The impact of mHealth interventions on health-promoting behaviors in pregnant women: a randomized clinical trial

Abstract

Background

Health-promoting behaviors are crucial for maternal and fetal well-being during pregnancy, yet many women face barriers to adopting these practices. Mobile health technologies offer innovative solutions by providing personalized interventions. This study evaluated the effectiveness of a comprehensive mHealth application in promoting health behaviors among pregnant women.

Methods

In a two-arm, parallel-group, randomized controlled trial, 80 pregnant women aged 18–45 years with gestational age between 18 and 36 weeks were recruited from prenatal clinics affiliated with Mashhad University of Medical Sciences, Mashhad, Iran between May and September 2023. Participants were randomly assigned (1:1) using computer-generated sequence and centralized allocation concealment to either an intervention group receiving the mHealth app or a control group receiving standard prenatal care. The app, developed by healthcare professionals, provided multimedia resources and step-by-step guidance on health responsibility, exercise, nutrition, spiritual improvement, interpersonal relations, and stress management. Data were collected at baseline and after four weeks using the Health Promoting Lifestyle Profile II. Statistical analyses, following the intention-to-treat principle, included analysis of covariance to assess the intervention effect while controlling for baseline scores, independent t-test, and chi-square test using SPSS Version 22. Missing data were addressed using Multiple Imputation.

Results

Both groups were homogeneous in terms of personal and obstetric characteristics at baseline. The intervention group exhibited significant improvements in health-promoting behaviors compared to the control group, with higher HPLP-II scores (mean difference: 29.5, P = 0.004). Significant improvements were observed across all HPLP-II subscales, including health responsibility, exercise, nutrition, spiritual improvement, interpersonal relations, and stress management. The effectiveness of intervention was consistent across various demographic subgroups.

Conclusions

The mHealth intervention significantly enhanced health-promoting behaviors among pregnant women, demonstrating the practical application of mHealth tools in supporting healthier lifestyles during pregnancy. These findings suggest that mHealth technologies can be effectively integrated into prenatal care to improve maternal and fetal outcomes. Future research should explore the long-term impacts and broader applicability of such interventions across diverse populations.

Trial registration

The trial was registered at https://irct.behdasht.gov.ir (identifier: IRCT ID: IRCT20180412039276N2) on May 23, 2022.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-27137-2.

Introduction

Pregnancy is a profoundly transformative period in a woman’s life, marked by significant physiological, psychological, and behavioral changes. Ensuring optimal maternal health during this critical time is of paramount importance for the well-being of both the mother and the developing fetus [1]. Health-promoting behaviors, including nutrition, exercise, stress management, health responsibility, spiritual improvement, and interpersonal relations play a crucial role in mitigating pregnancy-related complications and optimizing birth outcomes [2]. Barriers such as lack of access to reliable information, time constraints, social and cultural influences, and inadequate support systems can significantly impede the adoption of health-promoting behaviors. These challenges can have lasting consequences, potentially increasing the risk of pregnancy complications and compromising birth outcomes [3–5].

The rise of mobile health (mHealth) technologies has generated renewed hope for addressing these significant gaps. By utilizing the widespread accessibility of smartphones and wearable devices, mHealth interventions provide tailored support directly to pregnant women [6]. These interventions leverage various communication channels, such as mobile applications, text messaging services, and wearable sensors, to deliver tailored educational contents, reminders, and support to pregnant women, aiming to promote healthier behaviors and improve maternal health outcomes [6, 7]. However, the effectiveness of mHealth interventions can vary significantly depending on the sociodemographic context.

In a pilot study focused on pregnant adolescent girls in South Africa, while the overall appointment attendance did not significantly differ between intervention and control groups, those who engaged more with the mHealth messages demonstrated improved knowledge and attitude towards antenatal care [8]. Furthermore, a systematic review highlighted that mHealth tools effectively improve various health outcomes for pregnant women, with 82% of studies reporting positive results [9]. Additionally, mHealth-based lifestyle interventions have been linked to reduced risks of gestational diabetes and improved pregnancy outcomes, particularly among women with overweight or obesity [10]. However, challenges such as low engagement and high heterogeneity in study designs suggest the need for more tailored and rigorous trials.

