Pregnancy-associated plasma protein A in maternal serum for predicting early gestational diabetes mellitus: a systematic review and meta-analysis

Introduction

Gestational diabetes mellitus (GDM) is characterized by elevated blood glucose levels first detected during pregnancy, which are above the normal ranges but do not reach overt diabetes. Globally, around 14% of pregnant women suffer from GDM (Sweeting et al., 2024), which threatens the health of pregnant women and may lead to long-term adverse outcomes in fetuses and newborns (He, Yang & Wu, 2024). These consequences encompass large birth weight, neonatal hypoglycemia, breathing difficulties, and a higher likelihood of developing metabolic syndrome and type 2 diabetes (T2DM) later in life (Senthil Kumar, Mehboob & Lei, 2024). Approximately 60% of GDM women may develop T2DM within 5- to 10 years after childbirth (Borna et al., 2023). The standard approach to GDM diagnosis is an oral glucose tolerance test (OGTT) (American Diabetes Association Professional Practice Committee, 2024). According to the International Association of Diabetes and Pregnancy Study Groups (IADPSG), 75-gram anhydrous OGTT is recommended during 24 to 28 weeks of pregnancy. The established diagnostic criteria include fasting plasma glucose ≥ 5.1 mmol/L, 1-hour post-prandial plasma glucose ≥ 10.0 mmol/L, and 2-hour post-prandial plasma glucose ≥ 8.5 mmol/L (Metzger et al., 2010). Since this standard covers the diagnosis of various diseases and the evaluation of multiple physiological parameters, GDM diagnosis has become complex, which hinders rapid identification and diagnosis in clinical settings. Consequently, identifying effective indicators for rapid GDM diagnosis remains a significant challenge (Cui et al., 2023).

Pregnancy-associated plasma protein A (PAPP-A) is a glycoprotein synthesized by syncytiotrophoblast cells (Chen et al., 2025), with an approximate molecular mass of 200 kDa. This glycoprotein is primarily expressed in placental tissues and certain specialized cell types. It belongs to the metzincin superfamily of metalloproteinases and exhibits zinc-binding characteristics (Lovati et al., 2013). It mainly regulates the activity of insulin-like growth factor (IGF) and enhances endometrial receptivity, which is essential for the healthy development of the fetus (Visconti et al., 2019). Throughout pregnancy, PAPP-A levels rise gradually as gestational age advances (Ramezani et al., 2020). Recent research has increasingly recognized the potential diagnostic value of PAPP-A in GDM. However, there are notable discrepancies in its diagnostic performance. For instance, one study found that PAPP-A demonstrated strong diagnostic effectiveness, with an area under the curve (AUC) of 0.885 (Yildiz et al., 2023), whereas another study reported significantly lower accuracy, with an AUC of 0.479 (Visconti et al., 2019). Given these inconsistent findings, further research is required to determine the applicability and diagnostic precision of PAPP-A across different populations (Ren et al., 2020; Sovio et al., 2024).

Diagnostic meta-analysis is a reliable method for integrating data from various studies to precisely evaluate the sensitivity and specificity of diagnostic tests (Bossuyt et al., 2003). This approach effectively resolves discrepancies among research findings. In this study, we aim to integrate data to evaluate the utility of PAPP-A in the early diagnosis of GDM, thereby enhancing early detection and intervention strategies for GDM. This will provide evidence-supported guidance for clinical practice and offer evidence-based insights for healthcare professionals and public health policymakers while identifying potential avenues for future research.

Materials and Methods

Results

After database searches, we initially retrieved 3,587 relevant articles. After eliminating duplicates and thoroughly assessing the titles and abstracts, we selected 39 articles for full-text evaluation. During this process, 19 articles were ruled out because they failed to meet our inclusion criteria. Finally, 20 studies were enrolled in the systematic review. Due to inadequate diagnostic outcome data, eight studies were later removed. As a result, 12 studies were incorporated into the meta-analysis (Lovati et al., 2013; Syngelaki et al., 2015; Talasaz et al., 2018; Visconti et al., 2019; Ramezani et al., 2020; Ren et al., 2020; Kapustin et al., 2022; Borna et al., 2023; Cui et al., 2023; American Diabetes Association Professional Practice Committee, 2024; Amini et al., 2024; Lu et al., 2024). The research screening process is shown in the PRISMA flowchart (Fig. 1).

