First-Trimester Influenza Infection Increases the Odds of Non-Chromosomal Birth Defects: A Systematic Review and Meta-Analysis

Viral infections during pregnancy raise several clinical challenges, including birth defects in the offspring. Thus, this systematic review and meta-analysis aims to prove and highlight the risk of birth defects after first-trimester maternal influenza infection. Our systematic search was performed on 21 November 2022. Studies that reported maternal influenza infection in the first trimester and non-chromosomal congenital abnormalities were considered eligible. We used odds ratios (OR) with 95% confidence intervals (CIs) to measure the effect size. Pooled ORs were calculated with a random effects model. Heterogeneity was measured with I² and Cochran’s Q tests. We found that first-trimester maternal influenza was associated with increased odds of developing any type of birth defects (OR: 1.5, CI: 1.30–1.70). Moreover, newborns were more than twice as likely to be diagnosed with neural tube defects (OR: 2.48, CI: 1.95–3.14) or cleft lip and palate (OR: 2.48, CI: 1.87–3.28). We also found increased odds of developing congenital heart defects (OR: 1.63, CI: 1.27–2.09). In conclusion, influenza increases the odds of non-chromosomal birth defects in the first trimester. The aim of the present study was to estimate the risk of CAs in the offspring of mothers affected by first-trimester influenza infection.


Introduction
The recent outbreak of COVID-19 pneumonia, caused by SARS-CoV-2, has highlighted the role of viral infections during pregnancy as well [1,2]. The importance of these infections will probably increase as we face growing risks of pandemics, which may affect the pregnant mother and the fetus [3]. Viral infections during pregnancy raise several clinical challenges, including adverse pregnancy outcomes and birth defects in the offspring [4].
Influenza is an acute infectious disease caused by influenza A, B or C viruses. Most cases occur during epidemic outbreaks, generally between December and March. The latency period between infection and the manifestation of symptoms is about 48 h. Healthy individuals usually recover in 3-7 days; however, influenza may be followed by secondary bacterial infections of the respiratory system [5]. Pregnant women are among the groups that are at increased risk of complications [6].
Influenza may occur in pregnant women. Organogenesis takes place in the first trimester, so any environmental effect that occurs during this period may affect the development of the embryo. However, the role of influenza viruses is debated in the origin of congenital abnormalities (CAs). Some epidemiologic studies have shown a small increase in CAs in general or in some specific CAs (e.g., heart defects, esophageal atresia, anencephaly), while others did not find any increase of CAs after influenza epidemics [7][8][9][10][11][12]. However, data from these reports are difficult to assess because of lacking serologic proof of the microbial agents, unknown time of infection during pregnancy, restricted numbers of CAs and inadequate controls. One meta-analysis [13] on this topic also highlighted an association between maternal influenza and non-chromosomal birth defects. However, the included studies were not restricted to first-trimester infection It is not clear whether there is a direct link between influenza infection and the development of birth defects or it exerts an indirect teratogenic effect. High fever associated with influenza was assumed to play a causative role in this pathologic process [14].
Our study hypothesized that influenza in the first trimester increases the likelihood of developing congenital abnormalities. Thus, the present study aimed to estimate the risk of CAs in the offspring of mothers affected by influenza during the first trimester of the pregnancy. Worldwide, 3-5% of newborns are affected by a congenital anomaly [14]. These birth defects are to a great extent responsible for infant mortality [15]. The prevention of CAs is considered a public health priority due to their medical and financial consequences. Therefore, finding the role of influenza infection in the development of CAs is also of significant importance.

Materials and Methods
The systematic review and meta-analysis is reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 Statement (see Supplementary Table S1) [16] and the MOOSE Checklist (see Supplementary Table S2) [17]. We followed the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions [18]. The protocol was prospectively registered in PROSPERO (International Prospective Register of Systematic Reviews) under the registration number CRD42021283210 (https://www.crd.york.ac.uk/prospero, accessed on 19 October 2021). There was no protocol deviation.

Systematic Search
The systematic search was performed in three major medical databases: MEDLINE (via PubMed), Cochrane Library (CENTRAL) and Embase on 20 October 2021, and we reran our search on 21 November 2022. The search key contained the following main components: influenza, pregnancy, birth defect or congenital anomalies. The detailed queries are shown in the Supplementary Material. There were no filters or other restrictions used during the search.

