Maternal influenza vaccination during the first trimester of pregnancy and risk for birth defects: an updated systematic review and meta-analysis

Abstract

Background

Pregnant people have been recommended to receive influenza vaccine at any stage of pregnancy or before conception. However, a large number of pregnant people do not receive the vaccine because of safety concerns. The aim of this study was to investigate the risk of maternal influenza vaccination during the first trimester for birth defects (BDs).

Methods

A literature search of PubMed, Embase, the Cochrane Library, the Web of Science and Scopus databases was performed to identify eligible observational studies published from inception to 28th April 2023. The risk estimates were analyzed by pooling odds ratio (OR) with 95% confidence intervals (CIs). Potential sources of heterogeneity were identified by subgroup and meta-regression analysis. Several categories of sensitivity analyses were performed.

Results

31 studies involving 1,223,024 individuals were included in this study. Compared with the unvaccinated control group or those receiving influenza vaccination during second or third trimester, no increased risk of total BDs (28 studies, unadjusted OR 1.15, 95% CI 0.97–1.36; 10 studies, adjusted OR 1.03, 95% CI 0.98–1.07) or major BDs (16 studies, unadjusted OR 1.11, 95% CI 0.97–1.25; 6 studies, adjusted OR 1.03, 95% CI 0.98–1.09) was observed among those vaccinated during the first trimester. No potential risks were observed based on different vaccine type, the use of adjuvants, the assessment of BDs or control type. Results were consistent across sensitivity analyses for primary results.

Conclusions

This study provided updated evidence that influenza vaccination in early pregnancy did not increase the risk for BDs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12884-026-09238-3.

Introduction

Influenza virus infection has a significant global disease burden, causing 3–5 million cases of severe illness [1, 2]. Compared to the general population, pregnant people are more susceptible to influenza virus infection due to physiologic changes and immunosuppression, which can increase the risk of several adverse pregnancy outcomes (e.g., miscarriage and preterm birth) [3, 4]. The Centers for Disease Control and Prevention (CDC) and the American College of Obstetricians and Gynecologists (ACOG) recommend that all pregnant people receive influenza vaccine at any stage of pregnancy [5, 6]. However, several surveys revealed persistently low uptake of influenza vaccine among pregnant people, such as 61.2% in the US [7], and 1.7%-40.6% in European Union [8, 9]. In China, only 11.4% of pregnant people expressed willingness to receive an influenza vaccine during pregnancy [10]. One of the major reasons for this low uptake is concern about potential adverse events following maternal influenza vaccine, such as stillbirth, preterm birth, low birth weight, and birth defects (BDs) [11].

The first trimester is a crucial stage of embryonic development, during which the organs of the embryo are formed. This period is particularly vulnerable to external factors that may induce BDs, such as drugs and infection. Vaccination during pregnancy has been consistently shown to effectively prevent influenza in expectant mothers and protect infants under six months of age through the transmission of protective antibodies across the placenta [12, 13]. Although influenza vaccination during pregnancy is generally considered safe, some studies have explored potential concerns regarding first-trimester vaccination. For example, Pasternak et al. reported a slight increase in the risk of BDs in the H1N1-vaccinated population during the first trimester of pregnancy in unadjusted analyses (unadjusted OR 2.26, 95% CI 1.40–3.66); although this association was no longer statistically significant after adjustment for confounders [14]. Chambers et al. found no significant association between influenza vaccination and BDs (adjusted OR 0.79, 95% CI 0.26–2.42) [15], however, the wide confidence interval reflects limited precision.

To date, several meta-analyses have explored the association between influenza vaccination and BDs [16–18]. However, few studies have specifically focused on those vaccinated during the first trimester and the sample sizes in these studies are relatively small. Moreover, these analyses have not scrutinized potential associations between BDs and other pivotal variables related to influenza vaccination, such as differentiating between pandemic monovalent and seasonal multivalent vaccines. The impact of influenza vaccination during early pregnancy on the occurrence of BDs remains uncertain. In order to address this issue and provide further evidence in the field of reproductive health, we conducted this updated systematic review and meta-analysis of all available evidence to assess the association between maternal influenza vaccination during the first trimester and the risk of BDs.

Methods

This study was conducted and reported following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement [19]. The review protocol was prospectively registered in PROSPERO (CRD42023450641).