This research encompassed four retrospective cohort studies, four prospective cohort studies, one cross-sectional study, and three case-control studies. These studies involved 25,183 adult participants from various countries in Asia and Europe, including China, Iran, Israel, Italy, Bulgaria, and Turkey, with sample sizes varying between 99 and 7,936 individuals. The detailed study features and diagnostic outcomes are presented in Table 1. Among these studies, three provided specific PAPP-A threshold values: 4.95 IU/L (international units per liter), 16.34 ng/L (nanograms per liter), and 1.39 mg/L (milligrams per liter). The other nine studies reported PAPP-A cutoffs using median multiples of the median (MoM), ranging from 0.32 MoM to 2.21 MoM, as outlined in Table 1.

Quality assessment

All studies were rated as either “low risk” or “unclear”. Regarding patient selection, five studies were classified as “unclear”, whereas seven were deemed to have “low risk”. For the process and timing, all articles had “low risk of bias”. For the indicator test, most studies were rated as “low bias risk”. Despite potential biases in the design and execution of these studies, the overall bias risk remained relatively low (Fig. 2).

Pooled effect size

This meta-analysis evaluated the diagnostic accuracy of PAPP-A for gestational diabetes mellitus (GDM), revealing that its overall diagnostic capability was limited, with an area under the curve (AUC) of 0.70 (95% CI [0.65–0.73]). The sensitivity was 0.71 (95% CI [0.60–0.80]), and the specificity was 0.62 (95% CI [0.55–0.68]). Subgroup analysis indicated that studies with smaller sample sizes (<1,000) demonstrated better diagnostic performance (AUC = 0.76, 95% CI [0.72–0.80]) than those with larger sample sizes. Additionally, the time-resolved fluorescence immunoassay (TRFIA) method exhibited superior performance (AUC = 0.77, 95% CI [0.73–0.80]) to non-TRFIA methods. The pooled sensitivity was 0.71 (95% CI [0.60–0.80]), specificity was 0.62 (95% CI [0.55–0.68]), AUC was 0.70 (95% CI [0.65–0.73]), positive likelihood ratio (PLR) was 1.9 (95% CI [1.4–2.5]), negative likelihood ratio (NLR) was 0.47 (95% CI [0.30–0.72]), and diagnostic odds ratio (DOR) was 4 (95% CI [2–8]). Significant heterogeneity was observed in the results, with I² values exceeding 90%. These findings highlight the need for further research to establish standardized protocols and enhance the diagnostic utility of PAPP-A for GDM (Table 2, Figs. 3 and 4).

Table 2:

PAPP-A diagnostic analysis for GDM: overall and subgroup assessments.

