Cardiovascular risk prediction in women with gestational diabetes using risk assessment tools: a scoping review

Abstract

Background

Gestational Diabetes Mellitus (GDM) is a glucose intolerance diagnosed during pregnancy and is associated with an increased risk of developing cardiovascular disease (CVD) later in life. Despite this well-established link, it remains unclear how effectively existing CVD risk assessment tools account for the unique risks faced by women with a history of GDM. This scoping review aims to systematically explore the literature to examine how CVD risk prediction tools have been applied to this population and whether they adequately capture their long-term risk.

Methods

A comprehensive search was conducted across multiple databases, including Medline (via Ovid), Embase (via Ovid), Scopus, PubMed, PsycINFO, CINAHL, along with grey literature from Google Scholar. The overall search was conducted from the database’s inception date to January 28, 2026. A total of 508 studies underwent title and abstract screening, with 18 selected for full-text data extraction. Three reviewers independently screened studies and extracted data, ensuring consistency and reliability. A PRISMA flow diagram was used to illustrate the study selection process.

Result

Among the identified CVD risk prediction tools, the Framingham risk score (FRS) 10-year estimator was one of the most frequently applied in 7 (38.9%) of the included studies. This was followed by the ASCVD-Pooled Cohort risk equations (ASCVD-PCE) risk calculator and Framingham-30 years risk lipid-based calculator with 5 (27.8%) and 4 (22.2%).

Conclusion

Despite their widespread use, these tools primarily assess traditional cardiovascular risk factors and do not incorporate pregnancy-related indicators, which are significant predictors of future CVD risk in women with a history of GDM. This scoping review underscores the need for the development or adaptation of risk prediction models that specifically address the long-term cardiovascular risk profile of this population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13098-026-02154-8.

Introduction

Gestational diabetes Mellitus (GDM) is defined as a glucose intolerance of varying degrees, with onset or first recognition during pregnancy [1, 2]. It is the most prevalent medical complication associated with pregnancy. According to the International Diabetes Federation (IDF), the global prevalence of GDM is estimated at 14.0% [3]. Australia, like many other countries, is experiencing a rapid increase in the prevalence of GDM. From 2021 to 2022, Australian hospital records revealed that more than one in six women aged 15–49 who gave birth were diagnosed with GDM, amounting to 17.9% of this demographic [4].

The implications of GDM extend beyond the perinatal period, leading to various maternal and fetal complications [5]. These encompass an increased risk of preterm delivery, which is associated with escalated morbidity and mortality rates for both mother and infant [6]. Additionally, preeclampsia- a hypertensive disorder related to pregnancy- is more prevalent in most GDM-affected pregnancies, which complicates maternal management and increases the likelihood of adverse outcomes [5–7]. Despite clear evidence that GDM imposes a substantial risk for post-partum women, follow-up remains limited in most parts of the world [8–12] which might be associated with both the healthcare system and personal factors, and patient-related barriers [13–15].

Women with a history of GDM have a twofold higher risk of developing CVD [16–18]. This higher risk of cardiovascular dysfunction in people with GDM is linked with increased blood pressure, adverse changes in triglycerides and cholesterol, and other cardiovascular risk variables [19]. Many strategies exist to reduce CVD burden, and an essential approach is risk prediction and identifying individuals at high risk of developing CVD [20–22].

Moreover, CVD prevention guidelines have shifted from targeting individual risk factors, like high blood pressure, to adopting an absolute risk approach that considers multiple risk factors to estimate the likelihood of a cardiovascular event within the next 5–10 years [23]. The probability of developing CVD was estimated using various risk prediction tools. It is also clear that the tools must be designed for particular populations, taking into account their specific characteristics, such as the levels of risk factors, the relative impact of each risk factor, and the population’s average risk of CVD [24].

