Draft Manuscript For Review. Reviewers should submit their review at http://mc.manuscriptcentral.com/humrep
Influenza and congenital anomalies: a systematic review
and meta-analysis
Journal: Human Reproduction
Manuscript ID: HUMREP-13-1069.R1
Manuscript Type: Meta-Analysis
Date Submitted by the Author: 12-Nov-2013
Complete List of Authors: Luteijn, Johannes; University of Ulster, School of Nursing Brown, Mary; University of Ulster, School of Nursing Dolk, Helen; University of Ulster, School of Nursing
Keywords: Influenza, Congenital Anomalies, Meta-analysis, Observational Studies,
Public Health
Specialty: Reproductive Epidemiology
Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online.
Luteijn_eAppendix Figures 12-51.zip
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Please find our responses to the comments made by the reviewers below. Reviewer comments in bold 1
green, our responses in red. 2
Reviewer: 1 3
Comments to the Author 4
This is an interesting manuscript about the association between influenza infection during the first 5
trimester of pregnancy and congenital anomalies. 6
7
The paper is well written. I only have relatively minor but still important comments. 8
1. Search strategy: Please add that MESH terms were used. 9
The fact that the search terms for PubMed and Embase databases were MeSH terms was added. Line 10
89. 11
2. Explain what AOR is in the M&M. 12
The following sentence was added in the methods section: “Meta-analysis was performed combining 13
adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for confounders).” Lines 132-133. 14
3a. Please state in the M&M that you combine OR and AOR. You state in the discussion that 15
adjustment for confounders showed a moderate effect on OR in studies that did report both 16
crude OR and adjusted OR. 17
Initially we did not combine OR and AOR as we extracted 2x2 tables and solely calculated crude ORs. 18
The sentence that was referring to the “moderate effect” was referring to the effect witnessed within 19
studies, rather than aggregate meta-analysis results. Based on comment #2 and comment #3 of 20
reviewer 1, it was decided to redo the meta-analysis on a combination of crude and adjusted ORs. This 21
had a number of consequences: 22
1. We are now able to include the study by Oster (Oster et al. 2011) in the meta-analysis. This 23
was not possible earlier since we did not get a reply from Oster et al to our request for the 24
core 2x2 tables which we required to calculate crude ORs. 25
Page 1 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
2. Since we now use AORs/ARRs, rather than ORs/RRs, whenever possible (and we include 26
Oster), slight changes to overall pooled meta-analysis estimates have occurred. For example, 27
our estimate for “all non-chromosomal CA” went from 2.25 (95% CI: 1.77-2.85) to 2.00 (1.62-28
2.48). 29
The following sentence was added in the methods section: “Meta-analysis was performed combining 30
adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for confounders). Only four case-control studies 31
provided AORs (Table IV) and none of the cohort studies made statistical adjustments (Table V). 32
Where adjusted estimates were not available, crude estimates were used. ” Lines 132-135. 33
3b. Please do add a subgroup analysis combining only OR and AOR. You do not need to add all 34
figures in my view but you can add a statement in the results section. 35
Solely three studies reported AOR and either crude OR or means of calculating the crude OR.(Botto et 36
al. 2001, Li et al. 2007, Lynberg et al. 1994) Granroth (Granroth et al. 1978) did report OR (2.5) and 37
AOR (1.7), however due to the lack of confidence intervals, we were unable to use Granroths AOR in 38
meta-analysis. Oster solely reported AOR. 39
We decided to do a subgroup analysis comparing adjusted estimates versus crude estimates for all 22 40
studies enrolled in the meta-analysis and enrolling Botto, Li and Lynberg in both categories (with 41
corresponding AOR and OR, respectively). The adjusted estimate pool consisted of solely 4 studies 42
(Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994, Oster et al. 2011), half of which reported 43
exclusively on NTDs. 44
The following statement was added in the results section “No differences were detected between 45
pooled adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77).” Lines 210-211. 46
Results of this subgroup analysis are not included in Table I. 47
4. Can you translate the OR in an absolute value in text or table, at least for the main outcomes? 48
It would be valuable to know how this translates as an RR can imply an increase from 40 to 80% 49
Page 2 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
or an increase from 0.1% to 0.2%. For instance compared to a risk in the control population of 50