Population	Studies (n)	Sensitivity	Specificity	AUSROC	LR+	LR-	dOR	Publication bias (p-value)
Overall	12	0.71 [0.60, 0.80]	0.62 [0.55, 0.68]	0.70 [0.65, 0.73]	1.9 [1.4, 2.5]	0.47 [0.30, 0.72]	4 [2, 8]	0.400
subgroup
sample size >1,000	5	0.66 [0.46, 0.82]	0.56 [0.47, 0.64]	0.61 [0.57, 0.65]	1.5 [0.9, 2.4]	0.61 [0.31, 1.23]	2 [1, 8]
sample size <1,000	7	0.74 [0.64, 0.81]	0.68 [0.59, 0.75]	0.76 [0.72, 0.80]	2.3 [1.6, 3.3]	0.39 [0.25, 0.60]	6 [3, 13]
TRFIA	6	0.69 [0.53, 0.82]	0.58 [0.51, 0.64]	0.64 [0.60, 0.68]	1.6 [1.1, 2.4]	0.53 [0.30, 0.96]	3 [1, 8]
Non TRFIA	6	0.71 [0.60, 0.80]	0.62 [0.55, 0.68]	0.70 [0.65, 0.73]	1.9 [1.4, 2.5]	0.47 [0.30, 0.72]	4 [2, 8]
Published ≤2020	5	0.67 [0.57, 0.75]	0.65 [0.53, 0.75]	0.71–0.67, 0.75]	1.9 [1.2, 2.9]	0.51 [0.34, 0.77]	4 [2, 9]
Published >2020	7	0.73 [0.55, 0.86]	0.60 [0.52, 0.68]	0.68 [0.63, 0.72]	1.8 [1.2, 2.8]	0.45 [0.22, 0.92]	4 [1,13]
case–control study	4	0.63 [0.54, 0.70]	0.60 [0.46, 0.72]	0.65 [0.61, 0.69]	1.6 [1.0, 2.4]	0.63 [0.43, 0.92]	2 [1, 6]
Non case–control study	8	0.75 [0.60, 0.85]	0.63 [0.55, 0.71]	0.72 [0.68, 0.76]	2.0 [1.4, 3.0]	0.40 [0.22, 0.74]	5 [2, 14]

DOI: 10.7717/peerj.19825/table-2

Notes:

Abbreviations

AUSROC

Area under summary receiver operating characteristic curve

CI

confidence interval

dOR

diagnostic odds ratio

LR+

positive likelihood ratio

LR−

negative likelihood ratio

PAPP-A

pregnancy-associated plasma protein-A

TRFIA

time-resolved fluorescence immunoassay

Subgroup analysis

Subgroup analyses were implemented based on publication year, sample sizes, gestational weeks, study design, cut-off, and testing method. The diagnostic threshold was evaluated for each subgroup using PAPP-A levels. Our results suggested that the variability in sample sizes might significantly contribute to the heterogeneity observed in specificity (p < 0.01) (Fig. 5).

Publication bias assessment

Deek’s funnel plot illustrated that the points representing individual studies in the population exhibited an approximately symmetrical distribution. Deek’s test suggested no publication bias (p = 0.400) (Fig. 6).

Discussion

In recent years, increasing research has examined the link between PAPP-A and GDM, but the results across studies remain inconsistent. For instance, one study reported a relatively high diagnostic efficacy for GDM with an AUC of 0.82 (Amini et al., 2024), whereas another study concluded a lower efficacy with an AUC of 0.542 (Borna et al., 2023). The present research conducted a comprehensive review and meta-analysis to thoroughly examine data from various studies, thereby offering a more reliable and unbiased assessment. The results indicated that PAPP-A exhibited diagnostic value in GDM risk among pregnant women across different geographical regions, albeit with certain limitations. Specifically, the pooled sensitivity was 0.71 (95% CI [0.60–0.80]), specificity was 0.62 (95% CI [0.55–0.68]), DOR was 4 (95% CI [2–8]), and AUC was 0.70. Although PAPP-A is promising as a diagnostic indicator for GDM, its precision needs further improvement.

The PLR of PAPP-A was 1.9 (95% CI [1.4–2.5]), indicating that a positive PAPP-A result is 1.9 times more likely in pregnant women with GDM than those without GDM, highlighting its role in identifying high-risk pregnant women and prompting further diagnostic testing. The NLR was 0.47 (95% CI [0.30–0.72]), less than 1, suggesting a negative PAPP-A result can effectively rule out GDM, helping to minimize unnecessary diagnostic procedures for low-risk individuals. While PAPP-A alone has moderate diagnostic accuracy (AUC = 0.67), its PLR and NLR offer meaningful support for GDM screening.

