Impact of low-dose aspirin exposure on obstetrical outcomes: a meta-analysis

Abstract

Objective

To assess the impact of low-dose aspirin (LDA) on obstetrical outcomes through a meta-analysis of placebo-controlled randomized controlled trials (RCTs).

Methods

A systematic search of the PubMed, Cochrane Library, Web of Science and Embase databases from inception to January 2024 was conducted to identify studies exploring the role of aspirin on pregnancy, reporting obstetrical-related outcomes, including preterm birth (PTB, gestational age <37 weeks), small for gestational age (SGA), low birth weight (LBW, birthweight < 2500g), perinatal death (PND), admission to the neonatal intensive care unit (NICU), 5-min Apgar score < 7 and placental abruption. Relative risks (RRs) were estimated for the combined outcomes. Subgroup analyses were performed by risk for preeclampsia (PE), LDA dosage (<100 mg vs. ≥100 mg) and timing of onset (≤20 weeks vs. >20 weeks).

Results

Forty-seven studies involving 59,124 participants were included. Compared with placebo, LDA had a more significant effect on low-risk events such as SGA, PTB and LBW. Specifically, LDA significantly reduced the risk of SGA (RR = 0.91, 95% CI: 0.87–0.95), PTB (RR = 0.93, 95% CI: 0.89–0.97) and LBW (RR = 0.94, 95% CI: 0.89–0.99). For high-risk events, LDA significantly lowered the risk of NICU admission (RR = 0.93, 95% CI: 0.87–0.99). On the other hand, LDA can significantly increase the risk of placental abruption (RR = 1.72, 95% CI: 1.23–2.43). Subgroup analyses showed that LDA significantly reduced the risk of SGA (RR = 0.86, 95% CI: 0.77–0.97), PTB (RR = 0.93, 95% CI: 0.88–0.98) and PND (RR = 0.65, 95% CI: 0.48–0.88) in pregnant women at high risk of PE, whereas in healthy pregnant women LDA did not significantly improve obstetrical outcomes, but instead significantly increased the risk of placental abruption (RR = 5.56, 95% CI: 1.92–16.11). In pregnant women at high risk of PE, LDA administered at doses ≥100 mg significantly reduced the risk of SGA (RR = 0.77, 95% CI: 0.66–0.91) and PTB (RR = 0.56, 95% CI: 0.32–0.97), but did not have a statistically significant effect on reducing the risk of NICU, PND and LBW. LDA started at ≤20 weeks significantly reduced the risk of SGA (RR = 0.76, 95% CI: 0.65–0.89) and PTB (RR = 0.56, 95% CI: 0.32–0.97).

Conclusions

To sum up, LDA significantly improved neonatal outcomes in pregnant women at high risk of PE without elevating the risk of placental abruption. These findings support LDA’s clinical application in pregnant women, although further research is needed to refine dosage and timing recommendations.

Keywords:

• Aspirin
• low-dose
• pregnancy
• obstetrical outcomes
• meta-analysis

Introduction

Aspirin, a cost-effective and widely available non-steroidal anti-inflammatory drug (NSAID), plays a critical role in mitigating the risk of preeclampsia (PE) in pregnant women [1]. Its mechanism involves the inhibition of cyclooxygenase isozymes (COX-1 and COX-2). At low doses (60–150 mg/day), aspirin selectively acetylates COX-1, reducing thromboxane A2 (TXA2) synthesis in platelets. The placenta in PE may have excessive vasoconstriction caused by TXA2 overexpression [2]. So, aspirin’s TXA2-mediated platelet aggregation inhibitory characteristics offset poor placental perfusion, including placental vasoconstriction and blood clotting, hence lowering hypertension disorders and the risk of PE in pregnancy [3]. Therefore, following the American College of Obstetricians and Gynecologists (ACOG) guidelines, low-dose aspirin (LDA) is recommended from the late first trimester for women with a history of severe PE or multiple pregnancies affected by the condition [4].

It has been observed that aspirin can traverse the placental barrier and enter fetal circulation at modest dosages [5–7], potentially exerting an impact on the infant. However, the precise influence of LDA administration during pregnancy on obstetrical outcomes remains uncertain. One study demonstrated a significant reduction in perinatal mortality with LDA usage [8]. Another study indicated a decreased risk of preterm delivery with LDA treatment, although this difference between groups did not reach statistical significance [4]. Conversely, a publication in The Lancet reported no substantial effect of LDA on the risk of fetal or infant death [9]. Furthermore, evidence suggests that while LDA reduces PE incidence, it may elevate the risk of placental abruption [8]; however, other studies have shown no association between LDA and placental abruption [10–13], yielding inconsistent results.