Eligibility criteria

Cohort studies and case-control studies were included if they (1) evaluated the association between influenza vaccination during pregnancy and adverse pregnancy outcomes; (2) included at least one exposure group that received influenza vaccine during the first trimester; (3) included total BD, major BD (major structural anomalies with serious health or developmental consequences) or organ system specific BDs (e.g., congenital malformations of the circulatory system) as outcomes among live-born infants.

Studies were excluded if they (1) were conference abstracts, recommendations or review; (2) did not report the original number of events and total participants in vaccination and control group, or adjusted effect size and 95% confidence intervals (CIs).

Literature search

A systematic literature search was conducted on PubMed, Embase, Cochrane Library, Web of Science and Scopus databases from the earliest available date (i.e., 1981) to 28th April 2023. A pre-designed and modified search strategy (see detailed search strategy in Appendix 1), which combined of MeSH terms and free-text words, was used to identify eligible studies, including but not limited to pregnancy, maternal, gestation, pregnant, influenza vaccine, flu vaccine, influenza virus vaccine, flu vaccination, influenza vaccination, influenza virus vaccination. Furthermore, the reference lists of the included articles and relevant systematic reviews, which investigated the influence of influenza vaccination on BD, were manually checked for additional relevant literature.

Study process

Two reviewers (C.C. and X.L.) independently screened titles, abstracts, followed by full texts to check the eligibility for inclusion. Any disagreements were resolved by a third reviewer (Y. W.). Data from each eligible study was extracted using standardized and pilot-tested forms. The forms included general study information (e.g., first author name, year of publication, country of origin, sample size and study design), the vaccination information (e.g., vaccine type, the use of adjuvants and the timing of vaccination), BDs information (e.g., classification, the assessment of BDs, number of BDs), and adjusted effect estimates with 95% confidence intervals.

Risk of bias assessment

The risk of bias in the included studies was assessed independently by two reviewers (X.L. and Y.Z.) using the Newcastle-Ottawa Scale (NOS) [20]. The scale scores studies on three categories: selection of subjects, comparability of study groups, and assessment of exposure, with a maximum score of 9 points. Studies with a quality score of 7 or higher were considered to be of high quality.

Exposure and outcomes

First trimester vaccination (exposed group) was defined as pregnant people receiving influenza vaccine during 0–13^+ 6 gestation weeks. People who received the influenza vaccine during the second or third trimester, or who did not receive the influenza vaccine during pregnancy were classified as the control group.

The primary outcome in this study was total BDs [21]. The second outcomes were major BDs and 10 organ system specific BDs. Major BDs refers to structural congenital anomalies present at birth that have significant medical, social, cosmetic, or functional consequences for the affected individual and typically require medical or surgical intervention [22]. The organ system specific BDs included congenital malformations of the nervous system, eye, ear, face and neck, the circulatory system, the respiratory system, cleft lip and cleft palate, the digestive system, genital organs and the urinary system, the musculoskeletal system, other congenital malformations, chromosome abnormalities. The number of BD was defined as the number of infants who were with one or more type of BD. In instances where number of infants was unavailable, number of maternal was substituted.

Statistical analysis

In this study, all unadjusted and adjusted data were separately pooled. Between-study statistical heterogeneity was determined by Cochran’s Chi-squared test and the I² statistic. Substantial statistical heterogeneity was defined as I² ≥ 50%. OR with 95% confidence intervals (CIs) were calculated as effect indicators by using the Mantel-Haenszel random-effects model. Because the incidence of the outcome in this study is usually less than 5%, we treat OR, hazard ratio (HR), prevalence risk (PR), and relative risk (RR) as the same effect measures. When pooling the risk estimates, we prioritized using the unvaccinated group during pregnancy as the control group. In cases where this group was not available, vaccinated during the second trimester, third trimester, or any combination of them was chosen as the control group.

Subgroup analyses were performed based on several factors: vaccine type (pandemic monovalent vs. seasonal multivalent), the use of adjuvants (nonadjuvanted vs. adjuvanted vs. nonadjuvanted/adjuvanted), study design (cohort study vs. case-control study), the assessment of BDs (medical records vs. self-reported vs. medical records/self-reported) and control type (unvaccinated control vs. the second trimester control vs. the third trimester control vs. the second or third trimester control). The statistical difference of the effects among these subgroups was tested by interaction test. Additionally, meta-regression analysis was employed to evaluate heterogeneity.