Framingham’s CVD risk score (FRS) [25], Systematic Coronary Risk Evaluation (SCORE) [26], Quality Risk calculator for Integrated System-wide Knowledge(QRISK) [27], Australian Cardiovascular Disease risk calculator (AusCVD) [28], and World Health Organization (WHO) risk prediction tools [29] are some of the CVD risk assessment tools that are being used in various areas of the globe. While the tools are useful for assessing general cardiovascular risk, they may not fully capture the elevated risk profile associated with GDM without additional clinical context or risk assessments tailored specifically to this population. The existing literature does not provide a unified synthesis on whether current cardiovascular risk prediction tools include pregnancy-related factors, nor does it identify any tools specifically developed for women with a history of gestational diabetes. It is also unclear how general population tools are being applied to this group or what limitations have been reported in their use. Therefore, this scoping review aimed to collate and examine the current body of literature to understand how cardiovascular risk assessment tools have been applied in women with GDM.

Methods

This scoping review process was structured based on the framework provided by Arksey and O’Malley [30], which includes the following stages: identifying the research question, identifying relevant studies, selecting studies, charting data, and reporting results. Additionally, advanced methodological recommendations proposed by Levac, Colquhoun, and O’Brien were incorporated to enhance the rigour of the review [31]. The screening process was conducted using Covidence, ensuring a systematic and transparent selection of studies. To further strengthen the review’s methodological rigour, a standardised data extraction form was developed, accompanied by detailed guidance for its use [32].

Eligibility

This scoping review is structured under the PCC framework (Population, Concept, and Context), as guided by the Joanna Briggs Institute (JBI) to develop a clear objective and eligibility criteria [33]. The review is limited to English-language written publications.

Participants

The population for this review was women who were diagnosed with gestational diabetes in their at least one pregnancy, regardless of pregnancy outcome (including live birth, stillbirth, miscarriage, or termination).

Concept

In this scoping review, studies that assessed cardiovascular disease risk assessment tools, such as risk scoring tools (FRS, ASCVD-PCE, SCORE, QRISK, AusCVD, WHO risk prediction tools, etc.), were used to quantify future CVD and predict outcomes such as complications, morbidity, or mortality among women with gestational diabetes.

Context

The context of this review was the care provided in the individual’s home residence or Health institution, with no restrictions on the setting.

Types of studies

This scoping review included quantitative, qualitative and mixed research. Non-empirical studies, such as reviews, theoretical papers, policy reports, commentaries, letters to the editor, conference proceedings, and pilot studies, were excluded.

Outcomes

The review used the definition of CVD from the Australian CVD Risk guideline, which refers to all conditions affecting the heart and blood vessels, including (Myocardial infarction (MI), Angina, Other coronary heart disease (CHD), Stroke, Transient ischemic attack, Peripheral vascular disease, Congestive heart failure, Other ischemic CVD-related conditions) [28].

The design of this scoping review is also grounded in a framework adapted from critical appraisal criteria forwarded by Moons, K.G., et al. [34]. Using the framework as shown in Table 1, this scoping review clearly outlines the detailed criteria employed to comprehensively explore the literature.

Table 1.

A framework adapted from a study to review studies about risk prediction tools [34]

Evaluation criteria	Detailed explanation
Prognostic versus diagnostic prediction model	The aim is to predict the future occurrence of CVD
Purpose of using the model	To inform clinicians’ therapeutic decision-making
Type of prediction modelling studies	Prediction models with or without external validation
Target population	Pregnant women with GDM, per the diagnostic criteria in the clinical setting
The outcome to be predicted	CVD based on ICD-10
The time span of prediction	CVD occurring after the occurrence of GDM
Intended moment of using the model	Models to be used at the time of assessing CVD risk in clinical or community settings

Quality assessment

Even though quality appraisal is not considered a mandatory component of scoping review, scholars have reported that including it can enhance the robustness of the findings [33, 34]. Therefore, methodological quality appraisal of included studies was conducted using the JBI checklist matched to each study design (Cross-sectional and cohort), which is consistent with the scoping review methodology aimed at mapping available literature [35]. However, the studies included met the criteria defined in the review process, and there was no study excluded based on its quality level. Three reviewers rated each study, and from all studies included, 61% of them got a high-quality rating, and the rest were rated as good quality (Supplemental document II).