Consequently, the objective of this meta-analysis was to assess the impact of exposure to LDA during pregnancy on obstetrical outcomes, aiming to ascertain the broader clinical justification for aspirin use. Furthermore, subgroup analyses were conducted in this study to investigate the optimal dosage and timing for initiating aspirin therapy, as well as the influence of high-risk PE status on obstetric outcomes.

Methods

The review was prospectively registered on PROSPERO (ID: CRD42023430911). To assess the impact of LDA on obstetrical outcomes, we performed a systematic review and meta-analysis of placebo randomized controlled trials (RCTs).

Search strategy and selection criteria

A systematic literature search was conducted across PubMed, Cochrane Library, Web of Science, Embase and Google Scholar from inception to January 2024. The search formula used was “(acetylsalicylic acid OR aspirin) AND (pregnant OR pregnancy OR preeclampsia OR preterm OR pregnancy complications)”, with medical subject headings (MeSH) and text words including “pregnancy”, “children”, “aspirin” and “acetylsalicylic acid”.

Our selection criteria for trials aimed to assess the impact of LDA exposure during pregnancy on obstetrical outcomes, guided by the PICOS framework [14]. The inclusion criteria were as follows: Population (P) – pregnant women with aspirin administered at dosages below 150 mg; Intervention (I) – any dose of aspirin less than 150 mg at any time during pregnancy; Comparators (C) – placebo dose and duration matched to aspirin received by pregnant women; Outcome measures (O) – outcomes reflecting poor obstetrical outcomes: preterm birth (PTB, gestational age <37 weeks), small for gestational age (SGA), perinatal death (PND), low birth weight (LBW) < 2500 g (LBW), admission to the neonatal intensive care unit (NICU), placental abruption, 5-min Apgar score < 7; study design (S) – (RCTs published in English).

We excluded studies that: (1) lacked data on key obstetrical outcomes; (2) published in languages other than English; (3) focused on other antiplatelet agents or NSAIDs; and (4) represented duplicate publications, for which we selected the report with the most comprehensive dataset. Studies with multiple arms, including aspirin, placebo and other medicines, were still included if their two experimental groups were aspirin and placebo, with the exclusion of the third.

Identification of relevant studies and data extraction

The title and abstract were screened by two reviewers (Xiaoyan Lin, Jingchao Yong), and if the study appeared to be possibly eligible, the full text was retrieved and screened by two reviewers. Discrepancies between reviewers led to adjudication by a third reviewer. Data from studies meeting inclusion criteria were systematically extracted into a spreadsheet using a standardized form. This process was independently performed by the initial reviewers, with any differences resolved by a third reviewer (Jiangbo Du). The accuracy of the data extraction was assessed on separate occasions. The following data were extracted: author, year of publication, time of study, original inclusion criteria, total number of patients included in the study, aspirin dose and duration of application. Relevant studies were also identified through the reference lists of reviewed articles and additional sources.

Statistical analyses

R (version 4.3.2) and RStudio were used to perform statistical analyses. Pairwise dichotomous outcomes were analyzed, presenting relative risks (RRs) and 95% confidence intervals (95% CIs). Risk ratios and summary RR were plotted as forest plots, while trial heterogeneity was evaluated using the I² statistic, considering I² > 50% as high heterogeneity [15], applying the DerSimonian and Laird random effects model; otherwise, the Mantel–Haenszel fixed effects model was used. Sensitivity analyses were also carried out to test the reliability of the results of the meta-analysis.

Subgroup analyses

We divided the pregnant women into subgroups based on their risk for PE. A pregnant woman was defined as being at high risk for PE by any of the following criteria: insulin-treated gestational diabetes, chronic hypertension, multiple pregnancies or a history of PE in prior pregnancies [16]. And a healthy pregnant woman is defined as: pregnant women without chronic hypertension, renal disease, diabetes mellitus or other medical illnesses [17].

Following this, we combined subgroup analyses of pregnant women at high risk for PE with aspirin dosage (<100 mg vs. ≥100 mg) and LDA onset (≤20 weeks vs. >20 weeks). This criterion was selected since a prior meta-analysis revealed that pregnant women taking aspirin 100–150 mg before 20 weeks of gestation were successful in lowering the risk of PE [18].

Outcomes

The main outcomes were as follows: low-risk outcome (PTB, SGA and LBW), high-risk outcome (PND, admission to the NICU, and 5-min Apgar score < 7) and placental outcome (placental abruption). These outcomes were chosen because they responded to poor obstetrical outcomes, and unless otherwise stated, outcomes were defined by the authors, with definitions of the primary outcomes listed in the study characteristics (Table 1).