1% the risk after influenza was found to be between 2 and 3%. 51
We recognize the importance of describing public health implications of findings. The following 52
sentence was added at the beginning of the discussion: “The twofold increase in risk of non-53
chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT Central Registry 2013) to 54
3.6% of births among 1st
trimester influenza exposed pregnancies.” Lines 264-266. 55
The EUROCAT congenital anomaly prevalence tables cited here allow for fast calculation of absolute 56
value effects of (A)ORs reported for any congenital anomaly in the systematic review and/or meta-57
analysis. 58
5. Data were pooled when at least 3 studies were available (page 7 line 129). In table II CA's are 59
described that are not included in the meta-analyses but I notice that some of the anomalies 60
are described by three or more studies. 61
This is correct. Not all studies included in the systematic review were enrolled in the meta-analysis. For 62
example, all ecological studies were excluded from meta-analysis since ecological study outcomes 63
cannot be combined into a summary estimate with outcomes of other studies. Some estimates 64
reported in table 2 are from the BWIS study for which we were unable to obtain data after contacting 65
the authors. 66
For additional clarification, “Note for some CA, ≥3 estimates were available from studies that could 67
not be included in the meta-analysis.” was added below Table II. 68
Page 3 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Reviewer: 2 69
Comments to the Author 70
The authors pose the question whether first trimester maternal influenza infection increases the risk of 71
non-chromosomal anomalies. Overall, the systematic review was methodical and the study well 72
written. The study question is clearly formulated and answerable in the context of a systematic 73
review. The search for applicable studies was extensive and the selection process only biased by study 74
reporting in PubMed or EMBASE and publication in English or Dutch. Authors of prior studies were 75
contacted to obtain unavailable data. Inclusion and exclusion criteria for the selected studies were 76
clearly stated and appropriate, particularly exclusion of chromosomal anomalies. The quality of the 77
studies selected and risk of bias was assessed using a modification of the Newcastle-Ottawa scale, 78
which is appropriate. Sources of bias were discussed in detail in the text. Limitations were 79
thoughtfully considered and thoroughly discussed in the conclusion. The risk of bias was well 80
explored in the discussion and sources of heterogeneity and weighting discussed. 81
82
I have only a few minor Comments: 83
1. Please check that all abbreviations are defined in the text. It is not clear if “CA” is defined 84
earlier in the text. Was CHD defined? 85
Congenital anomy (CA) is now defined in the main body of text on the first occurrence (Line 78). 86
Abbreviations for specific CA (CHD, CNS, NTD and VSD) were removed from the manuscript as it was 87
felt these abbreviations might be less common. 88
2. Please describe and define EUROCAT. 89
EUROCAT is now defined. EUROCAT is described as “EUROCAT is a network of population-based 90
congenital anomaly registries which surveys over 1.7 million births annually in 23 European 91
countries.” Lines 119-120. 92
3. The sentence on lines 180-181 is confusing. Please reword. 93
Page 4 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
The sentence has been reworded as follows: 94
“Overall, our meta-analysis involved 29,542 CA cases of which 1,112 were exposed to influenza in the 95
first trimester of pregnancy and 53,089 controls of which 1,382 were exposed to influenza in the first 96
trimester of pregnancy from case-control studies.” Lines 192-194. 97
4. Please reword the sentence on lines 266-268. 98
The sentence has been reworded as follows: 99
“However, as long as maternal reports are collected prospectively prospective to the mother being 100
aware of the malformation (e.g. from medical records or interviews during pregnancy), there is no 101
reason to believe misclassification of influenza exposure will differ between arms cases and controls., 102
and t Therefore prospective maternal reports will not lead to a spurious association, but rather bias 103
the estimate toward the null as cases and controls are subject to similar misclassification.” Lines 289-104
293. 105
5. In line 333, please insert “but” after the comma. 106
“but” was added in the corresponding sentence. Line 366. 107
Page 5 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
6
Associate Editor's comments to Author: 108
This is a carefully performed meta-analysis of an important clinical topic. I really do have nothing to 109