One randomized controlled trial was conducted in Iran, demonstrated that the telemidwifery app effectively reduced fear of childbirth and increased self-efficacy among primiparous women, leading to a lower cesarean section rate compared to the group which received routine care [11]. Additionally, an Iranian qualitative research highlighted the perceived needs of pregnant women, indicating a strong desire for comprehensive self-care applications that address nutrition, health monitoring, and risk factors [12]. However, a review of existing Iranian pregnancy-related apps revealed that while their quantity and user acceptance have increased, only a small percentage are developed with obstetrician involvement, raising concerns about the credibility of the information provided [13].

In countries like Iran, where cultural norms and limited access to healthcare can add an extra layer of complexity, addressing these challenges through a one-size-fits-all approach might not be sufficient [14]. While mHealth interventions hold immense potential, there is a lack of robust evidence to confirm their effectiveness in such specific settings [15].

This study seeks to address this critical gap by evaluating the impact of mHealth interventions on health-promoting behaviors in Iranian pregnant women. Through a rigorous randomized controlled trial, we exam the hypothesis that mHealth interventions can effectively promote healthier lifestyles during pregnancy. The aim of this study was assessing the effectiveness of a targeted mHealth intervention in improving the health-promoting behaviors during pregnancy. The findings of this study will contribute to the growing body of evidence on the potential of mHealth technologies to empower pregnant women, enhance maternal health outcomes, and ultimately promote the well-being of both mothers and their developing children.

Methods

This study employed a two-arm, parallel-group, randomized controlled trial design to evaluate the effectiveness of a comprehensive mHealth intervention in promoting health behaviors among pregnant women. The study adhered to CONSORT guideline, ensuring rigorous methodology and transparent reporting.

Participants were randomly assigned (1:1 allocation ratio) to either the intervention group receiving the mHealth application or the control group receiving standard prenatal care. The randomization sequence was generated using computer generated random numbers, and allocation concealment was guaranteed by employing a centralized, opaque randomization service, thereby minimizing the risk of selection bias. Blinding of participants and healthcare providers was not feasible due to the nature of the intervention (using an app). To minimize potential performance bias due to the lack of blinding, several strategies were employed. The control group received structured educational materials on health-promoting behaviors. Participants in the intervention group had access to the mHealth application for one months to ensure comparable health information exposure. Both groups were contacted at similar intervals to monitor their participation and provided study-related reminders. Additionally, data analysts remained blinded to group allocation during statistical analyses to reduce detection bias.

Sample size calculation

The sample size was determined using G Power, considering an effect size of 0.5, a power of 80%, and a significance level of 0.05. This calculation suggested a sample size of 64 participants (32 per group). To account for potential attrition (loss to follow-up) during the study, we increased the sample size to 80 participants (40 per group).

Recruitment and inclusion/exclusion criteria

Pregnant women were recruited between May and September 2023, from selected prenatal clinics affiliated with Mashhad University of Medical Sciences, Iran. Recruitment strategies included flyers, posters, and information sessions.

Inclusion criteria were

• Pregnant women aged 18–45 years.

• Willingness to participate in clinical trial.

• Minimum educational qualification of a high school diploma.

• No significant pre-existing medical or background diseases, such as diabetes, heart disease, hypertension, or other chronic illnesses that could interfere with participation.

• Gestational age between 18 and 36 weeks at the time of recruitment.

• Ownership of a smartphone compatible with the mHealth application.

• Obtaining a minimum electronic literacy score of more than 50.

Exclusion criteria were

• High-risk pregnancies (e.g., multiple gestations, preeclampsia), diagnosed through medical records and self-report.

• Participation in other clinical trials or mHealth programs during the study period.