Table 1. Characteristics of the included 47 randomized controlled trials reporting the impact of low-dose aspirin on pregnancy outcome.

Study ID	Inclusion criteria	Patients recruited (n)			Intervention		Outcomes of interest	GA (aspirin)
		Total	Aspirin	Placebo	Aspirin dose (mg)	GA commenced (weeks)
Wallenburg et al. [19]	Angiotensin-II-sensitive nulliparity	44	21	23	60	28	PTB, SGA	40
Benigni et al. [20]	Hypertension or previous obstetrical history	33	17	16	60	12	PND, SGA	39 ± 2
Schiff et al. [21]	Nulliparity, positive roll-over test	65	34	31	100	28–30	SGA, PTB, 5-min Apgar score <7, NICU	272 ± 8 (days)
McParland et al. [22]	Nulliparity, high risk of pregnancy-induced hypertension	100	48	52	75	24	LBW, PND, VLBW	39.5
Schröcksnadel et al. [23]	Primigravidae with positive roll-over test	41	22	19	80	28–32	PTB, HBW, SGA	None
Sibai et al. [17]	Nulliparity	3135	1570	1565	60	13–25	SGA, PND, NICU, PTB, HBW, placental abruption	38.6 ± 3.3
Viinikka et al. [24]	Women with preexisting hypertension or a history of severe preeclampsia in their previous pregnancy	197	97	100	50	12–18	NICU, PND	38.6 ± 2.1
Davies et al. [25]	Nulliparous women with a high hemoglobin concentration in the second trimester	118	58	60	75	18	NICU, placental abruption	39.9
Kyle et al. [26]	Angiotensin-II-sensitive nulliparity	80	44	36	60	28	SGA, PND	278.3 (days)
Newnham et al. [27]	FGR and abnormal UtA Doppler	51	25	26	100	30–33	SGA, NICU, 5-min Apgar score <7	35.7
Bower et al. [28]	Abnormal UtA Doppler	60	31	29	60	24	IUGR, PTB	None
Morris et al. [29]	Nulliparity, abnormal UtA Doppler	102	52	50	100	18	SGA, PTB	None
Wang and Li [30]	Pregnant women at high risk of IUGR	84	40	44	75	28–30	IUGR, PTB, PND	38.6 ± 1.5
Bar et al. [31]	History of PE, FGR, chronic HTN, renal disease, MA, multiple pregnancy or current PE or FGR	87	41	46	60	12–24	IUGR, NICU	38 ± 3
Bar et al. [32]	History of PE, FGR, chronic HTN, renal disease, MA, multiple pregnancy or current PE or FGR	53	25	28	60	12–32	IUGR, NICU	38 ± 3
Gallery et al. [33]	Pregnant women with preexisting hypertension or renal disease, or a history of early severe preeclampsia in a previous pregnancy	108	58	50	100	17–19	PTB, LBW	39
Byaruhanga et al. [34]	Pregnant women with either a previous history of pregnancy-induced hypertension or preexisting chronic hypertension	230	113	117	75	20–28	IUGR, PTB, PND, NICU	37.6
Caritis et al. [16]	Pregnant women with pregestational insulin-treated diabetes mellitus, chronic hypertension, multifetal gestations, and women who had had preeclampsia during a previous pregnancy.	2503	1254	1249	60	13–26	Placental abruption, PTB, SGA, PND	None
Rotchell et al. [35]	All women	3641	1819	1822	75	12–32	PTB, PND, LBW, NICU, placental abruption, miscarriage	None
McCowan et al. [36]	Singleton pregnancy with ultrasound evidence of a small for gestational age fetus (abdominal circumference <10%); previous anatomy scan c 20 weeks and no evidence of fetal abnormality; gestation between 24 and 36 weeks; umbilical artery Doppler resistance index >95% for gestation; no previous aspirin treatment in pregnancy; and no contra-indication to aspirin treatment.	99	49	50	100	24–36	SGA, NICU, PND	34.4
Pattison et al. [37]	Women with a history of recurrent miscarriages (≥3) and antiphospholipid antibodies	40	20	20	75	6–12	PTB, SGA, PND	None
Subtil et al. [38]	Nulliparity	3274	1634	1640	100	14–20	SGA, PTB, PND, HBW, NICU	39.3
Yu et al. [39]	Abnormal UtA Doppler at 23 weeks	554	276	278	150	22–24	Placental abruption, PTB, NICU, PND	None
Kaandorp et al. [40]	Women with a history of unexplained recurrent miscarriage	241	120	121	80	6	SGA, PTB, miscarriage	39.5 ± 1.3