add to the two high quality review reports. 110
111
Finally, the format of abstracts published in Human Reproduction has recently changed (please refer 112
to the journal's 'Information for Authors'). Attached is a pro-forma for the new, extended version that 113
should be used in the resubmitted article. 114
Page 6 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
7
Botto LD, Lynberg MC and Erickson JD. Congenital heart defects, maternal febrile illness, and 115
multivitamin use: A population-based study. Epidemiology 2001:12:485-490. 116
EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013:2013. 117
Granroth G, Haapakoski J and Saxen L. Defects of the central nervous system in Finland: V. 118
Multivariate analysis of risk indicators. Int J Epidemiol 1978:7:301-308. 119
Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural 120
tube defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 121
Teratol 2007:79:295-300. 122
Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: a 123
population-based case-control study. Am J Epidemiol 1994:140:244-255. 124
Oster ME, Riehle-Colarusso T, Alverson CJ and Correa A. Associations between maternal fever and 125
influenza and congenital heart defects. J Pediatr 2011:158:990-995. 126
Page 7 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Manuscript title 1
Influenza and congenital anomalies: a systematic review and meta-analysis 2
Suggestion for a running title 3
Influenza and congenital anomalies 4
Authors full names 5
JM Luteijn. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 6
Ulster, Jordanstown Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 7
MJ Brown. Institute of Nursing Research/School of Nursing, University of Ulster, Jordanstown 8
Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 9
H Dolk. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 10
Ulster, Jordanstown Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 11
Page 8 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
Abstract 12
Study question: Does 1st
trimester maternal influenza infection increase the risk of non-13
chromosomal congenital anomalies? 14
Summary answer: 1st trimester maternal influenza exposure is associated with raised risk of a 15
number of non-chromosomal congenital anomalies, including neural tube defects, hydrocephaly, 16
congenital heart defects, cleft lip, digestive system defects and limb reduction defects. 17
What is known already: Hyperthermia is a well-established risk factor for neural tube defects. 18
Previous studies suggest influenza may be a risk factor not only for neural tube defects, but also 19
other congenital anomalies. No systematic review has previously been undertaken. 20
Study design, size, duration: Systematic review and meta-analysis. A search of EMBASE and PUBMED 21
was performed for English and Dutch studies published up to July 2013. A total of 33 studies (15 22
case-control, 10 cohort and 8 ecological) were included in the systematic review of which 22 studies 23
were included in the meta-analysis. 24
Participants/materials, settings, methods: A total of 29,542 babies with congenital anomaly (1,112 25
exposed) from case-control studies and 1,608 exposed pregnancies resulting in 56 babies with 26
congenital anomaly from cohort studies were included in the meta-analysis. Maternal influenza 27
exposure was defined as any reported influenza, influenza-like illness, or fever with flu, with or 28
without serological or clinical confirmation during the 1st trimester of pregnancy. Data for 24 29
(sub)groups of congenital anomaly available from ≥ 3 studies were analysed using the DerSimonian-30
Laird random effects model. The hypothesis of publication bias was assessed using funnel plots and 31
risk of bias of included studies was assessed using a slightly modified version of the Newcastle-32
Ottawa Scale. 33
Main results and the role of chance: 1st trimester maternal influenza exposure was associated with 34
an increased risk of any congenital anomaly (adjusted odds ratio 2.00, 95% CI: 1.62-2.48), neural tube 35
Page 9 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
defects (OR 3.33, 2.05-5.40), hydrocephaly (5.74, 1.10-30.00), congenital heart defects (1.56, 1.13-36
2.14), aortic valve atresia/stenosis (AOR 2.59, 1.21-5.54), ventricular septal defect (AOR 1.59, 1.24-37
2.14), cleft lip (3.12, 2.20-4.42), digestive system (1.72, 1.09-2.68) and limb reduction defects (2.03, 38
1.27-3.27). The increased risk for cleft lip (but not for cleft palate), was also reported by ecological 39
studies not included in the meta-analysis. Study outcomes reported for 27 subgroups of congenital 40
anomaly could not be included in the meta-analysis. Visual inspection of funnel plots did not suggest 41
evidence for publication bias. 42