Sensitivity analyses were conducted employing four approaches (1) adding 1 BD to both exposed and control group in these studies that reported no BD observed in all participants [23]; (2) deleting the studies with a low quality (NOS score less than 7); (3) sequentially excluding one study at a time and recalculating the pooled estimate to assess whether any single study disproportionately influenced the results; (4) using the trim and fill method to impute potentially missing studies and adjust the pooled estimate for potential publication bias [24]. To assess publication bias, we conducted Egger’s and Begg’s tests and generated a funnel plot. All statistical analyses were performed using Stata version 17.0 (Stata Corp LLC, College Station, TX).

Results

Characteristics of included studies

We initially identified 14,339 relevant studies from the literature search, and 31 eligible studies involving 1,223,024 participants were ultimately included [14, 15, 23, 25–53]. The flow chart illustrating the study selection process is shown in Fig. 1.

Fig. 1.

PRISMA (Preferred reporting Items for Systematic Reviews and Meta-Analyses) flow diagram of the literature search and study selection process. ††Of the 31 included studies, 29 reported total birth defects (28 were included in pooled analysis), 16 reported major birth defects, and 12 reported organ system specific outcomes

There were 17 prospective studies [15, 23, 25, 27, 32–36, 38, 41, 42, 45, 49–52], 10 retrospective studies [14, 26, 28–31, 37, 40, 46, 53] and 4 case-control studies [39, 44, 47, 48]. Of the 31 studies, 28 reported total BDs, 16 reported major BDs, and 12 reported organ system specific BDs. All vaccines mentioned in the included studies were inactivated influenza vaccines, with 14 studies using pandemic monovalent vaccination, 13 studies using seasonal trivalent or quadrivalent influenza vaccines, 2 studies using any of the aforementioned types, and 2 studies not reporting the vaccine type. 23 studies reported BDs based on medical records, while 3 studies relied on self-reporting. 4 studies used a combination of medical records and self-reported data and 1 study did not mention the method. Further descriptions of the included studies showed in Table 1. The included studies exhibited a low risk of bias, 26 (83.87%) achieved a score of 7 or higher.

Table 1.