Abramovici et al. [41]	Women at high risk (prior preeclampsia, chronic hypertension, diabetes mellitus or multifetal pregnancy)	2500	1252	1248	60	None	NICU, PTB, LBW	None
Ayala et al. [42]	Family history or previous history of PE or PIH, chronic HTN, CVS or endocrine or bleeding or metabolic disorder, history of miscarriage, multiple pregnancy, obesity, MA <18 or >35	350	176	174	100	12–16	PTB, IUGR, PND	39.2 ± 1.9
Ohno et al. [43]	Low- and high-risk women	4945	2480	2465	60	13–25	Placental abruption	38
Schisterman et al. [44]	Women who had one or two pregnancy losses	1228	615	613	81	None	PTB, miscarriage	38.8 ± 2.0
Abramovici et al. [45]	Low- and high-risk women	2978	1483	1495	60	13–25	PTB, LBW, SGA, NICU	None
Moore et al. [46]	Women at high risk for preeclampsia	523	265	258	60	13–26	SGA	None
Allshouse et al. [47]	Pregnancies with at least one risk factor for preeclampsia: preexisting insulin-dependent diabetes (IDDM), HTN, multiple gestation or preeclampsia in a previous pregnancy (PrePreE)	1789	903	886	60	13–26	PTB	37
Euser et al. [48]	Women with at least one risk factor for preeclampsia (multiple gestation, chronic hypertension, insulin-dependent diabetes or preeclampsia in a prior pregnancy)	225	106	119	None	12–26	SGA	34.6
Sentilhes et al. [49]	Nulliparity, at high risk for preterm preeclampsia	1620	798	822	150	11–14	PTB, PND, SGA, NICU	39
Scazzocchio et al. [50]	Nulliparity, abnormal UtA Doppler	155	80	75	150	11–14	FGR, SGA, NICU	39.4 ± 1.6
Andrikopoulou et al. [51]	Nulliparity with absence of hypertension, renal disease, diabetes, other endocrine disorders, seizures, heart disease or collagen vascular disease.	2543	1262	1281	60	13–25	PTB, placental abruption	None
Blomqvist et al. [52]	Women with at least three consecutive miscarriages in the first trimester within the couple	400	200	200	75	24	PTB, NICU, SGA, 5-min Apgar score <7, miscarriage	None
Sutton et al. [53]	Nulliparous, low-risk women	2947	1465	1482	None	13–25	SGA, PTB, NICU, placental abruption	38.1 ± 4.1
Tolcher et al. [54]	Women with pregestational insulin-treated diabetes mellitus, chronic hypertension, multiple gestations or a history of preeclampsia in a previous pregnancy	2539	1254	1249	60	13–26	PTB, SGA, PND, placental abruption	36.2
Hauspurg et al. [55]	Women at high-risk for preeclampsia	1020	512	508	60	13–26	PTB, NICU, SGA, placental abruption	None
Gu et al. [56]	Pregnant women at high risk for preeclampsia	1105	821	284	25, 50, 75	12	IUGR, placental abruption	None
Handal-Orefice et al. [57]	Nulliparous, obese women (pre-pregnancy BMI > 30) with singleton, non-anomalous pregnancies	258	124	134	None	None	Placental abruption, PTB, NICU	39.1 ± 2.8
Hoffman et al. [58]	Nulliparity	11,558	5787	5771	81	6–13	PTB, SGA, PND, LBW, VLBW, miscarriage	None
Tolcher et al. [59]	Low- and high-risk women	2985	1485	1500	60	13–26	PTB, SGA, PND, placental abruption	38.6
Xiang et al. [60]	Females with mild to moderate chronic hypertension in the first trimester (6–10 weeks) and systemic blood pressure between 140 and 159 mmHg or diastolic blood pressure between 90 and 109 mmHg without medication and disease in the target organ	255	129	126	None	12	Placental abruption, SGA, PTB, NICU, PND	None
Ling et al. [61]	High risk women	428	208	220	150	11–13	PTB, SGA, PND	39.3
Bauserman et al. [62]	Nulliparous women	1446	724	722	81	6–13	PTB, SGA, PND, LBW	None
Landman et al. [63]	Women with a singleton pregnancy and a history of spontaneous preterm birth of a singleton between 22 and 37 weeks	387	194	193	80	8–16	PTB, SGA, PND, NICU	38

PTB: gestational age <37 weeks; SGA: small for gestational age; PND: perinatal death; LBW: birthweight < 2500 g; NICU: neonatal intensive care unit.