Limitations, reasons for caution: This study enrolled observational studies which can be subject to 43
limitations such as confounding, retrospective maternal exposure reports and non-response of 44
intended participants. Influenza exposed pregnancies can also have been exposed to influenza 45
related medication. 46
Wider implications of the findings: Prevention of influenza in pregnant women may reduce 47
congenital anomaly risk, and would be relevant to more than just neural tube defects. More research 48
is needed to determine whether influenza and/or its related medication is teratogenic, to determine 49
the role of hyperthermia in teratogenicity and the role of other environmental factors such as 50
nutritional status in determining susceptibility. 51
Study funding/ competing interests: Funded by the EC, under the framework of the EU Health 52
Programme 2008-2013, Grant Agreement 2010 22 04 (Executive Agency for Health & Consumers). No 53
competing interests. 54
Page 10 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Introduction 55
Both during seasonal influenza and pandemic influenza outbreaks, pregnant women have been at 56
risk of increased morbidity and mortality from influenza infection compared to the general 57
population.(Dodds et al. 2007, Harris 1919, Neuzil et al. 1998, Siston et al. 2010) Women in the later 58
stages of pregnancy are particularly vulnerable to adverse health outcomes after influenza infection 59
(Mak et al. 2008), perhaps because of immunological changes that take place during 60
pregnancy.(Jamieson et al. 2006) 61
Unravelling the question of teratogenicity of influenza is complex. In observational studies, influenza 62
exposure can affect the foetus not only via viral infection of the foetus, influenza-induced 63
hyperthermia and toxic metabolites associated with fever (Edwards 2006), but also via antiviral and 64
antipyretic use. A recent systematic review found strong evidence of an association between 65
maternal hyperthermia and neural tube defects.(Moretti et al. 2005) Evidence on other anomalies 66
with respect to hyperthermia or fever is scarce. Animal models have associated maternal 67
hyperthermia with arthrogryposis (Edwards 1971), congenital heart defects (Cockroft and New 1975, 68
Cockroft and New 1978), club foot(Edwards 1971), microcephaly (Edwards 1969, Edwards 1969), 69
microphthalmos (Germain et al. 1985) and others.(Edwards 2006) 70
The primary method of protecting pregnant women and their unborn child against influenza 71
infection is vaccination. In an increasing number of countries, pregnant women are advised to be 72
vaccinated against seasonal influenza infection.(Mak et al. 2008, Mereckiene et al. 2010) However, 73
vaccination policies in European countries vary both for seasonal influenza vaccination and during 74
the 2009 H1N1 influenza pandemic, especially with respect to trimesters eligible for 75
vaccination.(Luteijn et al. 2011, Mereckiene et al. 2010) In the absence of consensus on whether or 76
not to vaccinate 1st trimester pregnant women, the hypothesis of a causal relationship between 77
congenital anomalies (CA) and influenza virus deserves renewed attention. A better understanding of 78
the possible relationship between influenza and CA will allow for better understanding of the 79
Page 11 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
benefit-risk balance of vaccinating 1st
trimester pregnant women and women of childbearing age 80
against influenza. 81
The objective of our review is to identify and summarize the available epidemiologic evidence 82
regarding the risk of CA associated with 1st
trimester exposure to maternal influenza. 83
Materials and methods 84
Search strategy 85
This systematic review was informed by PRISMA guidelines (Moher et al. 2009) and the MOOSE 86
group guidelines.(Stroup et al. 2000) Two of the authors (JML and MJB) conducted the various steps 87
of the review and resolved any disagreements by discussion and consensus. The PubMed® and 88
Embase® databases were searched using the MeSH terms ("Influenza, Human") AND (pregnancy OR 89
congenital abnormality) and (Influenza) AND (pregnancy OR congenital abnormality), respectively on 90
July 1st, 2013. No publication or date restrictions were set. Where the papers’ abstract, title or 91
indexed MeSH terms suggested the possibility of reporting any fetal outcomes after maternal 92
exposure to influenza, the full paper was obtained. Reference lists of enrolled papers were reviewed. 93
Eligibility criteria 94
Case-control, cohort and ecological studies investigating CA outcomes following maternal exposure 95
to influenza were eligible for inclusion. No quality criteria were set for inclusion, although risk of bias 96
analysis was performed (eAppendix). Influenza was defined as any reported influenza, influenza-like 97