Characteristics of included studies

ID	Study	Country	Study design	Study center	Study duration	Total number of participants	Age of the participants at conception(y)	No. of exposed groups	No. of BDs in exposed groups	No. of control groups	No. of BDs in control groups	Vaccine type	Assessment of birth defects	Control typea	Adjustment factorsc	NOS
1	Deinard (1981)	US	Prospective cohort study	4	NA	706	21.97	41	7	517	67	H1N1 (A/New Jersey/76) monovalent pandemic, nonadjuvanted	Medical records	a1 a2 a3 a4 a5	NA	7
2	Ka¨lle´n (2012)	Sweden	Retrospective cohort study	Multicenter	2009–2010	238,824	NA	3197	200	111,297	2997	H1N1 2009 monovalent pandemic, AS03-adjuvanted	Medical records	a1	c1	9
3	Ludvigsson (2016)	Sweden	Prospective cohort study	7	2009.10-2011.10	137,660	39.8	14,385	732	56,886	2825	H1N1 2009 monovalent pandemic, AS03-adjuvanted	Medical records	a1 a5	c2	9
4	Moro (2016)	US	Retrospective cohort study	Multicenter	2010.7-2016.5	671	31	116	5	69	1	Trivalent, seasonal, nonadjuvanted	Medical records	a2 a3 a4	NA	7
5	Moro (2020)	US	Retrospective cohort study	Multicenter	2013.7-2020.5	391	29.5	92	1	248	11	Trivalent, seasonal, nonadjuvanted	Medical records or Self-reported	a4	NA	7
6	Peppa (2021)	UK	Retrospective cohort study	Multicenter	2010.9-2016.3	71,124	NA	6872	565	46,669	3289	Trivalent or quadrivalent, seasonal, nonadjuvanted	Medical records	a1 a2 a3 a4 a5	c3	9
7	Sheffield (2012)	US	Retrospective cohort study	Multicenter	2003.10-2008.3	84,843	26.0-26.8	447	5	76,919	1163	Trivalent, seasonal, nonadjuvanted	Medical records	a1	NA	7
8	Tavares (2011)	UK	Prospective cohort study	87	2009.10-2009.12	267	30.9	38	0	217	5	H1N1 2009 monovalent pandemic, AS03-adjuvanted	Medical records	a3 a4	NA	7
9	Mackenzie (2011)	UK	Prospective cohort study	1015	2009.11-2010.4	128	NA	9	1	7	0	H1N1 2009 monovalent pandemic, nonadjuvanted or AS03-adjuvanted	Self-reported	a1 a2 a3 a4 a5	NA	6
10	Robinson (2022) a*	US	Prospective cohort study	5	2017–2020	665	28	173	2	212	5	Quadrivalent, seasonal, nonadjuvanted	Medical records or Self-reported	a2 a3 a4	NA	7
11	Robinson (2022) b*	US	Prospective cohort study	5	2017–2021	483	29.8	171	2	111	0	Quadrivalent, seasonal, nonadjuvanted	Medical records or Self-reported	a2 a3 a4	NA	7
12	Pasternak (2012)	Denmark	Retrospective cohort study	Multicenter	2009.11-2010.9	53,432	NA	345	18	22,917	545	H1N1 2009 monovalent pandemic, AS03-adjuvanted	Medical records	a1	c4	9
13	Ohfuji (2020)	Japan	Prospective cohort study	117	2013.10-2014.12	10,330	32	1121	33	1705	55	NA	Medical records	a1	NA	7
14	Kharbanda (2019)	US	Retrospective cohort study	Multicenter	2004.1-2013.9	425,944	NA	52,552	861	282,091	4284	Trivalent or quadrivalent, seasonal, nonadjuvanted	Medical records	a1 a5	c5	8
15	Auffret (2013)	France	Prospective cohort study	47	2009.11-2010.4	781	30.9	24	0	322	6	H1N1 2009 monovalent pandemic, nonadjuvanted or AS03-adjuvanted	Self-reported	a2 a3 a4	NA	6
16	Huang (2014)	China	The case-control design	1	2009.11-2010.9	34,244	NA	161	7	10,783	1051	H1N1 2009 monovalent pandemic, nonadjuvanted or MF59-adjuvanted	Medical records	a1	c6	9
17	Cleary (2014)	UK	Retrospective cohort study	1	2008.12-2010.9	6894	NA	246	9	3898	110	H1N1 2009 monovalent pandemic, nonadjuvanted or AS03-adjuvanted	Medical records	a1	NA	7
18	Vries (2014)	Netherlands	Prospective cohort study	1	2009–2010	278	31.2	23	1	135	0	H1N1 2009 monovalent pandemic, MF59-adjuvanted or AS03-adjuvanted	Self-reported	a2 a3 a4	NA	6
19	Heikkinen (2012)	Finland	Prospective cohort study	36	2009.11-2010.8	4508	31.6	94	2	2213	41	H1N1 2009 monovalent pandemic, MF59-adjuvanted	Medical records	a1 a2 a3 a4 a5	NA	7
20	Chambers (2013)	US or Canada	Prospective cohort study	1	2009.10-2012.4	1032	NA	328	9	188	6	H1N1 2009 monovalent pandemic or trivalent influenza vaccine, nonadjuvanted	Medical records	a1	c8	9
21	Chambers (2016)	Italy	Prospective cohort study	1	2010–2014	1730	NA	457	26	427	13	Trivalent or quadrivalent, seasonal, nonadjuvanted	Medical records	a1	NA	7
22	Sarna (2022)	Australia	Retrospective cohort study	Multicenter	2012.4-2016.4	125,866	NA	2811	62	112,125	2491	Trivalent or quadrivalent, seasonal, nonadjuvanted	Medical records or Self-reported	a1	c9	9
23	Louik (2016)	US	The case-control design	5	2011–2012, 2012–2013, 2013–2014	7054	NA	711	461	3566	2405	Trivalent or quadrivalent, seasonal, nonadjuvanted	Medical records	a1	c11	9
24	Oppermann (2012)	Germany	Prospective cohort study	1	2009.9-2010.3	1652	32–33	52	6	1198	138	H1N1 2009 monovalent pandemic, nonadjuvanted, AS03-adjuvanted or MF59-adjuvanted	Medical records	a1	NA	7
25	Horiya (2011)	US	Prospective cohort study	1	2009.10	210	34.8–35.7	15	0	29	2	H1N1 2009 monovalent pandemic, nonadjuvanted or AS03-adjuvanted	NA	a2 a3 a4	NA	6
26	Mohammed (2020)	Australia	Prospective cohort study	2	2015.3-2017.12	1253	29.5	141	2	1066	21	Trivalent, nonadjuvanted	Medical records	a5	c12	9
27	Chavant (2013)	France	Prospective cohort study	17	2009.11-2010.3	2415	31	92	1	1162	34	H1N1 2009 monovalent pandemic, nonadjuvanted, AS03-adjuvanted or MF59-adjuvanted	Medical records	a2 a3 a4	NA	7
28	Lin (2012)	China	Retrospective cohort study	3	2009.10-2010.2	396	32.4–32.8	10	0	206	7	H1N1 2009 monovalent seasonal, nonadjuvanted	Medical records	a1 a2 a3 a4 a5	NA	7
29	Nwoji (2022)	US	Prospective cohort study	4	2014.6-2019.5	115	NA	30	0	37	0	Quadrivalent, seasonal, nonadjuvanted	Medical records	a2 a3 a4	NA	6
30	Palmsten (2021)	US	The case-control design	10	1997.10-2011.12	4937	NA	NA	NA	NA	NA	NA	Medical records	a5	c7	9
31	Louik (2013)	US	The case-control design	4	2009.10-2010.7, 2010.8-2011.7	4191	NA	NA	NA	NA	NA	H1N1 2009 monovalent pandemic or trivalent influenza vaccine, nonadjuvanted	Medical records	a1	c10	8