Selection and Eligibility Criteria
We used EndNote 20 (Clarivate Analytics, Philadelphia, PA, USA) to select the retrieved articles. After removing duplications, two independent authors screened the library separately by title and abstract, then by full text (Á.M., A.F.). We calculated Cohen's kappa coefficient (κ) after each selection process to measure interrater reliability [19]. Disagreements were resolved after each step by a third author (B.T.).
We used the population-exposure-outcome (PEO) framework to structure our research question. Case-control studies and cohort analyses, including pregnant women (P), who were investigated for influenza infection (E) during the first trimester of pregnancy, were found eligible. The outcomes of interest were all types of non-chromosomal birth defects (O).
The exclusion criteria concerned conference abstracts, articles reporting on common cold or fever instead of influenza, and influenza before or after the first trimester. To identify all relevant articles on the topic, the reference list of eligible full texts, and studies that cited the key articles were also checked. If the articles were not written in English, we sought the help of a translator. We contacted the authors if there was no available text Viruses 2022, 14, 2708 3 of 17 or data from an article. When the study population was overlapping for an outcome, we included the study with a larger sample size.

Data Extraction
Two authors independently (Á.M., A.F.) extracted the data. We created a standardized data collection Excel sheet (Office 365, Microsoft, Redmond, WA, USA), which included first author, year of publication, title of the study, digital object identifier (DOI) number, study design, study type, study period, country, number of the participant centers, demographic characteristics of the included patients, number of participants, population characteristics of the study, data about the exposure (definition, time), raw data regarding the outcomes and unadjusted and adjusted results separately. A third author (B.T.) resolved the disagreements between the two authors.

Data Synthesis and Analysis
Qualitative data synthesis was done in the R programming language (R Core Team, 2012, Vienna, Austria, R version 4.2) using the meta v5.5.0 [20] and dmetar v0.0.9000 packages. A meta-analysis was performed if there were a minimum of three studies for one outcome. Where pooling effect sizes could not be carried out, we only visualized results on forest plots.
When possible, pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated from 2 by 2 tables raw data or from crude ORs for the specific outcomes. Where possible, pooled adjusted ORs (AORs) were also estimated. A p-value less than 0.05 was considered statistically significant. Since we anticipated considerable between-study heterogeneity, a t random-effects model was applied in case of raw data using the Mantel-Haenszel [21,22] method by metabin function from meta package, and we applied the Paule-Mandel method to estimate the between-study variance [23]. Other cases we used already calculated crude or adjusted ORs and their standard errors for pooling with restricted maximum likelihood methods [24] with the help of metagen function. Heterogeneity was tested with I 2 and Cochran Q tests; p < 0.1 indicated significant heterogeneity [25].
Following the recommendations of IntHout et al. (2016), where applicable, we also reported the prediction intervals (i.e., the expected range of effects of future studies) of the pooled estimates.
There were less than 10 studies for each outcome; therefore, a publication bias assessment could not be performed.

Risk of Bias Assessment and Quality of Evidence
The risk of bias assessment was performed by two review authors (Á.M., B.P.) independently using the Quality in Prognosis Studies (QUIPS) tool [26]. Disagreements were resolved by a third review author (B.T.). We visualized our results in figures with the Riskof-bias VISualization (robvis) tool [27]. The overall assessment was considered low-risk if only low-risk bias domains existed. If one domain was at high risk or three out of six domains were at moderate risk, the overall result was reported as having a high risk of bias. The quality evaluation of the evidence was performed following the recommendations of the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) workgroup. We prepared the summary of findings tables with the GRADEPro Guideline Development Tool [28].