Only first author of each study is given.

Study quality and risk of bias assessment

Two reviewers (Xiaoyan Lin, Jingchao Yong) assessed the risk of bias in randomized trials using the Cochrane Risk of Bias 2.0 tool [64], and any discrepancies were settled by a third reviewer. Our evaluation consists of six domains of bias: selection bias, performance bias, detection bias, attrition bias, reporting bias and other bias. We judged the bias of the included studies as high, low or unclear. The assessment of risk of bias is summarized in Figure S1. Publication bias and symmetry were evaluated using funnel plots and the Egger test [65]. A p value greater than .05 and symmetric funnel plots suggest a low risk of publication bias. The assessment results for publication bias are presented in Figure S2.

Meta-regression analysis and cumulative meta-analysis

To investigate potential sources of heterogeneity, we conducted univariate meta-regression analyses on the methodological and clinical analysis variables of interest [66]. Additionally, for illustrative purposes, we generated weighted bubble plots with fitted meta-regression lines. Furthermore, to assess the cumulative trend of research findings over time and evaluate the impact of each study on the composite results, we performed a repeated meta-analysis by progressively incorporating studies based on their publication year.

Results

Literature review

The methodology employed for literature selection in this study is summarized in Figure 1. A comprehensive search was conducted across multiple databases including PubMed, Cochrane Library, Web of Science and Embase, resulting in the identification of a total of 11,479 publications. After excluding duplicates and non-RCTs, 747 papers were deemed potentially eligible for detailed evaluation. Further exclusions were made based on criteria such as review articles, commentaries, protocols, non-English papers without full text availability and studies unrelated to the design and evaluation of alternative treatments (e.g. other antiplatelet drugs and NSAIDs). This rigorous screening process yielded a final set of 98 studies that underwent thorough review. Following exclusion criteria related to irrelevant study outcomes (n = 15), inability to extract required data from reported sources (n = 4), absence of placebo-control groups (n = 14) or inclusion of data subsets from other included studies (n = 18), a total of 47 studies [16,17,19–63] involving a combined participant pool comprising 59,124 individuals were ultimately included in this meta-analysis.

Figure 1. Study selection diagram.

The characteristics of the RCTs mentioned in this article are summarized in Table 1. Studies recruited pregnant women with different risk factors: two of them recruited women with angiotensin II sensitivity [19,26], two studies recruited women with positive roll-over tests [21,23], five studies recruited women with abnormal Doppler ultrasound [27–29,39,50], seven studies recruited nulliparous women [17,38,51,53,57,58,62], one study recruited women with high hemoglobin concentrations in mid-pregnancy [25], five studies recruited women with a history of miscarriage or premature delivery [37,40,44,52,63], 23 studies recruited women at risk of hypertension or PE in pregnancy [16,17,20,22,31–34,36,41–43,45–50,54–56,59–61], one study enrolled women at risk for IUGR [30] and one study enrolled all women [35]. In addition, except for four studies that were ambiguous [48,53,57,60], 34 studies used aspirin <100 mg [16,17,19,20,22–26,28,30–32,34,35,37,40,41,43–47,51,52,54–56,58,59,62,63] and 11 studies used aspirin ≥100 mg [21,27,29,33,36,38,39,42,49,50,61]. Except for 18 studies with ambiguous or long onset windows from early to late pregnancy [16,17,31,32,35,41,43–48,51,53–55,57,59], 12 studies started LDA at >20 weeks of gestation [19,21–23,26–28,30,33,34,36,39] and 19 studies started LDA at ≤20 weeks of gestation [20,24,25,29,33,37,38,40,42,49,50,56,58,60–63].

Two outcomes, namely PTB defined as gestational age <37 weeks and LBW defined as birthweight <2500 g, exhibited potential publication bias. However, the presence of publication bias was assessed to be generally acceptable for all other outcomes based on passing the funnel plot and Egger’s test (p > .05).