Search strategy

The search strategy is built using three concepts “cardiovascular disease” “risk score” and “Gestational diabetes mellitus” which are combined using Boolean operators like “OR” or “AND”. The databases searched for this scoping review were Medline (via OVID), CINAHL (via EBSCO), PsycINFO (via EBSCO), SCOPUS, EMBASE (via OVID), and Google Scholar. Besides, the references of each eligible study were searched, and the overall search was conducted from the database’s inception date to January 28, 2026. The complete search strategy is found in the Supplemental document I, Section I.

Title and abstract screening

Three independent reviewers assessed all 687 imported articles based on the pre-established inclusion and exclusion criteria. Any conflicts of decision between any two reviewers were resolved by the third reviewer. At this stage, 508 studies went through the screening process, and 103 studies proceeded to the next stage of full-text review.

Full-text screening

To assess the eligibility and relevance of the 103 studies, two independent reviewers screened the full text against the predefined inclusion criteria. The rigorous screening led to the exclusion of 84 studies due to reasons such as wrong outcome, wrong study type, wrong tool and misalignment with our review questions and objectives. This approach ensured review questions and objectives. Ultimately, 18 studies were included for data extraction (Fig. 1).

Fig. 1.

PRISMA flow of article selection

Charting data

A total of 19 papers reached this stage; however, two of these originated from the same dataset and were therefore counted as one study [36, 37]. The one with the larger sample size and the recent publication is chosen, and this results in 18 unique studies to be included in the final review synthesis. Data extraction was designed to capture key details of each study, including the title, authors, year of publication, population, study context, study aims, study design, CVD risk assessment tools used, findings and other relevant information (Supplemental document I, Section II). Data extraction was conducted independently by two reviewers, and discrepancies were resolved through discussion and involvement of the third reviewer to reach a consensus before finalising the extracted data.

Collating, summarising and reporting results

Data extracted on the data charting form became organised, collated and summarised. The Information obtained is reported using narration, tabular, and pictorial ways. The review was reported in accordance with the PRISMA-ScR extension for Scoping Review guideline [35] (Supplemental document I, section III).

Results

Study setting and year of publication

Among 508 studies included in the title and abstract screening, 18 studies have gone to the analysis, yielding 3,099,466 participants. Of the participants included in the studies, 249,816 (8.1%) were found to be gestational diabetic. Notably, the majority of the studies were conducted from data in North America, which is among the 18 studies included in this scoping review, 5 (27.7%) were conducted in the United States, followed by 4 (22.2%) in the United Kingdom (UK) and 3 (16.6%) in Canada. Other studies were from Finland, India, Germany, Barbados, Hong Kong, Israel, Thailand, Poland, Norway, Iraq, South Africa, and Iran (Fig. 2).

Fig. 2.

Countries where studies are being published

Moreover, the number of publications has increased significantly in recent years, with most of the studies 16 (88.89%) being published after 2020 (Fig. 3).

Fig. 3.

Year of Publication of studies included in the scoping review

Studies design

The majority of studies included in this review (n = 13, 72.2%) employed a cohort study design, ten of which were followed prospectively [36, 38–46] and the others were retrospective cohort studies [47–49]. Five studies used a cross-sectional study design [50–54] (Supplemental document I, Section IV, Table 1).