Basic Characteristics of Included Studies
Baseline characteristics of the enrolled analyses are shown in Table 1. We included 13 case-control studies [8,[29][30][31][32][34][35][36][37][38][39][40][41], and one retrospective cohort analysis [33]. The overall study period of the articles is placed between 1960 and 2020. The largest database included in the meta-analysis was the Hungarian Case-Control Surveillance of Congenital Abnormalities [29,34]. The diagnosis of influenza infection was in most cases based on the symptoms of the disease by using retrospective questionnaires. There was only one study

Basic Characteristics of Included Studies
Baseline characteristics of the enrolled analyses are shown in Table 1. We included 13 case-control studies [8,[29][30][31][32][34][35][36][37][38][39][40][41], and one retrospective cohort analysis [33]. The overall study period of the articles is placed between 1960 and 2020. The largest database included in the meta-analysis was the Hungarian Case-Control Surveillance of Congenital Abnormalities [29,34]. The diagnosis of influenza infection was in most cases based on the symptoms of the disease by using retrospective questionnaires. There was only one study in which the disease was confirmed by a polymerase chain reaction (PCR) test [33]. In most of the studies, the age of the mothers was between 20-34 years. The summary table of the outcomes of the included studies was shown also in the Supplementary Material (Table S3).

Non-Chromosomal Congenital Malformations
Only one article reported a total number of non-chromosomal congenital malformations [30]. On the basis of their results, first-trimester maternal influenza was associated with 1.5 times increased odds (CI: 1.30-1.70) of developing all types of non-chromosomal birth defects.

Non-Chromosomal Congenital Malformations
Only one article reported a total number of non-chromosomal congenital malformations [30]. On the basis of their results, first-trimester maternal influenza was associated with 1.5 times increased odds (CI: 1.30-1.70) of developing all types of non-chromosomal birth defects.

Non-Chromosomal Congenital Malformations
Only one article reported a total number of non-chromosomal congenital malf mations [30]. On the basis of their results, first-trimester maternal influenza was asso ated with 1.5 times increased odds (CI: 1.30-1.70) of developing all types of non-chrom somal birth defects.

Congenital Heart Defects
Several studies [33,36,37,42] provided data for congenital heart de found 1.62 times higher odds (CI: 1.17-2.24), and Czeizel et al. [34] rep higher odds (CI: 1.50-2.40) for developing all types of congenital hear studies also found a significant association between influenza and ao septal and ventricular septal defects. Odds ratios for the subtypes of con fects are presented in Table 2.

Congenital Heart Defects
Several studies [33,36,37,42] provided data for congenital heart defects. Li et al. [8] found 1.62 times higher odds (CI: 1.17-2.24), and Czeizel et al. [34] reported 1.90 times higher odds (CI: 1.50-2.40) for developing all types of congenital heart defects. Several studies also found a significant association between influenza and aortic, conotruncal, septal and ventricular septal defects. Odds ratios for the subtypes of congenital heart defects are presented in Table 2. Three studies [8,32,37] reported adjusted odds ratios (AORs). The pooled results showed a more than 1.5 times increase in the odds of developing congenital heart defects (AOR: 1.63, CI: 1.27-2.09; I 2 = 0%, CI: 0-90%; Figure 5). Three studies [8,32,37]reported adjusted odds ratios (AORs). The pooled showed a more than 1.5 times increase in the odds of developing congenital heart (AOR: 1.63, CI: 1.27-2.09; I 2 = 0%, CI: 0-90%; Figure 5). We also found a positive association between influenza and specific types of ital heart defects when the analysis was performed with multiple logistic regressi results of the studies showed an almost fourfold increase in the development o coarctation, an almost threefold increase in the development of left-sided obs heart defects and a two-and-a-half-fold increase in the development of right-si structive heart defects. The findings of the studies showed an almost threefold inc the development of atrial septal defects, a twofold increase in the development of ular septal defects, more than one-and-a-half-fold increase in the development o truncal defects. These results are shown in Table 3.

Odds Ratio
Lower with influenza Higher with influenza Figure 5. Adjusted odds ratio of developing congenital heart defects after influenza infection in the first trimester [8,32,37].
We also found a positive association between influenza and specific types of congenital heart defects when the analysis was performed with multiple logistic regression. The results of the studies showed an almost fourfold increase in the development of aortic coarctation, an almost threefold increase in the development of left-sided obstructive heart defects and a two-and-a-half-fold increase in the development of right-sided obstructive heart defects. The findings of the studies showed an almost threefold increase in the development of atrial septal defects, a twofold increase in the development of ventricular septal defects, more than one-and-a-half-fold increase in the development of conotruncal defects. These results are shown in Table 3. Table 3. Detailed odds ratios for developing specific congenital heart defects after influenza in the first trimester.