illness, or fever with flu, with or without serological or clinical confirmation. 98
Solely studies reporting influenza exposures during the 1st trimester of pregnancy were included in 99
the systematic review and meta-analysis. In order to be included in the meta-analysis, case-control 100
and cohort studies needed to allow for the calculation or report odds ratios (OR) or relative risks 101
(RR). For financial reasons, only English and Dutch language papers were eligible for inclusion. No 102
Dutch papers were included. 103
Page 12 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
6
Data extraction 104
Study characteristics were extracted by JML and MJB (eAppendix, Table III-V). Crude and adjusted 105
ORs and RRs and 2x2 tables relevant for meta-analysis were extracted by JML and MJB from cohort 106
and case-control studies. We contacted three authors of studies published ≤10 years ago(Czeizel et 107
al. 2008, Kelly et al. 2012, Oster et al. 2011) to obtain core data in order to create aggregate groups 108
such as orofacial clefts and for crude OR calculation and received the complete dataset for one 109
study.(Czeizel et al. 2008) In case of studies distinguishing between flu with fever and flu without 110
fever where it was impossible to combine the two, such as the stratified study by Lynberg (Lynberg et 111
al. 1994), the flu with fever dataset was extracted. We extracted data regarding malformed controls, 112
if available, rather than non-malformed controls since use of malformed controls reduces the impact 113
of differential recall bias. We recognize this can lead to underestimation of effect size if CA in the 114
control group are related to influenza exposure. In the meta-analysis, influenza exposures outside of 115
the 1st
trimester were added to the non-exposed cohort and for one cohort study without controls 116
(Doll et al. 1960), this allowed us to form a control group. 117
CA were classified into European surveillance of Congenital Anomalies (EUROCAT) defined subgroups, 118
excluding minor anomalies as specified by EUROCAT.(EUROCAT Central Registry 2009) EUROCAT is a 119
network of population-based congenital anomaly registries which surveys over 1.7 million births 120
annually in 23 European countries. CA were classified down to the greatest level of precision possible. 121
For example, a study that reported anomalies only as neural tube defects without further 122
specification contributed data to the analysis of neural tube defects but could not contribute data to 123
an analysis of anencephaly or spina bifida specifically. We excluded chromosomal syndromes since 124
their etiology is not related to maternal exposures. 125
Risk of bias assessment 126
In order to assess the validity of included studies’ findings we assessed the risk of bias of case-control 127
and cohort studies included in the meta-analysis using a slightly modified version of the Newcastle-128
Page 13 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
7
Ottawa scale (NOS).(Wells et al. 2004) Judgement criteria and deviations from the NOS are 129
summarized in the eAppendix. 130
Quantitative data summary and synthesis 131
Meta-analysis was performed combining adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for 132
confounders). Only four case-control studies provided AORs (Table IV) and none of the cohort studies 133
made statistical adjustments (Table V). Where adjusted estimates were not available, crude 134
estimates were used. 135
We assumed similarity between OR and RR because CA are rare events.(Davies et al. 1998) Statistical 136
analysis was performed using Stata version 9.2 [StataCorp, College Station, TX]. Meta-analysis was 137
performed on all CA combined and, for EUROCAT defined subgroups of CA (if n studies ≥ 3). The 138
DerSimonian & Laird random effects model (DerSimonian and Laird 1986) was used since the studies 139
in this meta-analysis involved varying countries, time periods and influenza strains. Subgroup analysis 140
was performed based on study type, publication date, risk of differential recall bias and adjustment 141
for confounders in order to assess the impact of these variables on study outcome. Subgroup 142
analyses combined all studies in the relevant categories, using the estimate for all non-chromosomal 143
CA combined where available and if not available, the estimate for the specific CA subgroup studied. 144
Due to scarce numbers and imbalance between some study arms, we used an alternative continuity 145
correction based on the OR of (other) studies with >0 events in both arms and group ratio imbalance 146
as discussed by Sweeting et al.(Sweeting et al. 2004) Heterogeneity between studies was assessed 147
using the I-squared statistic. Values of I-squared equal to 25%, 50% and 75% were considered to 148