CVD risk assessment tools utilised

The most widely 7 (38.9%) used CVD tool was the Framingham risk score, which was used to estimate the 10-year CVD risk among women with gestational diabetes. The second most used were the ASCVD-Pooled Cohort risk equations (ASCVD-PCE) risk calculator and Framingham-30 years risk lipid-based calculator, with 5 (27.8%) and 4 (22.2%). The other Framingham-30-year CVD risk estimator using the BMI-based calculator accounted for the same frequency, 2 (11.1%), as SCORE and the Lifetime CVD risk calculator. Lifetime CVD risk, QRISK 2, QRISK 3, PARCCS- Prediction of Acute Risk for Cardiovascular Complications in the Peripartum Period Score, WHO chart, the Finland’s FINRISK calculator, the New Zealand’s PREDICT, JBS- Joint British Societies, and Life’s Simple 7 (LS7) score were also used by studies (Fig. 4).

Fig. 4.

CVD Risk assessment tools used in studies included in this scoping review

Application of the CVD risk assessment tools

CVD risk comparison

Five studies utilised the tools to compare the CVD risk profile among women with and without a history of GDM [38–40, 53, 54]. Two studies estimated and compared risk scores between tools. The study of Moe, Sugulle et al. made a comparison between ASCVD, FRS and JBS 3 risk calculator [46] and the other study conducted by Doust et al. has compared the predictive accuracy of the New Zealand’s PREDICT, QRISK 2, and ASCVD-PCE [45]. Another study used the Framingham 30-year CVD risk tool to evaluate the CVD risk among individuals with diet-controlled and insulin-controlled GDM [42]. Additionally, one study applied a CVD risk assessment tool to compare risk profiles between two maternal clinics [47] (Supplemental document I, Section IV, Table 1).

CVD risk and post-pregnancy complications

Most studies (n = 11, 61.1%) utilised the tools to assess the occurrence of CVD following various forms of pregnancy complications [36, 41–43, 45–51, 55]. From those, two studies categorised GDM with Hypertensive Disorders of Pregnancy (HDPs), preterm birth and stillbirth as adverse pregnancy complications [49, 50]. Other studies examined the association of CVD risk and pregnancy complications by grouping GDM alongside preterm delivery, hypertensive disorders, abnormal birth weight, and anomalous size for gestational age, hypertensive disorders of pregnancy (pre-eclampsia), hypertensive disorders of pregnancy (gestational hypertension), small for gestational age, postnatal depression, stillbirth, miscarriage, preterm birth, and placental abruption [36, 43–45, 48, 51]. A further study examined the CVD risk by analysing the types of Hyperglycemia First Detected during Pregnancy (HFDP), specifically GDM and Diabetes in Pregnancy (DIP) [41]. From studies included in this scoping review, one study assessed the CVD risk among women referred to clinics specialising in maternal health. It evaluated the history of GDM-specific complications, considering factors such as hypertensive disorders, preterm birth, gestational hypertension, preeclampsia, intrauterine growth restriction, multiple gestation, and placental abruption [42].

Developing a novel CVD risk prediction tool

Unlike the other studies, one study by Zahid, Hashem et al. developed its risk prediction tool using a machine learning technique to estimate peripartum CVD complications during delivery [52]. This tool, the Prediction of Acute Risk for Cardiovascular Complications in the Peripartum Period Score (PARCCS), integrates 14 variables according to clinical relevance and predictive ability. In performance evaluation, the tool has shown an Area Under the Curve (AUC) of 68%, and the value of this scoring system ranges from 0 to 100, with defined score cut-offs to categorise and convert the predicted risk levels [52]. Another study was conducted to assess the impact of pregnancy-related factors in enhancing the 10-year predictive performance of the QRISK 3 tool. The overall discrimination and calibration of the tool in predicting 10-year post-partum cardiovascular risk remained similar to the original model. However, adding pregnancy-related factors into the tool has improved the model’s clinical utility [48].