Risk of Bias Assessment and Level of Evidence
The evaluation was based on six criteria. Every domain was scored as carrying low, moderate or high risk of bias for all the included outcomes. The results of the risk of bias assessment are presented in the Supplementary Material (Figures S4-S14). As the prognostic factor measurement was not carried out with PCR and the chances of recall bias were high, the overall risk of bias was high in most of the included studies. Also, the quality of evidence was very low for most of the results (Table S5).

Discussion
The importance of viral infections will increase as we face growing risks of pandemics, which may affect the pregnant mother and the fetus [42]. Viral infections during pregnancy raise several clinical challenges, including adverse pregnancy outcomes and birth defects in the offspring [3]. During embryogenesis, the development of different organs and organ systems occurs one after the other and parallel to others. Almost every organ system develops in the first 12 weeks of pregnancy, so any effect on the embryo during this period may significantly influence the development of birth defects.
This systematic review and meta-analysis evaluated the effects of influenza infection in the first trimester of pregnancy on the development of non-chromosomal birth defects.
We know from a previous study by Ács et al. [30] that the odds of developing all types of non-chromosomal birth defects are 1.5 times higher if influenza is present in the first trimester (OR: 1.40; CI: 1.30−1.60). Oster et al. [43] got the same association (OR: 1.11; CI: 0.91−1.35) between influenza and birth defects in the USA. On the basis of the results of this meta-analysis, we could also confirm the association between first-trimester maternal influenza and the development of non-chromosomal congenital malformations. Moreover, our results showed increased odds of developing three major types of CAs, namely congenital heart defects, neural tube defects and oral clefts after influenza infection in the first three months of pregnancy.
Neural tube defects are congenital brain, spine, or spinal cord abnormalities. Many studies [44][45][46] have investigated the pathogenesis of NTDs. Among the epidemiology and risk factors chromosomal anomalies (trisomy 13, 18 and triploidy) represent less than 10% of all NTDs [47]. Maternal diabetes, obesity, maternal hyperthermia (use of sauna, hot water tube, and fever), medication (valproate) and low folate intake were also considered as risk factors for developing NTDs [48]. The quantitative and qualitative synthesis of our findings in the present study confirmed the association between influenza and neural tube defects [34][35][36]. Their significance is outstanding as they are amongst the most common causes of genetic abortions. Most importantly, we can decrease the development of these congenital malformations with folic acid supplementation [49].
Our findings also confirmed the association between the development of cleft lip and palate and first-trimester maternal influenza. Orofacial clefts are common birth defects, estimated at 1.5 per 1000 live births worldwide [50]. Newborns with these anomalies have feeding difficulties and often develop conductive hearing loss and speech problems [51]. Environmental agents and genetic factors were also linked with the development of these birth defects. Moreover, orofacial clefts were also associated with teratogenetic agents, such as alcohol or anticonvulsant drugs. However, folic acid deficiency can also increase the chances for these birth defects [52,53]. In a population-based case-control study, L. Ács et al. [29] confirmed that influenza is a risk factor for developing cleft palate like other lifestyle factors (gender, birthweight and smoking) or illnesses. Waller et al. [7] examined common cold/flu during the periconceptional period, and they assessed that the odds of developing a cleft lip was 1.23 (CI: 1.05−1.45).
Congenital heart defects are among the most prevalent birth defects, and they remain an important causative factor of neonatal and infant deaths. The risk of these birth defects was associated with low maternal education, pregestational diabetes, self-reported maternal clotting disorders, prescriptions for anticlotting medication [43] and maternal influenza infection [43,54]. Our results could prove the association between influenza infection and congenital heart disease [8,34].
Organogenesis takes place during the first 12−13 weeks of pregnancy. This time-frame is the so called critical period for the development of birth defects. Therefore, our metaanalysis focused only on the first three months of pregnancy. We identified several studies, which analyzed the periconceptional period (i.e., three to one months before pregnancy and the first trimester) [8,44,55]. These results do not differ significantly from the results we have obtained. Further studies have shown that if fever during an influenza infection was relieved effectively with antipyretic drugs, malformations are less likely to develop [30]. Several studies have also shown that taking folic acid and maternal vitamins during the first trimester of pregnancy can reduce the odds of developing birth defects [30,56].