• Discontinuation of app use for or unwillingness to use the mHealth application regularly.

• non-use of the app for more than seven consecutive days (Criteria related to app non-use were applied only for per-protocol analyses and did not affect inclusion in the intention-to-treat analysis).

• Failure to continue the follow up period, such as withdrawal from the study, or the occurrence of significant life events like death of a close relative, severe financial distress (e.g., bankruptcy), or other major disruptions that could impact participation.

• Failure to meet criteria for active usage of mHealth app (logging in at least once per week and completing one educational module per week).

Intervention

The intervention group received access to a pre-designed mobile health application specifically designed to promote healthy behaviors during pregnancy. The health promotion lifestyle app for pregnancy was developed by a team of healthcare professionals specializing in maternal health. This app provided expectant mothers with practical self-management recommendations designed to support health-promoting behaviors.

The app included a variety of multimedia resources such as audio guides, images, and videos that offered step-by-step instructions focusing on the key areas of health-promoting behaviors such as exercise, stress management, health responsibility, spiritual improvement, and interpersonal relations. Additionally, all materials could be downloaded for offline use, ensuring continuous access regardless of internet connectivity. By providing a comprehensive toolkit for self-care, this app aimed to empower pregnant women to take an active role in their health, ultimately contributing to better health-promoting outcomes for both mother and baby.

To ensure equitable access to the intervention, the study provided participants with support for internet access. Considering that the cost of mobile data could present a barrier for some participants, we reimbursed participants for their data usage or provided subsidized data plans, where necessary. This arrangement was discussed with participants during the informed consent process, ensuring they were fully informed about the internet costs and their options for support.

Participants in the intervention group were provided with unique login credentials to access the app on their smartphones. The application sent regular reminders, educational content, and interactive activities tailored to each participant’s stage of pregnancy. Participants in the intervention group had access to the mHealth application for one month, a timeframe chosen based on prior evidence suggesting that sustained short-term digital interventions can effectively promote behavioral change [11].

Participants’ engagement with the app was tracked and monitored to assess the frequency and duration of use. The app automatically recorded comprehensive usage data. The operational definitions for these metrics were:

• Login: Defined as the successful entry into the application interface using unique credentials. To exclude rapid re-entry, sequential logins within a 5-minute period were treated as a single session.

• Active usage time: Calculated based on the time difference between consecutive logged interactions (e.g., screen transitions, button clicks). Inactive sessions (no recorded interaction) exceeding 10 min were truncated to 10 min to mitigate the impact of participants leaving the app open in the background.

• Completion of educational module: Defined as viewing all content within a module and successfully passing the associated end-of-module comprehension check (if applicable).

The study defined active usage for the per-protocol analysis as logging into the app at least once per week and completing a minimum of one educational module per week over the four-week period. This usage data was collected and analyzed to ensure participants were actively engaging with the app as intended. The research team regularly reviewed the usage data to monitor adherence.

Control group

The control group received routine prenatal care, which included standard prenatal visits and health education through pamphlets on health-promoting behaviors. Both groups received these routine services; however, only the intervention group had access to the additional support provided by the mHealth app.

Data collection

Data collection employed a pre-test, post-test design with a four-week interval between assessments. Prior to randomization, all participants completed a comprehensive questionnaire, included an assessment of participants’ prior health behaviors and social support using a structured questionnaire. This questionnaire collected information on dietary habits, physical activity levels, stress management practices, and the presence of social support systems, such as family or community assistance during pregnancy. These factors were assessed to ensure baseline comparability between the intervention and control groups and to identify any potential confounding variables that could influence the study outcomes. Moreover, an electronic literacy questionnaire was administered to assess participants’ ability to effectively use digital tools, particularly the mHealth application. Also, Health Promoting Lifestyle Profile II (HPLP-II), a validated tool for assessing health-promoting behaviors. The HPLP-II, developed by Susan Noble Walker and colleagues in 1995, consists of 52 items divided into six subscales: health responsibility (9 items), exercise (8 items), nutrition (9 items), spiritual improvement (9 items), interpersonal relations (9 items), and stress management (8 items). Each item is rated on a 4-point Likert scale ranging from 1 (never) to 4 (routinely). The total score of the HPLP-II is obtained by summing the scores of all items, with possible scores ranging from 52 to 208, where higher scores indicate a greater frequency of health-promoting behaviors. Internal consistency reliability for the total scale and subscales has been reported to be high, with Cronbach’s alpha coefficients typically ranging from 0.70 to 0.90 in various studies [16, 17]. In our context, we conducted additional pilot testing to ensure the instrument’s cultural appropriateness and internal consistency within our study population, with Cronbach’s alpha coefficients for the HPLP-II subscales ranging from 0.72 to 0.88, indicating good reliability.