Outcome measure analyses

The obstetrical outcomes associated with LDA treatment are summarized in Figure 2. The study revealed a significant reduction in the risk of adverse events related to LBW, PTB and infants born with a birthweight less than 2500 g (LBW) due to LDA therapy. a total of 35 studies [16,17,19–21,23,26–32,34,36–38,40,42,45,46,48–50,52–56,58–63] involving 40,857 participants were included for the outcome of SGA. Overall findings demonstrated that compared to placebo-assigned controls, LDA therapy significantly reduced the risk of SGA (RR = 0.91, 95% CI: 0.87, 0.95). Additionally, a total of 26 studies involving 50,742 participants [16,17,33–35,38–42,44,45,47,49,51–55,57–63] were included for PTB analysis which showed that LDA therapy significantly decreased the risk of PTB compared to placebo-assigned controls (RR = 0.93, 95% CI: 0.89, 0.97). For LBW outcome analysis, a total of seven studies [22,33,35,41,45,58,62] involving 22,034 participants were included, and it was observed that LDA therapy slightly reduced the risk of LBW when compared to placebo-assigned controls (RR = 0.94, 95% CI: 0.89, 0.99).

Figure 2.  Forest plots of low-dose aspirin effects on obstetrical outcomes: (a) PTB, (b) SGA, (c) placental abruption, (d) 5-min Apgar score < 7, (e) LBW, (f) NICU admission and (g) PND.

However, LDA demonstrated limited efficacy in reducing the incidence of high-risk events such as admission to the NICU, PND and 5-min Apgar scores <7. For NICU admissions, a total of 22 studies involving 23,819 participants were included [17,21,24,25,27,31,32,34–36,38,39,41,45,49,50,52,53,55,57,60,62,63]. Compared to placebo-assigned controls, LDA therapy showed a slight reduction in the risk of NICU admission (RR = 0.93, 95% CI, 0.87, 0.99). Regarding PND outcomes, a total of 22 studies [16,17,20,22,24,26,30,34–39,42,49,54,58–63] involving 35,446 participants were included. No significant difference was observed between women assigned to the LDA treatment group and those assigned to the control group receiving placebo (RR = 0.89, 95% CI, 0.78, 1.0). For the outcome of 5-min Apgar score < 7, a total of three studies involving 462 participants were included [21,27,52]. No significant difference was found between women randomly assigned to receive LDA treatment and those assigned to receive placebo (RR = 0.35, 95% CI, 0.09, 1.26).

A total of 13 studies [16,17,25,39,43,51,53–57,59,60] involving a sample size of 22,436 participants were included to assess the impact of LDA treatment on placental abruption. The findings revealed a significant increase in the risk of placental abruption among individuals receiving LDA treatment compared to those assigned randomly to a placebo control group (RR = 1.72, 95% CI: 1.23, 2.43).

Subgroup analyses

The subgroup analysis of pregnant women at high risk for PE is summarized in Table 2. In contrast to aspirin treatment in healthy pregnant women, the administration of aspirin demonstrated a statistically significant reduction in the risk of small-for-gestational-age (SGA) infants (RR = 0.86, 95% CI: 0.77–0.97), PTB (RR = 0.93, 95% CI: 0.88–0.98) and PND (RR = 0.65, 95% CI: 0.48–0.88) among pregnant women with a high risk for PE. Additionally, LDA was found to increase the risk of placental abruption in healthy pregnant women (RR = 5.56, 95% CI: 1.92–16.11), while no statistically significant effect was observed in pregnancies with a high risk for PE (RR = 0.53, 95% CI, 0.21, 1.33).

Table 2. Subgroup analysis (women at high risk for preeclampsia).

Outcome	Trials (n)	Participants (n)	Aspirin (n)	Controls (n)	RR (95% CI)	I^2 (%)
SGA
Women at high risk for preeclampsia	12	9808	4886	4922	0.86 [0.77; 0.97]	20
Healthy pregnant women	5	21,584	10,791	10,793	0.96 [0.91,1.01]	16
PTB
Women at high risk for preeclampsia	11	13,306	6657	6649	0.93 [0.88; 0.98]	49
Healthy pregnant women	5	22,191	11,086	11,105	0.92 [0.85; 1.00]	9
PND
Women at high risk for preeclampsia	10	8225	4095	4130	0.65 [0.48; 0.88]	0
Healthy pregnant women	4	19,317	9653	9664	0.90 [0.78; 1.04]	15
NICU
Women at high risk for preeclampsia	8	5968	2968	3000	0.94 [0.86; 1.02]	2
Healthy pregnant women	3	9276	4615	4661	0.96 [0.84; 1.10]	0
LBW
Women at high risk for preeclampsia	3	2708	1358	1350	0.96 [0.87; 1.06]	4
Healthy pregnant women	2	12,675	6343	6332	0.93 [0.87; 1.00]	0
Placental abruption
Women at high risk for preeclampsia	2	3523	1766	1757	0.53 [0.21; 1.33]	0
Healthy pregnant women	2	5932	2950	2982	5.56 [1.92; 16.11]	0

The bold words are considered as statistically significant as the 95% confidence interval of the RR value does not encompass 1.

