This review adheres to the PRISMA-P (Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols) 2020 standards [17] and is registered with International Prospective Register of Systematic Reviews (PROSPERO; ID: CRD42023475665), which ensures transparency and reproducibility. The completed PRISMA 2020 checklist is provided in Checklist 1. All protocol amendments will be documented in PROSPERO.

Studies will be selected according to the Population Intervention (or Exposure, for observational studies) Comparator Outcomes Study design (PICOS) framework as follows.

The comparator groups will include pregnant women with no reported exposure to nonionizing EMWs, background environmental exposure only, or the lowest exposure category defined within each individual study. The comparator definitions will be extracted as reported and assessed for comparability across studies.

The search strategy will be developed and refined in consultation with an experienced university health sciences librarian to ensure methodological rigor and comprehensive coverage of the literature. The final search strategy will combine controlled vocabulary (eg, Medical Subject Headings terms) and free-text keywords related to EMWs, pregnancy, and reproductive outcomes and will be adapted for each database.

The electronic databases to be searched include PubMed/MEDLINE, Scopus, Web of Science, Embase, ScienceDirect, SpringerLink, Wiley Online Library, and Google Scholar.

In addition, clinical trial registries will be searched, including the WHO International Clinical Trials Registry Platform. Where relevant, other registries, such as ClinicalTrials.gov, will be screened to identify completed or ongoing studies with unpublished results.

In addition to bibliographic databases, gray literature sources will be searched to minimize publication bias. These include ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform. Embase and the Cochrane Library are included to ensure comprehensive coverage of biomedical, epidemiological, and clinical trial literature not fully indexed in PubMed.

Two reviewers will independently assess titles and abstracts for relevance. Full texts will then be reviewed for eligibility. Discrepancies will be resolved through discussion or by a third reviewer.

The selection process will be documented using a PRISMA 2020 flowchart (Figure 1).

Data will be extracted into Microsoft Excel, capturing the following:

When information is incomplete, the authors will be contacted for clarification.

Only primary human studies will be included in the risk of bias assessment and quantitative synthesis. The risk of bias in nonrandomized observational studies will be assessed using the Risk Of Bias In Nonrandomized Studies of Interventions tool, which evaluates bias across the domains of confounding, participant selection, exposure classification, deviations from intended exposure, missing data, outcome measurement, and selective reporting. Where randomized controlled trials are identified, the Cochrane Risk of Bias 2 tool will be applied. Each study will be categorized as having low, moderate, serious, or critical risk of bias according to the tool’s guidance. Studies judged to be at critical risk of bias will be excluded from quantitative synthesis but described narratively. Reporting quality will be evaluated descriptively using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist, without influencing risk of bias judgments.

Statistical heterogeneity will be assessed using the I² statistic and chi-square test. In accordance with the Cochrane Handbook, I² values will be interpreted cautiously and in context, recognizing that fixed thresholds can be misleading and that the importance of heterogeneity depends on multiple factors, including the magnitude and direction of effects and the strength of evidence for heterogeneity (eg, CIs around I² and the number of included studies).

Meta-analyses will be conducted using random-effects models when appropriate. Rather than relying on rigid I² cutoffs to define “substantial” heterogeneity, heterogeneity will be primarily addressed through clinical and methodological assessments, including differences in study populations, exposure metrics, outcome definitions, and confounder adjustment.

To better reflect the between-study variability and the expected range of true effects across different settings, prediction intervals will be calculated and reported for random-effects meta-analyses, in line with contemporary methodological guidance [18].

When quantitative synthesis is feasible, a random-effects meta-analysis will be performed using the DerSimonian–Laird or restricted maximum likelihood approach, depending on the data structure. Effect estimates will be reported with 95% CIs and, where applicable, with prediction intervals. Sensitivity analyses will be conducted to assess the robustness of the findings to study quality and exposure definitions.

Given that the available evidence is predominantly observational, included studies must demonstrate a priori adjustment for key confounders to be eligible for quantitative synthesis. At a minimum, studies must adjust for the following:

Studies failing to account for these minimum confounders will be excluded from meta-analysis and described narratively only. This minimum adjustment set will be applied a priori as a criterion for inclusion in quantitative synthesis. Residual confounding will be explicitly considered in the interpretation of findings, and causal language will be avoided in accordance with the limitations of observational evidence.

The overall quality and certainty of evidence for each outcome will be graded using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) system as high, moderate, low, or very low [19].

This study is a systematic review protocol based exclusively on the published literature and does not involve the collection of primary data or interaction with human participants. Therefore, ethical approval from an institutional review board or ethics committee is not required, in accordance with institutional and international guidelines for research involving publicly available data.