Risk classification

This scoping review identified that CVD risk assessment tools are utilised in various forms as continuous measures or categorical labels. Moreover, other tools are found to be used with varying classification categories:

Framingham risk score as a continuous value

This review has identified various studies that use different methods to analyse the Framingham risk score. Two studies have used the risk score of FRS as a continuous measure, presenting the results in terms of mean (SD) to provide the overall assessment of the population’s CVD risk [39, 43]. A study by Field et al. modelled the 10-year and 30-year ASCVD risk as a continuous measure, allowing them to assess the linear relationship between pregnancy-related variables and the risk level [36]. Similarly, studies by Essington, Pudwell et al. and McGrath, Pudwell et al. have analysed the score of FRS-lipid-based and FRS-BMI-based risk values using median (IQR) [47, 56]. One study compared the raw value of FRS, ASCVD and JBS3 to compare the CVD risk between the three CVD risk assessment tools [46].

Framingham risk score as a categorised value

In addition to continuous modelling, there are studies that utilised the FRS using various forms of categorisation. Two studies have used a two-tier classification of categories as low risk and high risk based on the classification of < 10% and ≥ 10%, respectively [44, 51]. Other studies (n = 2) used a three-tier classification of risk level as “low risk”, “Intermediate risk” and “high risk” based on < 10%, 10–20% and > 20%, respectively [41, 50]. However, another study further classified the risk level by merging intermediate and high-risk groups and ended up with low risk (< 10%) and intermediate-high risk (≥ 10%) [41].

ASCVD-pooled cohort risk equations (ASCVD-PCE) as categorised value

Even though there is no study that assesses the ASCVD-PCE as a continuous value, the two-tier classification (< 10% and ≥ 10%) was applied in studies that used ASCVD risk calculations [45, 54]. However, other studies (n = 2) categorised ASCVD risks into four groups: low (< 5%), borderline (5% to < 7.5%), intermediate (≥ 7.5% to < 20%), and high-risk (≥ 20%) [38, 49]. Unlike others, a study by Moe, Sugulle et al. used the ASCVD-PCE tool to calculate the lifetime risk at specific intervals of 8%, 27%, 39%, and 50%. For the analysis purpose, the authors categorised by defining 8% as the “optimal risk” level. Any score exceeding this threshold (27%, 39%, or 50%) is considered “suboptimal risk” [46].

(WHO/ISH) risk chart and lifetime CVD risk

The World Health Organization/International Society of Hypertension (WHO/ISH) risk chart was found to be used with categories of < 10%, 10–19%, 20–29%, and ≥ 40% [50]. In addition to short-term risk assessments, studies (n = 3) have assessed the lifetime CVD risk (which refers to the evaluation of the individual’s lifetime cumulative risk of developing CVD) by classifying individuals based on the presence and severity of risk factors using 8%, 27%, 39%, and 50% classification. For instance, a lifetime CVD risk of 8% reflects individuals with optimal values for all these variables, including cholesterol, blood pressure, smoking, and fasting glucose [46, 47, 56]. If one or more variables are not optimal, more than one is elevated, and if there is one major risk factor or more than two elevated variables, the lifetime CVD risk would be escalated to 27%, 39%, and 50%, respectively.

Association between GDM and CVD

Among studies (n = 14) that assessed the relationship between GDM and CVD, three of them did not find a significant difference in CVD risk based on GDM history [49, 53, 54]. However, in one of these studies, the cardiovascular health levels were found to be lower among women who had a history of GDM compared to those without such a history [54].

However, from all papers included in this review, most of the studies (n = 11, 61.1%) have shown that having a GDM history is significantly associated with an increased risk of CVD. The strength of this association was found to be influenced by the type of risk assessment tool utilised. This is evidenced by a study comparing three risk assessment tools (FRS, FINRISK and SCORE), which found that CVD risk identified using Framingham and FINRISK tools was significantly associated with GDM. However, during the utilisation of the SCORE tool, it did not demonstrate a statistically significant association [39]. Moreover, one study reported that the WHO/ISH score did not show significant differences in women with GDM, but the CVD risk difference using FRS was able to be identified among women with and without GDM [50]. In addition, one study found that all ASCVD, JBS3 and FRS were found to be unable to show the epidemiological high-risk profile among women with a history of GDM [46].