Outcome measures

The primary outcome of this randomized controlled trial was the change in health-promoting behaviors score, assessed using the Health Promoting Lifestyle Profile II (HPLP-II). This instrument measures overall health-promoting behaviors, with higher scores indicating a greater frequency of such behaviors. Both the total score and subscale scores were evaluated at two time points: baseline (pre-test) and after four weeks of intervention (post-test).

Data analysis

Statistical analysis was conducted using SPSS Version 22.0 (IBM Corp., Armonk, NY, USA). Baseline characteristics between the intervention and control groups were compared to ensure equivalence at the start of the study. The analysis was primarily conducted under the intention-to-treat (ITT) principle, including all participants as originally assigned regardless of their adherence to the intervention.

Descriptive statistics were used to summarize participant characteristics and baseline measures. The normality of the distribution of continuous variables was assessed using the Shapiro-Wilk test. Independent t-tests and chi-square tests were employed to assess baseline comparability.

The primary analysis of the post-test HPLP-II total and subscale scores was conducted using Analysis of Covariance (ANCOVA). This approach was chosen to maximize statistical power and precision by controlling for baseline imbalances and known covariates.

The ANCOVA model included the following variables:

• Dependent Variable: Post-test HPLP-II total score.

• Fixed Factor: Treatment Group (Intervention vs. Control).

• Covariates: Baseline HPLP-II total score (to adjust for pre-existing differences), age, and highest level of education.

Missing data were addressed using Multiple Imputation by Chained Equations (MICE) under the assumption of missing at random (MAR). This method was preferred over simple imputation to provide less biased and more efficient parameter estimates. We performed five imputations with 10 iterations each. The imputation model was inclusive, containing all pre- and post-test HPLP-II scores, group allocation, and auxiliary demographic variables (age and education) to ensure the robustness of the estimates. Effect sizes were calculated to quantify the magnitude of observed changes. A repeated-measures ANOVA was also performed as a sensitivity analysis to compare the changes in HPLP-II scores from pre-test to post-test between the groups. Secondary analyses were conducted based on per-protocol (PP) criteria, which included only participants who completed the study according to the protocol.

To report the RCT study, the CONSORT checklist was used (Supplementary File 1).

Results

Participants and group assignment

Out of 91 eligible women approached for the study, 80 participants (87.9%) provided consent and were randomized into the control (n = 40) and intervention groups (n = 40). Participant flow and attendance details are illustrated in Fig. 1. At follow-up (Time 2), the response rates were 75.0% for the intervention group and 80.0% for the control group. The dropout rates between groups were noted but did not significantly differ (p = 0.504).

Fig. 1.

The CONSORT flow diagram of the study. ITT = intention-to-treat

Demographic characteristics

Baseline demographic and clinical characteristics of the participants are summarized in Table 1. Both the control and intervention groups were comparable across all measured variables at enrollment. No statistically significant differences were observed between the groups (all p-values > 0.05), confirming the effectiveness of the randomization process in ensuring baseline equivalence.

Table 1.