* Robinson (2022) a and Robinson (2022) b were two studies with different vaccine process based on chicken embryos and cell-based, respectively

a control type:

a1 unvaccinated control

a2 the third trimester control

a3 the second trimester control

a4 the second or third trimester control

a5 the second trimester, third trimester or unvaccinated control

c adjusted factors:

c1 maternal age, year of birth, parity, smoking and BMI

c2 maternal age, BMI, sex, parity, smoking, education, mother’s country of birth, and disposable income

c3 maternal age, maternal ethnicity, region, and influenza season the number of weeks the first trimester overlapped with a period of influenza activity above baseline levels as well as the following maternal factors: IMD, number of children in the household, smoking status, hazardous drinking, extreme BMI, clinical risk group, chronic hypertension, and exposure to teratogenic drugs and/or live vaccines

c4 maternal age, place of birth, degree of urbanization at place of living, parity, smoking status, and pre-pregnancy body mass index; history of any birth defects, preterm birth, spontaneous abortion, and small size for gestational age; maternal comorbidities and use of drugs; and health care utilization: number of hospital admissions and hospital outpatient visits in the last 3 years, and number of drugs used in the last 6 months

c5 This article did not report adjustment factors

c6 maternal age, maternal medical risk factors, labor and delivery complications, and history of smoking during pregnancy

c7 maternal age at delivery, maternal race/ethnicity, plurality, cigarette smoking, alcohol use, low folate intake, and, for neural tube defects, folate antagonist medication use, education, body mass index, gravidity, hypertension during pregnancy, fever, and respiratory illness

c8 race, previous elective termination, infection and directly adjusted for autoimmune disease

c9 maternal age, mother’s ethnicity, parity, season of birth, smoking status, sex, antenatal care indicator, mother’s medical conditions, and disadvantage

c10 maternal age, maternal race, maternal education, family income, marital status, parity, study center, body mass index (BMI), family history of birth defects, pregnancy intention, periconceptional folic acid use, alcohol use, smoking, asthma, diabetes, LMP quarter, infertility treatment, treatment for high blood pressure or toxemia, inter-pregnancy interval, and season of exposure (2009–2010 or 2010–2011)

c11 maternal age, maternal race/ethnicity, maternal education, family income, marital status, parity, study center, pre-pregnancy BMI, pregnancy intention, periconceptional folic acid use, alcohol use, smoking, asthma, diabetes, high blood pressure or toxemia, interpregnancy interval, any infection in pregnancy, LMP quarter, infertility medication use, family history of birth defects, multiple births, coffee consumption, employment outside the home, illicit medication use, and history of miscarriage

c12 maternal age, ethnicity, total years of fulltime education, household income, gravidity, intake of alcohol and recreational drugs, smoking, pre-pregnancy body mass index (continuous), use of multivitamin supplements, Edinburgh Postnatal Depression Scale (EPDS), The State-Trait Anxiety Inventory (STAI), Perceived Stress Scale (PSS-10), physical activity, infertility treatment, asthma and estimated season of delivery and infant’s sex

Association between influenza vaccination during the first trimester and BDs

By including 28 studies involving 1,213,781 pregnant people, the pooled unadjusted results indicated that influenza vaccination during the first trimester did not increase the risk of total BDs (28 studies, unadjusted OR 1.15, 95% CI 0.97–1.36, I² = 81.7%; Fig. 2A). Similarly, the pooled adjusted data also showed no association between influenza vaccination during the first trimester and total BDs (10 studies, adjusted OR 1.03, 95% CI 0.98–1.07, I² = 0.0%; Fig. 2B).