represent low, moderate and high levels of heterogeneity, respectively. The hypothesis of 149
publication bias was assessed using funnel plots (eAppendix, Figures 28-51). 150
Results 151
Selection flow 152
Page 14 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
8
The PubMed® database search yielded 1369 papers and the Embase® database search yielded 2649 153
papers (Figure 1). After removing 1121 duplicates, a total of 2897 potentially relevant papers were 154
identified by the literature search. After screening by MeSH terms, titles and abstracts, 2615 papers 155
were excluded and full papers were retrieved for the remaining 282 papers. Of these, a total of 40 156
papers covering 27 studies met the inclusion criteria and were included in the systematic review. 6 157
additional eligible papers were detected by reference tracking, leading to a grand total of 46 included 158
papers covering 33 studies. 159
Study characteristics 160
The 46 enrolled papers were classified as 25 papers covering 15 case-control studies, 12 papers 161
covering 10 cohort studies and 9 papers covering 8 ecological studies. Enrolled studies are 162
summarized by study type in eAppendix, Table III-V and evidence provided by enrolled studies that 163
was not included in the meta-analysis is summarized in Table II. For one case-control study 164
information was limited to a conference abstract.(Choi and Klaponski 1970) Included papers were 165
published between 1953 and 2013, with the median year of publication 1971. 166
Risk of bias assessment 167
Visual inspection of the funnel plots did not suggest evidence for publication bias (eAppendix, 168
Figures 28-51). 169
Of the 15 case-control studies (25 papers), 10 studies did not take into account possible confounding 170
by maternal age, socioeconomic class or both. In the majority of these 10 studies some form of 171
matching between cases and controls (usually maternal ward, sex and day of birth) took place. Ten 172
papers relied on retrospective maternal reported influenza episodes (or timing of maternal 173
interviews was unknown), making these studies susceptible to differential recall bias. Of the 174
remaining 5 studies, 2 used serologic confirmation and 3 used prospectively collected antenatal 175
records. The last notable source of possible bias was that in 5 case-control studies over 20% of the 176
Page 15 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
9
cases intended for inclusion were not enrolled, making these studies susceptible to non-response 177
bias (eAppendix). 178
For the 10 cohort studies (12 papers), none took into account possible confounding by maternal age, 179
socioeconomic class or both. For 9 cohort studies, infants were not followed up for at least a year (or 180
unclear), making these studies susceptible to misclassification bias for some CA not apparent at birth. 181
6 of the studies used prospectively collected maternal reports for exposure, 3 serologic confirmation, 182
for 1 study exposure ascertainment was not described. The last notable source of possible bias was 183
that in 5 cohort studies the exposed cohort was not drawn from a clearly defined place and time, or 184
failed to enrol ≥80% of the population identified in specified place and time, raising questions over 185
representativeness of the enrolled exposed cohort. 186
Quantitative data summary and synthesis 187
Meta-analysis was possible for data from 22 studies, forming groups of ≥3 independent studies for 24 188
(sub)groups of EUROCAT defined major CA: any non-chromosomal major CA, neural tube defects, 189
anencephaly, encephalocele, spina bifida, hydrocephaly, congenital heart defects, orofacial clefts 190
(Figures 2-9) and sixteen other CA subgroups (eAppendix, Figures 12-27). 191
Overall, our meta-analysis involved 29,542 CA cases of which 1,112 were exposed to influenza in the 192
first trimester of pregnancy and 53,089 controls of which 1,382 were exposed to influenza in the first 193
trimester of pregnancy from case-control studies. From cohort studies, 1,608 exposed pregnancies 194
resulting in 56 CA plus 14,613 non-exposed pregnancies resulting in 347 CA were enrolled. The 195
enrolled cohort studies were relatively small with only the Coffey and Jessop study reaching over 50 196
CA (including minor anomalies) following maternal influenza exposure. Case-control studies more 197
readily enrol the large number required for research on CA and 6 out of 15 case-control studies 198
enrolled over 500 cases.(Botto et al. 2001, Czeizel et al. 2008, Granroth et al. 1978, Laurence et al. 199
1968, Oster et al. 2011, Saxen 1975) The larger numbers come at a cost and 11 out of 15 case-control 200
studies gathered exposure data by retrospective maternal reports. 201
