 35 studies (83.3%) did not employ a theoretical framework, which may limit their effectiveness and generalisability [66]. Future research should incorporate well-established psychological and behavioural models to optimise intervention design and delivery [67]. For example, interventions grounded in the theoretical frameworks of the Social Cognitive Theory of Self-Regulation [68], Stress, Appraisal, and Coping [69], and the Mechanisms of Mindfulness [70], which were employed in some of the included studies in this review, could more systematically target maladaptive thoughts and behaviours that contribute to perinatal psychological distress. Additionally, employing frameworks such as the Health Belief Model [71] or the Theory of Planned Behavior [72] may help in understanding and shaping health-related attitudes, beliefs, and intentions, thereby enhancing engagement and adherence to e-Health interventions. Utilising these models can provide clearer mechanisms of action and support the development of more robust, generalisable e-Health intervention programs for high-risk pregnant populations.

In terms of e-Health intervention delivery formats, websites and mobile applications were the most commonly used modalities, followed by phone or text messaging, email, and wearable devices. The diversity of delivery methods suggests that multimodal approaches may enhance engagement and intervention adherence. However, considerable variations in intervention design, duration, and follow-up assessments across studies pose challenges to determining the most effective strategies [73]. Standardising the key components of e-Health interventions, such as content structure, frequency of engagement, and duration, could help improve comparability across studies and facilitate broader implementation for high-risk pregnant women.

The meta-analysis results suggest that e-Health interventions are associated with improvements in anxiety, depression, and stress among high-risk pregnant women. These findings align with previous research suggesting the benefits of digital health interventions in managing perinatal mental health [74]. Notably, our subgroup meta-analyses indicated that psychological interventions were associated with reductions in anxiety, depression, and stress among pregnant women with physical, psychological, or combined conditions. For example, among pregnant women with physical conditions such as preterm labour, e-Health interventions such as internet-based relaxation training with cognitive restructuring exercises [50] and a supportive clinical decision-making app [53] were associated with reduced anxiety, with the former also showing reductions in stress. Similarly, for pregnant women with psychological conditions such as depression or anxiety, interventions including online mindfulness-based exercises targeting body awareness, emotion regulation, and the observation and acceptance of thoughts [24]; mindfulness meditation and breathing exercises delivered via a smartphone app [37], which was also associated with reduced stress; guided internet-based problem-solving therapy combined with supportive telephone calls to help manage daily worries and challenges [40]; and psychoeducation, cognitive restructuring, and interactive online feedback via email [47] were all associated with improvements in anxiety and depression.

These findings highlight the versatility and potential of e-Health interventions in addressing the psychological distress commonly experienced by pregnant women with diverse high-risk profiles, suggesting that digital approaches may help support their psychological needs. High-risk pregnancies, regardless of underlying clinical conditions, often involve substantial psychological challenges, highlighting the importance of appropriate support systems and interventions [75, 76]; targeted e-Health interventions may therefore contribute to improvements in these outcomes. Future research should explore the integration of multiple therapeutic elements within e-Health interventions and determine the optimal duration and intensity required to sustain these potential benefits from pregnancy through the postpartum period.

However, substantial heterogeneity was observed across the included studies. Differences in intervention type, duration, population characteristics, and outcome measures may explain the high heterogeneity observed. For example, the types of e-Health interventions (e.g., web-based education, mobile applications, telehealth counselling), intervention durations, and the intensity or frequency of engagement could have contributed to the variability in outcomes. Additionally, variations in participant characteristics, such as psychological risk profiles and the underlying clinical conditions defining high-risk pregnancies, may have further influenced interventional effectiveness. Differences in outcome measurement tools and timing of assessments across studies also present important contributors to heterogeneity. These factors strongly suggest the need for future research to adopt more standardised intervention protocols and outcome assessments, as well as to conduct detailed subgroup analyses to better identify which populations and intervention characteristics yield the greatest benefits.

The pooled effect sizes for anxiety and depression were large by conventional standards and warrant cautious interpretation. The large effect sizes (SMD > 1) may reflect methodological variability, small sample sizes, or differences in outcome measurement rather than true clinical magnitude. Consistent with this, the subgroup analysis restricted to psychological interventions, which showed low heterogeneity, yielded a more modest effect on anxiety, suggesting that the larger overall estimates, together with the heterogeneity noted above, may have been inflated. The pooled effect sizes should therefore be regarded as indicating a potential benefit rather than a precise estimate of clinical magnitude.