Fig. 2.

Forest plot for the risk of total birth defects associated with maternal influenza vaccination during the first trimester. (A) unadjusted data; (B) adjusted data. † †Of the 31 included studies, two reported only organ system specific birth defects, and one study reported no cases in either the exposed or control groups; therefore, 28 studies contributed to the pooled analysis of total birth defects ‡ In the analyses, 10 studies used maternal counts because infant BD data were unavailable, and 9 studies used second or third trimester vaccination as the comparison in the absence of an unvaccinated control group

In addition, no increased risk for major BDs was found in pooled unadjusted data (16 studies, unadjusted OR 1.11, 95% CI 0.97–1.25, I² = 44.7%; Fig. 3A) and pooled adjusted data (6 studies, adjusted OR 1.03, 95% CI 0.98–1.09, I² = 0.0%; Fig. 3B). Based on unadjusted data, no increased risk for BDs was found in 9 out of the 10 organ system specific BDs, with the exception of urogenital system (Table 2). For circulatory system BDs, adjusted data showed no increased risk (5 studies, adjusted OR 0.98, 95% CI 0.89–1.07, I² = 0.0%).

Fig. 3.

Forest plot for the risk of major birth defects associated with maternal influenza vaccination during the first trimester. (A) unadjusted data; (B) adjusted data

Table 2.

Maternal influenza vaccination during the first trimester and organ systems specific BDs

Organ system	No. of studies	BDs in exposed group (n/N)	BDs in control group (n/N)	The pooled unadjusted OR (95% CI)
Circulatory system	7*	857/81,290	8715/618,900	1.05 (0.97–1.13)
Cleft lip and cleft palate	7	138/54,274	673/286,075	1.00 (0.82–1.21)
Digestive system	3	114/56,484	677/395,921	0.93 (0.76–1.13)
Musculoskeletal system	3	73/56,484	1125/395,921	1.08 (0.77–1.52)
Nervous system	2	87/53,673	438/283,796	1.06 (0.84–1.34)
Respiratory system	1	1/1121	0/1705	4.57 (0.19-112.19)
Urogenital system	3	272/56,484	2023/395,921	1.19 (1.00-1.42)
Congenital malformations of eye, ear, face and neck	3	26/56,484	143/395,921	1.51 (0.74–1.77)
Other congenital malformations	1	73/52,552	385/282,091	1.02 (0.79–1.31)

*This included Palmsten’s (2021) study reporting the cardiovascular system, which comprises the majority of the circulatory system

Subgroup analysis and meta regression

The results of the subgroup analyses of different vaccine type, different adjuvants, and the assessment of BDs, different control type were consistent with the primary results. Although a slight increased risk of BD was observed in the pooled unadjusted data of cohort studies, the pooled adjusted data showed a non-significant association (Table 3). Meta-regression, which included variables such as vaccine type, the use of adjuvants, study design and the assessment of BDs, revealed no significant differences in heterogeneity between subgroups.

Table 3.

Subgroup analysis of association between total BDs and influenza vaccination during the first trimester based on the unadjusted data of 28 studies† and adjusted data of 10 studies