In recent years, the worldwide increasing incidence of obesity and changes in lifestyle patterns have resulted in a significant rise in GDM incidence (Eades, Cameron & Evans, 2017), which differs considerably across the globe, largely influenced by the specific diagnostic standards applied. The Carpenter and Coustan or National Diabetes Data Group criteria state a prevalence from 2.2% to 37.9%, while the IADPSG criteria indicate a range of 3.5% to 38.6% (Bilous et al., 2021; Wang et al., 2024). Currently, the OGTT is considered the gold standard for screening and diagnosing GDM (Zeck et al., 2007). However, this method has limitations, including complex procedures, time consumption, and potential burdens on pregnant women. Consequently, identifying more convenient and accurate biomarkers is significant for early detection and intervention of GDM. Recent research has explored various potential biomarkers, such as PAPP-A and placental IGF (Leeson, Odendaal & Quenby, 2023; Nurdiati et al., 2024).

Since PAPP-A was initially detected in maternal serum, the research discovered that this enzyme (Conover & Oxvig, 2023), which required zinc for its activity, was crucial in cleaving specific IGFBPs (Conover et al., 2004). It primarily targets IGFBP-4, leading to enhanced levels of free IGF-I available for biological behaviors (Giudice et al., 2002; Yu et al., 2025). PAPP-A regulates IGF bioavailability by proteolytically hydrolyzing IGFBP-2, -4, and -5 (Fagali Franchi et al., 2024). The physiological state of maternal adipose tissue and systemic glucose regulation influence IGF availability (Sessions-Bresnahan, Heuberger & Carnevale, 2018). Reduced expression of PAPP-A results in diminished hydrolysis of IGFBPs, consequently decreasing IGF bioavailability. This mechanism is more pronounced during pregnancy to satisfy the increased requirement for bioactive IGF (Gude et al., 2024; Harboe et al., 2024). Research revealed that serum PAPP-A levels were positively correlated with glycated hemoglobin concentrations and homeostatic model assessment for insulin resistance (HOMA-IR) in GDM patients (Ku et al., 2024; Choi, Duan & Bai, 2025). This suggests that PAPP-A could elevate blood glucose levels by influencing the insulin signaling pathway. Although PAPP-A shows promise for aiding in GDM diagnosis, it is not yet sufficiently effective to replace the current standard method, the OGTT. In this study, the AUC value of 0.70 indicated that PAPP-A had a relatively modest diagnostic accuracy, which may compromise its reliability as a sole clinical diagnostic marker. Additionally, specialized equipment and technical expertise are required for PAPP-A measurement, which limits its broad application in primary care settings. Therefore, PAPP-A might be more suited for preliminary screening in high-risk populations rather than serving as a routine diagnostic tool. Subgroup analysis indicated that sample size was likely the main source of heterogeneity (p < 0.05). Hence, future studies with larger sample sizes are required to improve the stability and reliability of our results. The QUADAS-2 tool demonstrated an overall high quality. Under the PRISMA guidelines, this meta-analysis significantly enhanced the reliability of the pooled results, surpassing that of individual studies. However, several limitations were noted. First, the search was confined to English-language databases. Second, the limited number of included studies (n = 12) could lead to an insufficient sample size and potential bias. Third, variations in sample sources and population characteristics across studies may introduce bias. Finally, the cut-off value used for identifying GDM was inconsistent.

These limitations should be tackled in future research to substantiate the utility of PAPP-A in GDM diagnosis.

Conclusion

This study evaluates the diagnostic accuracy of PAPP-A in GDM at an early stage. The results indicate that PAPP-A has limited capacity to predict individuals at risk for GDM. The current diagnostic effectiveness of PAPP-A alone is insufficient for clinical application. Future research should improve detection methods and establish standardized critical thresholds for PAPP-A. Additionally, the potential value of combining PAPP-A with other biomarkers to enhance GDM diagnostic performance warrants further exploration. Based on our findings, PAPP-A should not be used as a standalone screening or diagnostic test for GDM at present.

Supplemental Information