The subgroup analyses of primary outcomes associated with high-risk pregnant women with PE and different combinations of LDA are summarized in Supplementary Table S1. Among these subgroup analyses, it was observed that for pregnant women at high risk of PE, the subgroup receiving LDA doses ≥100 mg significantly reduced the risk of SGA infants (RR = 0.77, 95% CI: 0.66–0.91) and PTB (RR = 0.56, 95% CI: 0.32–0.97). Additionally, initiating LDA treatment at ≤20 weeks gestation significantly decreased the risk of SGA infants (RR = 0.76, 95% CI: 0.65–0.89) and PTB (RR = 0.56, 95% CI: 0.32–0.97).

Table 3 and Supplementary Figure S3 present the results of univariate meta-regression analysis using mean maternal age as a covariate. The impact of maternal age on the incidence of placental abruption showed significant heterogeneity compared to the placebo group (p = .01), accounting for 60% of the total variance. Nonetheless, maternal age did not significantly affect the other studied outcomes.

Table 3. Meta-regression analysis (covariate: average maternal age).

Outcomes	No. of studies	No. of patients	Stat. model	Coef. (95% CI)	p Value	Heterogeneity explained
SGA	35	40,821	Random-effects meta-regression	−0.019 (–0.045 to 0.007)	.15	−1.5%
PND	22	25,446	Random-effects meta-regression	−0.045 (–0.120 to 0.029)	.23	–
NICU	22	23,819	Random-effects meta-regression	−0.012 (–0.034 to 0.010)	.28	–
LBW	7	22,036	Random-effects meta-regression	−0.052 (–0.143 to 0.040)	.27	–
Placental abruption	11	22,436	Random-effects meta-regression	−0.20 (–0.36 to −0.04)	.01	60%

The bold words are deemed statistically significant due to a P-value <0.05.

The sensitivity analyses on each outcome are summarized in Supplementary Figure S4. Notably, the placental abruption outcome exhibited higher heterogeneity (I² = 66%). Subsequent sensitivity analyses led to the exclusion of several studies, followed by a revised meta-analysis. The findings remained largely consistent with initial results, indicating no significant alterations. The iterative exclusion of individual studies did not materially affect the other outcomes, suggesting the robustness of the pooled estimates against the influence of singular primary studies.

Furthermore, the findings from the cumulative meta-analysis spanning multiple years demonstrate a trend toward increased stability in most obstetrical outcomes over time. A comprehensive summary of these results can be found in Supplementary Figure S5.

Discussion

The obstetrical outcomes of LDA were investigated in this meta-analysis, which included 47 placebo-controlled RCTs with a total of 59,124 participants. The primary findings are as follows: (1) LDA treatment significantly reduces the risk of low-risk events. (2) LDA has a smaller effect on high-risk events and only significantly reduces the risk of NICU admission. (3) LDA treatment is associated with an increased risk of placental abruption. Subgroup analyses, focusing on the dosage and timing of LDA administration, revealed that (1) in pregnant women at high risk of PE, LDA could significantly reduce the risk of SGA, PTB and PND; specifically, when using a dose ≥100 mg or starting ≤20 weeks gestation, it could significantly reduce the risk of SGA and PTB. (2) In healthy pregnant women, LDA did not have a significant effect but instead increases the risk of placental abruption.

The trophoblastic invasion of the spiral arteries and placental development are typically completed by 16–18 weeks of gestation; however, this process is disrupted in individuals with PE, leading to increased inflammation, oxidative stress and endothelial dysfunction [67]. Low-dose aspirin can mitigate these effects through various mechanisms. It reduces trophoblast apoptosis [68,69], as well as inflammation and oxidative stress [70–72], by inducing carbon monoxide production in vascular endothelial cells. Additionally, LDA decreases TXA2 levels to achieve anticoagulation, thereby reducing platelet aggregation and vasoconstriction [20,73]. Not only does it decrease the risk of maternal hypertensive disorders and PE during pregnancy [3], but it also improves placental malperfusion which may enhance fetal growth and development. However, studies have shown that aspirin can cross the placental barrier and enter fetal circulation [5–7], potentially affecting the fetus. For instance, some studies have reported an increased risk of neonatal persistent ductus arteriosus associated with LDA use [63]. Therefore, to elucidate the impact of maternal LDA use on neonates comprehensively, we conducted this meta-analysis. To our knowledge, this is the first meta-analysis evaluating aspirin versus placebo-controlled RCTs to assess obstetrical outcomes related to LDA use during pregnancy. Based on our findings from current trials, data analysis showed that LDA significantly reduced the risk of adverse obstetric outcomes consistently with previous meta-analyses results [18].