In addition, the findings for postpartum outcomes were not statistically significant according to this review of e-Health interventions for high-risk pregnant women. Several factors may explain these results. Comparatively fewer studies specifically evaluated postpartum psychological outcomes, and some interventions had shorter follow-up periods that may not have captured long-term changes. Moreover, many interventions initiated during pregnancy did not extend into the postpartum phase, and the timing of postpartum assessments varied widely, limiting data available for meta-analysis. Natural psychological adjustments after childbirth may also have diminished the effects of antenatal interventions. Nonetheless, given that postpartum women typically have fewer in-person clinical visits compared to during pregnancy, the demand for and potential utility of e-Health support may be even greater in this period. Future research should therefore focus on designing interventions that ensure continuity from the antenatal through the postpartum period, while also accounting for the distinct psychological and social dynamics that characterize the postpartum experience.

Another consideration is the limited representation of certain maternal characteristics in the included studies that are well-established risk factors for psychological distress. Specifically, most included studies involved participants younger than 35 years, with only a small proportion reporting a mean maternal age of 35 years or older. Given that advanced maternal age is associated with heightened risks for perinatal psychological challenges [77], it highlights the need for future studies to explore whether maternal age may influence the acceptability and effectiveness of e-Health interventions on the psychological outcomes in this population.

Nevertheless, the potential benefits suggested by this systematic review and meta-analysis highlight the value of e-Health interventions as potentially effective alternatives or complements to traditional in-person therapies, particularly for populations such as high-risk pregnant women who often face barriers in accessing mental health services [78]. It remains unclear which specific e-Health intervention components are most effective for different high-risk subgroups. Future research should further explore the mechanisms through which these interventions may exert their effects, as well as the optimal intervention components and delivery strategies to enhance the psychological benefits for high-risk pregnant women. Such research should also examine the integration of diverse components within e-Health programs, carefully tailored to the unique risk profiles of this population and their contextual realities, such as accessibility to care and individual preferences, to optimise maternal psychological well-being and improve perinatal outcomes.

Limitations

This systematic review and meta-analysis had several limitations. First, substantial heterogeneity was observed across the included studies, with high statistical heterogeneity in the overall anxiety and depression analyses (I² > 80%) and clinical heterogeneity arising from the broad definition of high-risk pregnancy (encompassing both physical and psychological conditions) and variation in intervention design, delivery formats, and outcome measures. This may have contributed to an overestimation of the pooled effect sizes, which should therefore be interpreted with caution. Second, owing to the limited number of studies and inconsistencies in reporting essential variables, neither meta-regression nor sensitivity analyses could be conducted, which may further affect the robustness of the findings. Third, the subgroup analyses were limited by insufficient data, with some subgroups including only a small number of studies; subgroup analyses based on a limited number of studies (n < 5) should therefore be considered exploratory. This may reduce the reliability of the findings and highlight the need for further research. Fourth, the lack of theoretical frameworks in most studies may have affected the robustness and generalisability of the interventions. Fifth, the majority of included studies involved participants under 35 years of age, which restricts the applicability of the findings to older pregnant women who may have different psychological risks and responses to e-Health interventions. Additionally, factors such as digital literacy, Internet access, and user engagement have not been consistently addressed, indicating the need to explore barriers to sustained participation. Lastly, variations in follow-up periods limit our understanding of long-term effects. Despite these limitations, this review emphasises the potential of e-Health interventions for high-risk pregnant women and the need for further research to enhance their effectiveness and implementation.

Conclusions

This systematic review and meta-analysis examined e-Health interventions for psychological well-being among high-risk pregnant women. The findings suggest that these interventions may offer potential benefits for anxiety, depression, and stress; however, these findings should be interpreted cautiously given the high heterogeneity across studies. With their accessibility and scalability, e-Health interventions could complement existing perinatal mental healthcare, although their broader implementation in clinical and community settings should be guided by further evidence. Future research using standardized protocols and outcome measures is needed to confirm these effects and to clarify their long-term impact, cost-efficiency, and strategies to enhance user engagement for high-risk pregnant women.

Abbreviations

CI

Confidence interval

CINAHL

Cumulative Index to Nursing and Allied Health Literature

GDM

Gestational diabetes mellitus

JBI

Joanna Briggs Institute

PRISMA

Preferred Reporting Items for Systematic Review and Meta-Analysis

Authors’ contributions

JES conceptualized the study and received the research fund. GRK and JHY identified eligible papers, extracted data, and analysed the data. GRK, HP, and JAA wrote the initial draft of the manuscript. GRK and JHY prepared figures and tables. JES and JAA validated and supervised the study. All authors reviewed the final manuscript.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-RS-2024-00358533).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information file.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.