Baseline characteristics of participants

Variable	Control group (n = 40)	Intervention group (n = 40)	p-value
Age (years, Mean ± SD)	28.1 ± 4.7	26.8 ± 4.7	0.266
Educational level, n (%)			0.587
Diploma	2 (5.0%)	3 (7.5%)
Bachelor’s degree	33 (82.5%)	34 (85.0%)
Master’s degree	3 (7.5%)	2 (5.0%)
PhD	2 (5.0%)	1 (2.5%)
Income level, n (%)			0.545
Enough	35 (87.5%)	33 (82.5%)
Less than enough	5 (12.5%)	7 (17.5%)
Employment status, n (%)			0.389
Employee	17 (42.5%)	22 (55.0%)
Freelance	23 (57.5%)	18 (45.0%)
Residence, n (%)			0.317
Urban	31 (77.5%)	34 (85.0%)
Rural	9 (22.5%)	6 (15.0%)
Number of children (Mean ± SD)	1.2 ± 0.5	1.3 ± 0.6	0.560
Number of gravidity (Mean ± SD)	2.5 ± 0.8	3.3 ± 0.9	0.637
Gestational age at enrollment (weeks, Mean ± SD)	28.1 ± 4.2	26.3 ± 4.4	0.501

Primary outcome: HPLP-II scores

The primary outcome analysis focused on the change in Health-Promoting Lifestyle Profile II (HPLP-II) scores from baseline to post-intervention. Descriptive statistics and the primary ITT results (based on multiple imputation) are detailed in Table 2.

Table 2.

Changes in Health-Promoting Lifestyle Profile II (HPLP-II) scores from baseline to post-intervention by intention-to-treat analysis

Variable	Control group (n = 40) (Mean ± SD)	Intervention group (n = 40) (Mean ± SD)	Change in scores (Mean ± SD)	p-value (t-test)	p-value (ANCOVA)
Baseline HPLP-II scores	152.6 ± 13.2	162.0 ± 13.9	-	0.089	-
Post-intervention HPLP-II scores	168.1 ± 9.1	197.6 ± 7.8	-	0.004	0.004
Adjusted Mean Difference from ANCOVA	-	-	29.5 (95% CI: 10.7 to 48.3)	-	0.004
Change in HPLP-II scores	16.5 ± 11.7	35.6 ± 14.6	19.1 (7.3 to 30.9)	0.026	0.022 (Sensitivity)

• Baseline scores: No significant difference in baseline HPLP-II scores between groups (p = 0.089).

• Post-intervention scores: Significant improvement in the intervention group compared to the control group (p = 0.004). Effect size (Cohen’s d) for post-intervention scores was 1.32, indicating a large effect.

• Change scores: The intervention group showed a significantly greater mean change in HPLP-II scores compared to the control group (p = 0.026), with an effect size of 1.63, reinforcing the intervention’s effectiveness.

Although the baseline HPLP-II scores were slightly higher in the intervention group than in the control group (162.0 vs. 152.6), this difference was not statistically significant (p = 0.089).

The intervention effect was assessed using ANCOVA, demonstrated a significant difference between the groups. The intervention group showed significantly higher post-intervention scores compared to the control group (Adjusted Mean Difference [AMD] = 29.5, 95% CI: 10.7 to 48.3; p = 0.004).

This AMD suggests that participants in the mHealth group, when adjusted for baseline scores and key demographics, had an average HPLP-II score 29.5 points higher than the control group.

The secondary analysis using repeated-measures ANOVA on the raw change scores confirmed a significantly greater mean increase in the intervention group (35.6 ± 14.6) compared to the control group (16.5 ± 11.7) (p = 0.026). The per-protocol analysis further supported these findings, reinforcing the robustness of the intervention impact. The effect size (Cohen’s d) for the post-intervention comparison was 1.32, indicating a large effect.

Analysis of engagement metrics

Table 3 presents an analysis of participant engagement with the mHealth application and its correlation with changes in HPLP-II scores.

Table 3.