Discussion

This review systematically examines the existing evidence on the utilisation of CVD risk assessment tools among women with a history of GDM. Moreover, it aimed to identify and elaborate on the limitations of these tools, as highlighted by the included studies, to inform future improvements in risk assessment methodologies.

Utilisation of FRS

The FRS was the most frequently used risk assessment tool in the included studies. This is not surprising, as FRS is widely adopted in both research and clinical practice due to its early development, ease of use, and comprehensive nature [25]. However, despite its widespread application, its applicability to women with GDM remains uncertain. One key limitation of FRS is that it was developed based on data from the general population and does not explicitly account for pregnancy-related factors, including a history of GDM. None of the studies reported modifying or validating FRS specifically for GDM populations, which suggests a gap in prediction accuracy for this cohort.

Pregnancy-related factors

In this review, most of the tools used were for the purpose of assessing CVD risk after the occurrence of adverse pregnancy-related complications (gestational diabetes mellitus, hypertensive disorders of pregnancy (pre-eclampsia), hypertensive disorders of pregnancy (gestational hypertension), small for gestational age, postnatal depression, stillbirth, miscarriage, preterm birth, placental abruption). These pregnancy complications often serve as indicators of underlying CVD risk factors, predisposing individuals to future development of vascular and metabolic disease [57, 58]. Complications, including GDM, hypertensive disorders, and preterm birth, act as a physiological stress test that reveals underlying cardiovascular vulnerabilities [6, 59]. However, many women are not aware of this risk, and they are not receiving the recommended CVD prevention care [60, 61]. This underscores how pregnancy-related factors need to be considered in the process of CVD prediction and also strengthens the concept of considering pregnancy as a physiologically demanding phase that helps as a natural cardiac stress test [62, 63]. Failing to incorporate pregnancy-related factors into risk assessment tools can lead to underestimation of risk levels and missed opportunities for timely interventions on future CVD [64]. Therefore, having CVD risk screening during the period of pregnancy and postpartum is critically helpful in optimising effective preventive interventions and ensuring long-term cardiovascular health [65, 66]. Moreover, pregnancy serves as a pivotal period during which women exhibit increased motivation to adopt healthier lifestyles and dietary practices [67]. This is one of the reasons why CVD risk assessment at the time of pregnancy or postpartum is recommended to be considered as a priority area in education, research, clinical care, and policy development [68].

WHO/ISH tool limitations

The WHO/ISH risk prediction tool is another method used for assessing CVD risk. However, it has limitations, including ambiguous categorisation for certain risk scores and a lack of specificity for women with a history of GDM. Adjusting the tool to stratify risks more appropriately and incorporating pregnancy-related factors could enhance its clinical utility for this population. Studies suggest that simplifying the WHO/ISH risk categories into two groups, as low-risk (< 20%) and high-risk (≥ 20%), may enhance clinical applicability [69]. Additionally, it’s crucial to carefully select appropriate cutoffs to balance the risks of false positives, which could lead to unnecessary interventions in low-risk individuals [52].

Undiagnosed CVD in pregnancy

A significant concern is the prevalence of undiagnosed CVD among pregnant and postpartum women. Many women who experience cardiovascular complications during these periods have not been previously diagnosed, underscoring the need for improved screening and risk assessment. Early identification of at-risk women can lead to better management and prevention of adverse outcomes [70].