Multiple meta-analyses focusing on perinatal outcomes have consistently demonstrated the efficacy of LDA in reducing the incidence of PE [8,10–12]. However, there is still inconclusive evidence regarding its impact on maternal and perinatal outcomes. Some studies have reported a moderate reduction in eclampsia risk among high-risk women after 12 weeks of gestation with LDA prophylaxis (81 mg/day), without any adverse effects on fetal health [13]. Nevertheless, limited research exists concerning the offspring outcomes of LDA administration in normal pregnant women. One study indicated a significant decrease in PTB risk associated with LDA use [58], while another study suggested a potential increase in PTB risk [53], leading to conflicting results. Therefore, this study conducted subgroup analyses by stratifying pregnant women into those at high risk for PE and those who were normotensive to investigate the effects of LDA on offspring within different populations. Our findings demonstrate that LDA reduces neonatal events occurrence among high-risk women without increasing the risk of placental abruption. Conversely, no beneficial effect was observed for offspring born to normotensive pregnant women receiving LDA; instead, an increased risk of placental abruption was identified, posing threats to both maternal and infant safety. In summary, prophylactic administration of LDA not only benefits mothers but also their offspring when considering individuals at high risk for PE. However, caution should be exercised when recommending routine use of LDA as a preventive measure during pregnancy in healthy normotensive women due to potential adverse effects.

The prophylactic use of LDA in women at high risk of PE has demonstrated favorable outcomes for both maternal and neonatal health. However, limited research exists regarding the optimal dosage. The most recent guidelines from the National Institute for Health and Care Excellence (NICE) recommend a daily dose range of 75–150 mg aspirin for women at high risk of PE. Therefore, we categorized the aspirin dosage used in high-risk women as ≥100 mg or <100 mg, and found that doses ≥100 mg produced better outcomes, consistent with a previous study [18].

To determine the optimal timing for LDA administration in women at high risk of PE, our subgroup analysis stratified participants into two groups based on the start time of LDA intake (≤20 weeks and >20 weeks). We employed a cutoff point of 20 weeks to evaluate the initiation of LDA treatment, as PE is defined as hypertension occurring after 20 weeks of gestational age [74]. Previous meta-analyses have consistently demonstrated the efficacy of early initiation of LDA therapy (before 20 weeks) in reducing the risk of PE [18]. Our study further corroborates that administration of aspirin before 20 weeks gestation yields superior obstetric outcomes compared to its use after this threshold.

Limitations

Limitations of this study include the fact that we included studies limited to those published in English. Publication bias was present in both outcomes of PTB and LBW. High heterogeneity was observed in placental abruption outcomes (I² = 66%). Sensitivity analyses led to the exclusion of several studies, resulting in a revised meta-analysis (Supplementary Figure S4), which confirmed the robustness of findings, aligning closely with previous results without significant deviations.

Additionally, despite the fact that our meta-analysis encompasses a substantial number of studies and participants, it is noteworthy that some subgroup analyses involved a limited number of studies and subjects, potentially introducing bias. In addition, the inclusion of studies spanning a broad timeline from early to late pregnancy may complicate the precise classification of LDA initiation timing, presenting challenges in accurately assessing its onset effects.

Conclusions

LDA demonstrated significant efficacy in reducing the risk of SGA, PTB and PND among pregnant women at high risk of PE; however, its impact on obstetrical outcomes in healthy pregnant women was not statistically significant and instead showed an increased risk of placental abruption. Importantly, better results were observed when administering a dose ≥100 mg or initiating treatment ≤20 weeks gestation in pregnant women at high risk of PE.

Author contributions

Conceptualization, X.L. and J.D.; methodology, X.L., M.G. and J.Y.; software, X.L. and M.G.; validation, X.L., M.G., J.Y. and J.D.; formal analysis, X.L., M.G. and J.Y.; investigation, X.L. and J.Y.; resources, X.L.; data curation, X.L. and M.G.; writing – original draft preparation, X.L.; writing – review and editing, J.D. and S.T.; visualization, X.L. and M.G.; supervision, J.D. and S.T.; project administration, J.D. and S.T. All authors have read and agreed to the published version of the manuscript.

Ethical approval

Not applicable.

Consent form

Not applicable.

Acknowledgements

The authors would like to thank Xiao Lu and Yong Gao for their collaboration in the preparation of this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Not applicable.

Additional information

Funding

This research received no external funding.