Engagement metrics in the intervention group (n = 40)

Engagement metric	Mean ± SD	Range	Correlation with HPLP-II change (r)	p-value
Total logins during the study	12.8 ± 3.4	5–20	0.42	0.008**
Average weekly logins	3.2 ± 1.1	1–5	0.39	0.012**
Total time spent on app (minutes)	245.6 ± 58.2	120–380	0.51	0.002**
Educational modules completed	7.6 ± 1.8	3–10	0.47	0.004**
Feature usage frequency (%)
- Videos watched	85.0% (34/40)	-	0.35	0.028*
- Self-assessment tools used	72.5% (29/40)	-	0.31	0.045*
- Forum participation	50.0% (20/40)	-	0.29	0.056

p < 0.05 is considered statistically significant (*), p < 0.01 is highly significant (**)

Engagement analysis demonstrated that higher app usage was associated with greater improvements in health-promoting behaviors. Participants who logged in more frequently, spent more time on the app, and completed a higher number of educational modules exhibited significantly greater increases in HPLP-II scores. Specifically, total time spent on the app (r = 0.51, p = 0.002) and the number of completed educational modules (r = 0.47, p = 0.004) were the strongest predictors of behavioral improvement. Feature-specific analysis showed that video engagement and self-assessment tool usage were positively correlated with outcomes, whereas forum participation had a weaker association. These findings highlight the importance of sustained interaction with the app in driving meaningful health behavior changes.

Subgroup analysis

To explore the potential influence of demographic factors on the effectiveness of the intervention, ANOVA was performed within the intervention group, examining the impact of age, educational level, income, and employment status on post-intervention HPLP-II scores. As shown in Table 4, no significant interactions were found between these demographic variables and the intervention effect (all p-values > 0.05), suggesting that the mHealth intervention was effective across various subgroups of pregnant women,

Table 4.

ANOVA for post-intervention HPLP-II scores in intervention group by demographic variables

Variable	F-value	p-value
Age	0.45	0.503
Educational level	0.37	0.775
Income	0.62	0.431
Employment status	0.58	0.448

HPLP-II subscale scores

Further analysis of the HPLP-II subscales, using ANCOVA adjusted for baseline subscale scores and covariates, revealed significant improvements across all domains for the intervention group compared to the control group. The increases in subscale scores were as follows:

• Health responsibility: +18 points in the intervention group vs. +8 points in the control group.

• Physical activity: +23 points in the intervention group vs. +11 points in the control group.

• Nutrition: +23 points in the intervention group vs. +11 points in the control group.

• Spiritual growth: +23 points in the intervention group vs. +11 points in the control group.

• Interpersonal relations: +20 points in the intervention group vs. +9 points in the control group.

• Stress management: +22 points in the intervention group vs. +10 points in the control group.

These results indicate that the mHealth intervention significantly improved health-promoting behaviors across all dimensions measured by the HPLP-II.

Correlation and ANOVA for participant characteristics

Correlation and ANOVA analyses revealed no significant associations between participant characteristics and changes in HPLP-II scores (all p-values > 0.05), suggesting that the intervention’s impact was not influenced by these characteristics (Table 5).

Table 5.

Participant characteristics and changes in HPLP-II scores

Correlation analysis for continuous variables
Variable	Coefficient (r)	p-value
Age	-0.02	0.831
Income	0.15	0.124
Gestational age at enrollment	0.05	0.675

ANOVA for categorical variables
Variable	F-statistic	p-value
Educational level	0.83	0.479
Employment status	2.23	0.116
Residence	0.54	0.464

Secondary analyses: per-protocol analysis

The per-protocol analysis included 30 participants in the control group and 32 in the intervention group. Although this sample size was smaller than the ITT population, the results remained consistent with the ITT findings. The mean baseline HPLP-II scores were 160.8 (SD = 13.4) for the intervention group and 154.1 (SD = 12.1) for the control group (p = 0.102), confirming that the intervention significantly improved health-promoting behaviors. Sensitivity analyses, including multiple imputation for missing data, further supported the robustness of these findings.