Subgroup	Unadjusted effects			Adjusted effects
	No. of studies	OR (95% CI)	P value*	No. of studies	OR (95% CI)	P value*
Study design			0.11			0.40
Cohort study	26	1.23 (1.02–1.47)		8	1.03 (0.99–1.08)
Case-control study	2	0.67 (0.33–1.37)		2	0.75 (0.35–1.58)
Vaccine type			0.72			0.85
Pandemic monovalent	14	1.24 (0.81–1.89)		4	1.01 (0.89–1.14)
Seasonal multivalent	12	1.06 (0.95–1.18)		5	1.03 (0.98–1.09)
Pandemic monovalent or Seasonal multivalent	1	0.86 (0.30–2.44)		1	0.79 (0.26–2.41)
NA	1	-		-	-
The use of adjuvants			0.53			0.11
Nonadjuvanted	14	1.07 (0.97–1.17)		6	1.03 (0.98–1.09)
AS03-adjuvanted	4	1.65 (0.85–3.22)		3	1.03 (0.96–1.11)
MF59-adjuvanted	1	1.15 (0.27–4.83)		1	0.46 (0.22–0.98)
Any of the above	8	0.86 (0.48–1.54)		-	-
NA	1	-		-	-
The assessment of BDs			0.28			0.71
Medical records	20	1.18 (0.98–1.41)		9	1.03 (0.99–1.08)
Self-reported	3	3.31 (0.54–20.34)		-	-
Medical records or self-reported	4	0.93 (0.64–1.35)		1	0.98 (0.76–1.26)
NA	1	-		-	-
Control type			0.79			0.91
Unvaccinated control	18	1.18 (0.99–1.40)		8	1.02 (0.97–1.08)
The second trimester control	13	1.07 (0.96–1.19)		-	-
The third trimester control	12	1.28 (0.84–1.95)		-	-
The second or third trimester control	14	1.16 (0.84–1.62)		-	-
The second trimester, third trimester or unvaccinated control	8	1.08 (1.03–1.13)		3	1.03 (0.97–1.09)

*P value for interaction between subgroups. A two-sided P value < 0.05 was considered statistically significant

†Of the 31 included studies, two reported only organ-system–specific birth defects, and one study reported no cases in either the exposed or control groups; therefore, 28 studies contributed to the pooled analysis of total birth defects

Sensitivity analyses

By additionally including one study reported zero events in both the exposed and control groups [23], the results (29 studies, unadjusted OR 1.15, 95% CI 0.97–1.36; Figure S1) were consistent with primary pooled results. When deleting the low-quality studies, the result remains stable (23 studies, unadjusted OR 1.14, 95% CI 0.96–1.36; 10 studies, adjusted OR 1.03, 95% CI 0.98–1.07). By sequentially removing individual eligible studies, we observed that no individual study altered the pooled unadjusted or adjusted results (Figure S2A; Figure S2B). Furthermore, trim and fill method did not identify or supplement any additional studies using unadjusted data (28 studies, unadjusted OR 1.15, 95% CI 0.97–1.37) or adjusted data (10 studies, adjusted OR 1.03, 95% CI 0.98–1.07).

Publication bias

Funnel plot was symmetric across total BDs based on unadjusted and adjusted data (Figure S3A; Figure S3B). No potential publication bias was presented by Egger’s test (P = 0.93 for unadjusted pooled analysis; P = 0.04 for adjusted pooled analysis) and Begg’s test (P = 0.41 for unadjusted pooled analysis; P = 0.09 for adjusted pooled analysis).

Discussion

Summary of main findings

By including 31 studies involving 1,223,024 participants, this meta-analysis suggests that receiving influenza vaccine during the first trimester of pregnancy is not associated with a higher risk of total BDs (adjusted OR 1.03, 95% CI 0.98–1.07) or major BDs (adjusted OR 1.03, 95% CI 0.98–1.09), which represents the most comprehensive evidence to date. The results of several sensitivity analyses were consistent with the primary outcome, which substantiates the robustness of our findings.

Comparison with previous studies

In 2015, a meta-analysis first reported that maternal influenza vaccination was not associated with an increased risk of total BDs (15 studies, OR 0.96, 95% CI 0.86–1.07) [17]. Since then, several subsequent meta-analyses have yielded similar results [16, 18, 54]. However, there are still studies that have found some specific influenza vaccinations may be associated with a higher risk of BDs. For example, Zhang et al. reported an association between H1N1 vaccination during pregnancy and BDs (19 studies, adjusted OR 1.14, 95% CI 1.02–1.29) [55]; however, most vaccinations in their study occurred after the first trimester, which may contribute to variability in pooled estimates and complicate interpretation with respect to teratogenic risk. Although the safety profile of influenza vaccines during pregnancy is considered favorable [56], the level of evidence for BDs has very low certainty according to the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) framework [57], and the association between maternal influenza vaccination during the first trimester and BDs remains a much-debated topic. In this study, we defined the exposed group only as pregnant people who received vaccine during the first trimester, the critical stage of embryonic development, thus excluding the noncausal effect of vaccination during the second or third trimester. At the same time, in addition to including an unvaccinated control group, we also conducted the comparison to vaccination during the second trimester, third trimester, the second or third trimester and the second trimester, third trimester or unvaccinated control. Using these exposure and control group definitions, no significant association with increased risk was found in our study (Table 3).