Population-specific tool limitations

This review has shown the limitations of utilising a particular tool that was originally developed in a population of different cardiometabolic risk profiles compared to their cohort. For instance, the FRS was developed in a non-Hispanic white population with higher rates of tobacco use, individuals with pregestational diabetes, and varying definitions of hypertension [71]. This was found to be the source of limitation for most studies that aimed to look at the CVD risk among women with GDM [44, 56]. This finding is supported by another scoping review which shows the necessity of estimating CVD risk using a tool that best suits the individual, population characteristics [72]. Besides, it also strengthens the evidence of considering technical, methodological, and practical aspects of risk score development and validation before implementation in the new setting [73, 74]. In addition, this is also evidenced by studies which show the ASCVD risk tool, which was originally crafted based on data from the white population in the US, has been shown to overestimate the CVD risk when used on women from multiethnic backgrounds even within the same county [75, 76].

Need for tailored tools

In this review, there were studies that showed their recommendation on the variable that they believed to be incorporated into the risk prediction tool. One study highlighted the gap in the existing CVD risk assessment tool, specifically in missing some essential variables, such as adherence to treatment, and the variability of risk factors in terms of severity, fluctuation, and duration of exposure [56]. Moreover, another study suggested the variable gap in the ASCVD-PCE tool, which lacks parity, age at first birth or gravidity that may impact the occurrence of either adverse pregnancy outcome (e.g. GDM) or CVD [49].

The other evidence could be the AusCVD risk calculator, which was recalibrated using contemporary Australian population data and aligned with the Australian health system context to estimate the absolute five-year risk of a cardiovascular event. Compared with the 2012 Australian CVD Risk Management Guideline, the updated tool incorporates substantial methodological enhancements, including revised core risk factors, reclassification criteria, improved risk communication, explicit consideration of pregnancy-related factors, and updated risk categories. Risk estimation is based on age, sex, systolic blood pressure, smoking status, diabetes status, total cholesterol, high-density lipoprotein cholesterol, and use of cardiovascular medications (blood pressure–lowering, lipid-modifying, and antithrombotic therapies), with optional inputs such as postcode, atrial fibrillation history, and type 2 diabetes–related variables to improve precision [28]. However, since the Aus CVD tool was developed for the general population, its performance may be suboptimal in specific subgroups. Evidence from a study conducted in Melbourne and Hobart among women aged 40–70 years indicates that the tool may underestimate the presence and severity of subclinical atherosclerosis, demonstrating limited concordance with coronary calcium score–defined atherosclerotic burden and reduced sensitivity for identifying individuals with significant underlying disease [77].

Limitation

This study has a limitation in incorporating some of the newly designed or updated CVD risk assessment tools. For instance, since the Australian CVD risk assessment tool was updated in 2023, no published studies using this tool have been identified, likely due to its recency. Given that this is a scoping review, risk of bias and quality assessment were used solely for reporting purposes, not for excluding studies. Moreover, this review has a limitation of focusing only on English-language written studies.

Conclusion

This scoping review underscores the need for CVD risk assessment tools that are specifically tailored for women with a history of GDM. The existing tools, such as FRS, ASCVD, and WHO/ISH, fall short in accurately capturing the unique risk features of these women, mainly because they do not incorporate pregnancy-related factors that can serve as strong indicators of future CVD. The undiagnosed CVD during pregnancy and the post-partum period further underscores the necessity for modified risk assessment tools and guidelines that consolidate the process of early identification of CVD. Moreover, the development of a population-specific tool is essential, as the existing models may not be effective across diverse populations with varying cardiometabolic risk profiles. Ultimately, the implementation of comprehensive CVD risk assessment during and after pregnancy using population-specific tools provides a valuable opportunity for early identification and intervention. This will potentially improve the long-term cardiovascular health outcomes for women with a history of GDM.

Author contributions

All authors contributed to the conceptual design and manuscript writing. The first draft of the manuscript was written by TEW, while PS and JM provided critical feedback, contributed to the interpretation of findings, and assisted in manuscript drafting and the revision process. All authors have read and approved the final manuscript.

Funding

The authors have no competing interests to declare, and there are no relevant financial or non-financial interests to disclose.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

As no primary data collection was conducted, no formal ethical approval or consent to participate is required under current institutional guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.