In post-intervention analysis, the mean HPLP-II score in the intervention group significantly increased to 197.2 (SD = 7.3), compared to 170.4 (SD = 9.2) in the control group (p = 0.003). The mean change in scores was 36.4 (SD = 14.5) for the intervention group versus 16.3 (SD = 11.2) for the control group, demonstrating a statistically significant difference (p = 0.021).

The results from the ANCOVA analysis of the imputed datasets closely aligned with the complete case analysis and the Per-Protocol analysis, further reinforcing the reliability of the study findings. Consistency across the ITT, per-protocol, and complete case analyses suggests that missing data had a negligible impact on the study conclusions, thereby validating the robustness of the intervention effects.

No adverse events or unintended effects were reported by participants in either the intervention or control groups during the study period.

Discussion

The findings from this clinical trial suggest that the mHealth intervention significantly enhanced health-promoting behaviors among pregnant women, as indicated by higher HPLP-II scores in the intervention group compared to the control group. This supports the hypothesis that mHealth interventions can effectively promote healthier lifestyles during pregnancy.

Our results align with several previous studies that have demonstrated the efficacy of mHealth interventions in improving health behaviors among various populations [18]. Our results have shown that mHealth interventions can improve various aspects of maternal health during pregnancy, including health responsibility, nutrition, spiritual growth, interpersonal relations, and stress management. A systematic review found that mHealth interventions can effectively improve the intake of micronutrient supplements and nutritional status among pregnant women in low and lower-middle-income countries [19]. This suggests that mHealth can promote health responsibility and better nutrition during pregnancy. Additionally, a quasi-experimental study evaluated the impact of an mHealth intervention called “MIESRA mHealth” on marital satisfaction during pregnancy. The intervention, which included mindfulness exercises for couples, led to improved marital satisfaction scores compared to the control group [20]. This indicates mHealth can foster spiritual growth and strengthen interpersonal relations between partners during pregnancy.

However, our findings contrast with some studies that have reported limited or no significant effects of mHealth interventions on health behaviors. For example, a study investigating the impact of a toll-free hotline and mobile messaging service in a rural, resource-limited setting found no effect of the mHealth intervention on receiving the recommended four antenatal care visits during pregnancy and a negative effect on tetanus toxoid coverage. The study also reported no effect of the mHealth intervention on postnatal check-ups and vitamin A uptake among pregnant women [21]. This suggests that mHealth interventions alone may not be sufficient to improve certain maternal health behaviors without addressing other barriers. The discrepancy in results may be attributed to differences in study design, the specific features of the mHealth interventions, participant demographics, or the duration of the intervention.

While the reviewed studies did not directly assess other health behaviors like nutrition, spiritual growth, interpersonal relations, or stress management during pregnancy, the findings highlight the importance of tailoring mHealth interventions to specific contexts and addressing potential barriers to ensure their effectiveness in promoting various health behaviors among pregnant women.

The effectiveness of the mHealth intervention across various demographic subgroups in our study suggests that such interventions can be broadly applicable. This finding is crucial as it highlights the potential for mHealth tools to reach diverse populations, including those who may have limited access to traditional health promotion resources. Previous studies have shown varying degrees of effectiveness across different demographic groups. There is limited evidence directly assessing whether the effectiveness of mHealth interventions during pregnancy for improving health behaviors is associated with differences in outcomes across various demographic subgroups. For example, a systematic review on the effectiveness of mHealth-based educational interventions to improve self-care during pregnancy does not report on the association between demographic factors and differences in outcomes [22]. Also, the protocol for a systematic review on the efficacy of mHealth interventions for improving maternal and neonatal outcomes among pregnant women with hypertensive disorders does not provide any findings related to demographic differences in outcomes [23]. Conversely, a study by Jarke et al. (2019) noted that socioeconomic status and educational level could influence the effectiveness of mHealth interventions, with more significant benefits observed in higher-income and more educated populations [24]. More research is needed to specifically examine this aspect.