In addition to total BDs, several studies have focused on the impact of influenza vaccination on major BDs [43, 46], which have serious consequences for the health of the fetus, including long-term health problems, disability, and even death [58]. There were two previous meta-analyses that investigated the risk of major BDs after receiving influenza vaccines during pregnancy, revealing no significant association [17, 18]. Moreover, our study focused on vaccination during the first trimester and showed similar results (Table 2). However, limited studies have explored the risk of influenza vaccination for organ system specific BDs. In our study, we observed a possible week link between influenza vaccination during pregnancy and an increased risk of urogenital system BDs. Nonetheless, the results should be viewed with caution due to the small number of studies and unadjusted confounding factors. Further cohort studies with larger sample sizes may be required to confirm these associations for organ system specific BDs.

The potential difference in effects of seasonal and pandemic influenza vaccines on maternal and infant adverse outcomes is also a significant concern. Previous meta-analyses suggested that no significant difference in effect was found between seasonal and pandemic vaccines when focusing on preterm birth (seasonal vaccine, RR 1.12, 95% CI 0.88–1.46; pandemic vaccine, RR 1.15, 95% CI 0.93–1.46 ) [54] and low birth weight (seasonal vaccine, RR 0.74, 95% CI 0.52–1.05; pandemic vaccine, RR 0.89, 95% CI 0.78–1.02) [57]. By including 15,510 live births, a cohort study showed that maternal vaccination during pregnancy with pandemic H1N1 vaccine, compared with seasonal influenza vaccination, were not at increased risk of BDs (adjusted OR 1.14, 95% CI 0.96–1.35) [59]. According to the subgroup analyses, our study also showed either pandemic influenza vaccines (14 studies, adjusted OR 1.24, 95% CI 0.81–1.89) or seasonal influenza vaccines (12 studies, adjusted OR 1.06, 95% CI 0.95–1.18) administered during the first trimester did not increase the risk of BDs, which provide support pandemic influenza vaccination during pregnancy.

Implications for practice and research

This study suggests that influenza vaccination during early pregnancy did not increase the risk of BDs, which may help alleviate concerns about vaccination safety during pregnancy and support informed decision-making, particularly among individuals who were not vaccinated prior to conception. Although causal inferences cannot be made, influenza vaccination during pregnancy has the potential to contribute to the prevention of influenza-related complications, thereby supporting the overall quality of prenatal care.

Strengths and limitations

Our study has two main strengths. First, we conducted a comprehensive examination of the relationship between influenza vaccination during the first trimester and various aspects of BDs, including total BD, major BD, and organ system specific BDs. Secondly, we employed rigorous data analyses, including the pooling of unadjusted and adjusted effects, subgroup analyses, and sensitivity analyses, to substantiate the robustness of our findings.

This study also has four main limitations. First, some of the pooled results in this study had heterogeneity, which could be caused by differences in basic characteristics of the populations. Although random effect was used, the impact of heterogeneity on the pooled results was not wholly eliminated. Second, this study only examines BDs in live births. BDs in miscarriages or stillbirths are not included in the analysis because of insufficient reporting in the original studies. Third, because BDs are rare events, the number of BDs in some pooled analyses are small, especially in organ system specific BDs. For these specific BDs, prospective studies with larger sample sizes are needed to further verify the association. Fourth, most included studies were conducted over a decade ago, which may limit the generalizability of these findings to current clinical settings.

Conclusions

Compared with the unvaccinated control group or those receiving influenza vaccination during second or third trimester, no increased risk of BDs was observed among those vaccinated during the first trimester.

Authors’ contributions

YW: Methodology; Formal analysis; Writing- original draft; Writing-review & editing; Project administration. CL: Methodology; Data curation; Validation; Visualization; Writing - original draft. XQ: Methodology; Data curation; Validation; Visualization; Writing-original draft; YZ: Methodology; Data curation; Formal analysis; Validation; Visualization. YX: Conceptualization; Supervision; Project administration; Funding acquisition; Writing - review & editing.

Funding

This study was funded by National Natural Science Foundation of China (72004148), Zhejiang Provincial Traditional Chinese Medicine Science and Technology Project (2024ZL374), and Sichuan Provincial Natural Science Foundation (24NSFSC7783), Application of Real-World Data in Post-Marketing Drug Safety Surveillance and Evaluation (HX-H2403098).

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.