Strengths and Limitation
To our knowledge, this is the first meta-analysis that investigated the association between maternal influenza and all types of non-chromosomal abnormalities during the first trimester. We followed our protocol, which was registered previously. Our investigation covered a long study period with a large number of cases. The articles were from all over the world, giving a comprehensive picture of the disease and its effects.
In consideration of the limitations of this analysis, most studies defined the diagnosis of influenza based on the symptoms (headache, fever, running nose and muscle pain) and reported data that were retrospective and self-reported. Thus, the risk of bias was high because of the prognostic factor measurement. Most of the studies included were case-control investigations, where a high percentage of recall bias is to be expected. We could include only one study in which a serological test identified the disease. Because of the rigorous methodology and timing, there were few cases for one outcome.

Implications for Practice and Research
Translating scientific results into everyday practice has crucial importance [56]; therefore, we should make efforts to prevent influenza during pregnancy and highlight the importance of vaccination against flu and antipyretic therapy. We suggest that if influenza is presented in the critical pregnancy period, great emphasis should be placed on prenatal screening. Additional prospective observational studies should be performed with the use PCR to confirm first-trimester influenza infection.

Conclusions
In conclusion, our systematic review found that influenza in the first trimester is associated with non-chromosomal birth defects, primarily congenital heart defects, neural tube defects and oral clefts. Therefore, influenza prevention by vaccination before or during pregnancy is highly recommended. Moreover, if influenza has already manifested itself, proper antipyretic treatment should be started.

Supplementary Materials:
The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/v14122708/s1, Figure S1: Forest plot representing that the odds for developing specific types of neural tube defects was increased after influenza in the first trimester. Figure S2. Forest plot representing that the adjusted odds for developing limb reduction defects was increased after influenza in the first trimester. Figure S3. Forest plot representing that the adjusted odds for developing eye anomalies was increased after influenza in the first trimester. Figure S4. Risk of bias assessment at study and at domain level for all neural tube defects. Figure S5. Risk of bias assessment at study and at domain level for neural tube defects in the systematic review. Figure S6. Risk of bias assessment at study and at domain level for spina bifida. Figure S7. Risk of bias assessment at study and at domain level for cleft lip and palate. Figure S8. Risk of bias assessment at study and at domain level for cleft lip, palate. Figure S9. Risk of bias assessment at study and at domain level for congenital heart disease in the systematic review (adjusted). Figure S10. Risk of bias assessment at study and at domain level for congenital heart disease (adjusted). Figure S11. Risk of bias assessment at study and at domain level for congenital heart disease in the systematic review (unadjusted). Figure S12.
Risk of bias assessment at study and at domain level for congenital heart disease (unadjusted). Figure S13. Risk of bias assessment at study and at domain level for eye anomalies. Figure S14. Risk of bias assessment at study and at domain level for limb reduction. Figure S15. Search key of the systematic search. Table S1. PRISMA 2020 checklist. Table S2. MOOSE Checklist for Meta-analyses of Observational Studies. Table S3. Table of included studies with outcomes. Table S4. The odds ratios for developing specific types of congenital malformations after first trimester. Table S5. Quality of evidence.
Author Contributions: Á.M.: conceptualization, project administration, writing-original draft; B.T.: project administration, conceptualization, methodology, writing-review and editing; A.V.: formal analysis; visualization; P.H.: project administration, methodology, writing-review and editing; B.P. data curation, review, editing; A.F.: data curation; conceptualization, writing-review and editing; S.V.: conceptualization, writing-review and editing; B.L.: conceptualization, writing-review and editing; Z.M.: conceptualization, writing-review and editing; N.Á.: conceptualization; supervision; writing-original draft. All authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: No ethical approval was required for this systematic review with meta-analysis, as all data were already published in peer-reviewed journals. No patients were involved in the design, conduct, or interpretation of our study.

Data Availability Statement:
The datasets used in this study can be found in the full-text articles included in the systematic review and meta-analysis.

Conflicts of Interest:
The authors declare no conflict of interest.