This systematic review protocol provides a transparent methodological framework for evaluating the potential effects of nonionizing EMWs on human pregnancy and fetal development. Given the rapid expansion of mobile phone use and wireless technology, understanding the biological implications of prolonged exposure to EMWs has become an increasingly important public health concern [1,2]. Despite extensive research, the existing literature remains heterogeneous, with findings ranging from no observable adverse effects to possible reproductive or developmental impacts [13,17,20]. This protocol is designed to systematically assess and contextualize these inconsistencies through a rigorous methodological appraisal.

Biological tissues can absorb EMWs, with thermal effects related to localized tissue heating representing the most well-characterized interaction [21]. In addition, nonthermal biological interactions have been described primarily in experimental and in vitro models, including oxidative stress responses, calcium channel modulation, and alterations in gene expression; however, these findings remain inconsistent, and their relevance to human pregnancy outcomes is not firmly established [1,2].

Previous human studies have reported mixed associations between electromagnetic exposure and adverse pregnancy outcomes, including miscarriage and fetal growth restriction [3,12]. A systematic review by Wdowiak et al [14] reported potential links between chronic EMF exposure and lower birth weight, placental dysfunction, and developmental delay. Similarly, computational modeling by Takei et al [11] suggested that smartphone radiation may increase localized placental temperature, potentially affecting placental perfusion. Advanced modeling studies have further explored how RF radiation from commonly used mobile phones may influence electromagnetic exposure profiles in pregnant women, highlighting potential heat distribution effects that could have biological relevance; however, these findings require empirical validation in human populations [22].

Conversely, other large-scale epidemiological studies have not confirmed significant teratogenic or adverse pregnancy risks, underscoring the substantial heterogeneity in exposure assessment methods, study design, and confounder control [5,23]. Reviews of nonionizing radiation effects have also noted that while EMW exposure may induce cellular stress responses, such as oxidative stress or altered signaling pathways, the direct clinical relevance of these mechanisms to pregnancy outcomes remains insufficiently established, reinforcing the need for rigorous and balanced systematic synthesis [24].

These inconsistencies highlight the need for a comprehensive and methodologically rigorous systematic review. Many existing studies rely on self-reported device use or indirect exposure proxies, which may introduce recall bias and exposure misclassification [25]. Additionally, limited adjustment for confounding factors such as maternal stress, occupational radiation exposure, environmental pollutants, and thermal stress further complicates the interpretation of the available evidence [26].

The strengths of this protocol include adherence to PRISMA-P guidelines, a comprehensive search strategy, and the use of contemporary, domain-based risk of bias assessment tools appropriate for observational evidence, with explicit consideration of confounding and exposure misclassification. Nonetheless, heterogeneity in exposure metrics, outcome definitions, and study populations may limit the certainty of any conclusions that can be drawn; therefore, the findings will be interpreted with appropriate caution.

By synthesizing evidence from human observational studies, this systematic review aims to improve understanding of potential associations between nonionizing EMW exposure and pregnancy outcomes. Evidence from experimental and modeling studies will be discussed contextually in the Discussion to support biological plausibility, without being included in the quantitative synthesis. Experimental studies have suggested that RF exposure may influence oxidative balance, mitochondrial integrity, and calcium signaling in placental and fetal tissues under specific experimental conditions [20,27].

The anticipated contributions of this review are threefold: (1) to synthesize fragmented evidence regarding possible associations between EMW exposure and pregnancy outcomes while carefully accounting for study quality and bias; (2) to critically evaluate exposure assessment methodologies and identify research gaps, particularly regarding real-time dosimetry and gestational timing of exposure; and (3) to support evidence-informed discussion and future research directions rather than definitive risk conclusions, consistent with precautionary principles proposed by international organizations such as the WHO and ICNIRP.

The findings of this systematic review will be disseminated through publications in peer-reviewed open-access journals and presentations at relevant scientific conferences. In addition, the results may be communicated to academic and public health stakeholders to facilitate transparent, evidence-based interpretation and inform future research priorities.

This protocol provides a structured and reproducible framework for systematically synthesizing evidence of nonionizing EMW exposure during pregnancy. By applying rigorous methodological standards and transparent quality assessments, the forthcoming review aims to clarify the current state of evidence while explicitly acknowledging its limitations. The conclusions will be framed cautiously, reflecting the strength and consistency of the available data and are intended to support responsible scientific interpretation and future research rather than draw definitive clinical or causal inferences.