Strengths and limitations

This study’s strength lies in its comprehensive evaluation of health-promoting behaviors using HPLP-II subscales, providing a detailed assessment of the intervention’s impact. Additionally, the inclusion of diverse demographic subgroups enhances the generalizability of findings.

Despite the positive findings, several limitations must be considered in accordance with CONSORT reporting standards. First, regarding potential sources of bias, blinding of participants and healthcare providers was not feasible due to the nature of the digital intervention. This may have introduced performance bias, although we mitigated this through blinded data analysis and a structured education protocol for the control group. Furthermore, the use of self-reported HPLP-II scores carries a risk of social desirability bias or recall errors; future studies should integrate objective measures, such as wearable trackers, to improve data accuracy.

Second, in terms of imprecision, while our sample size was based on power calculations, the relatively small cohort and the short follow-up period (four weeks) limit our ability to determine the long-term sustainability of the observed behavior changes. Third, the generalisability of these results may be restricted, as the study was conducted in a specific urban setting in Iran among women with prior mHealth experience; therefore, the findings may not apply to rural populations or those with lower digital literacy.

Fourth, we acknowledge the potential for multiplicity of analyses given the numerous HPLP-II subscales tested. Although the consistent significance across all domains suggests a robust effect, the risk of Type I error should be considered. Finally, despite randomization, a slight baseline imbalance in HPLP-II scores was noted. While we addressed this using ANCOVA adjustments to ensure the intervention’s effect remained statistically valid, future research should utilize stratified randomization to further minimize such imbalances. Addressing these factors through objective tracking and extended follow-up will provide deeper insights into the long-term impact of mHealth interventions.

Implications for practice

This study demonstrates that mHealth interventions can effectively enhance health-promoting behaviors among pregnant women, making them a valuable addition to prenatal care. Healthcare providers can use these digital tools to deliver personalized health education and provide ongoing support, thereby improving maternal and fetal outcomes. The success of intervention across diverse demographic groups suggests that it could be widely implemented, particularly in underserved areas. Integrating mHealth into routine prenatal care could lead to better engagement, adherence, and long-term health benefits, making it a promising approach for improving maternal health on a broader scale.

Conclusion

In conclusion, our study adds to the growing body of evidence supporting the use of mHealth interventions to promote health-promoting behaviors among pregnant women. The significant improvements observed in the intervention group across all HPLP-II subscales underscore the potential of mHealth tools to facilitate healthier lifestyles during pregnancy. Future research should continue to explore the long-term effects of such interventions and identify strategies to enhance their effectiveness across diverse populations.

Abbreviations

mhealth

Mobile health

HPLP-II

Health Promoting Lifestyle Profile II

RCT

Randomized controlled trial

ITT

Intention-to-treat

PP

Per-protocol

ANCOVA

Analysis of covariance

MICE

Multiple Imputation by Chained Equations

AMD

Adjusted Mean Difference

CI

Confidence interval

SD

Standard deviation

Authors’ contributions

FA, SEZ and RLR contributed substantially in the conception and design of the study. FA carried out the data collection and analysis. RLR supervised the data collection and verified the analytical methods and statistical interpretations. FA drafted the manuscript. RLR and SEZ reviewed the manuscript critically for important intellectual content. All authors read and approved the final manuscript and agreed to be accountable for all aspects of the work.

Funding

This study is part of the first author’s (FA) doctoral thesis supervised by Professor Robab Latifnejad Roudsari, and funded by Vice Dean for Research, Mashhad University of Medical Sciences, Mashhad, Iran under grant number of 4000201.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information file 1.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Institutional Review Board of Mashhad University of Medical Sciences, Mashhad, Iran (IR.MUMS.REC.1400.179). Written informed consent was obtained from all participants prior to enrollment in the study. Participants were given the option to withdraw from the study at any time without any prejudice. Data security measures were implemented to protect panticipants’ information, including encrypted data storage and restricted access to study data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.