Page 16 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
10
Meta-analysis discovered statistically significant associations between 1st
trimester influenza 202
exposure and a large number of CA subgroups (Table I) including all non-chromosomal CA combined 203
(OR 2.00, 95% CI: 1.62-2.48). Medium heterogeneity was detected for the pooled estimate of all non-204
chromosomal CA (64%). Subgroup analysis showed lower odds ratios for pooled case-control study 205
outcomes (OR 1.84, 95% CI: 1.49-2.27), than for pooled cohort study outcomes (OR 2.12, 95% CI: 206
1.20-3.75) while pre-1970 studies reported higher odds ratios (OR 2.47, 95% CI: 1.50-4.70) than 207
studies published after 1970 (OR 1.71, 95% CI: 1.41-2.08). Overall, studies susceptible to differential 208
recall bias reported a lower risk (OR 1.92, 95% CI: 1.35-2.72) than studies not susceptible to 209
differential recall bias (OR 2.12, 95% CI: 1.54-2.92). No differences were detected between pooled 210
adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77). 211
Central nervous system defects 212
Associations were found for all neural tube defects (OR 3.33, 2.05-5.40) and the neural tube defect 213
subgroups anencephaly (OR 3.52, 1.69-7.32) and spina bifida (OR 2.20, 1.48-3.28). The majority of 214
the 2,500 neural tube defects were reported by Czeizel (n=1,202, AOR 2.40, 1.30-4.40), Li (344, AOR 215
3.06, 1.40-6.67), Laurence (n=551, OR 3.93, 1.37-11.27) and Lynberg (331, AOR 1.70, 1.10-2.50). The 216
lower OR reported by Lynberg is related to our preference for malformed controls over healthy 217
controls for the study by Lynberg, which lead to more conservative estimates. The study by Lynberg 218
reports higher AOR when using healthy controls for neural tube defects (AOR 3.0, 1.9-4.7). We 219
discovered significant heterogeneity in the aggregate groups and neural tube defects, anencephaly, 220
encephalocele and hydrocephaly (Table I). The heterogeneity for neural tube defects, anencephaly, 221
encephalocele and hydrocephaly seems to be driven by the Coffey, Hirvensalo, Pleydell, Saxen (1960) 222
and Wilson studies which all contributed OR of >10 in at least one CA subgroup. The ecological study 223
on 1957 pandemic influenza by Hakosalo, enrolling 27 neural tube defects reported suggestive 224
evidence for a relationship between influenza and neural tube defects, but two larger ecological 225
studies by Leck (n=2,484 (Leck et al. 1969) and n=162 (Leck 1963)) did not find such evidence for any 226
neural tube defects subgroup. The study by Hakosalo calculated cases back to last menstrual period, 227
Page 17 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
11
while none of the studies by Leck had access to gestational length, making these studies susceptible 228
to misclassification of exposure introduced by assumptions around gestational age. Of all 8 ecological 229
studies included, only three corrected for gestational age.(Busby et al. 2005, Hakosalo and Saxen 230
1971, Saxen et al. 1990) 231
Orofacial clefts 232
Our meta-analysis found an association between orofacial clefts and 1st
trimester influenza exposure 233
(OR 1.96, 95% CI: 1.33-2.91). There was a significant association for cleft lip with or without palate 234
(OR 3.12, 2.20-4.42), but not for cleft palate (OR 1.05, 0.60-.1.84) and no heterogeneity was detected 235
in these pooled groups (Table II). The majority of the orofacial clefts (n=2,773) were reported by 236
Czeizel (n=1,956) and Saxen (1975a n=591 and 1975b, n=194) and these studies report ORs between 237
1.90 and 2.32. Four of the ecological studies, all by Leck, also reported the association between 238
orofacial clefts and influenza and two of these four studies reported associations for cleft lip ± cleft 239
palate, but not for isolated cleft palate.(Leck 1963, Leck et al. 1969) 240
Congenital heart defects 241
Our meta-analysis found an association between congenital heart defects and 1st trimester influenza 242
exposure (OR 1.56, 95% CI: 1.13-2.14). The vast majority of the congenital heart defects reported in 243
the meta-analysis were reported by Botto (n=829, AOR 2.1, 0.8-5.5), Czeizel (n=4,479, OR 1.6, 1.3-1.9) 244
and Oster (n=2,361, AOR 1.11, 0.91-1.35). It should be noted the study by Botto suffered from a 2-12 245
year delay between delivery and maternal interview while influenza exposure in the study by Oster 246
could occur from 3 months before pregnancy to the 3rd
month of pregnancy. 247
With respect to specific types of congenital heart defects (Table I), meta-analysis showed aortic valve 248
atresia/stenosis and ventricular septal defect to be associated with 1st
trimester influenza exposure 249
(OR 2.59, 1.21-5.54 and OR 1.59, 1.24-2.04, respectively). No associations were found for atrial septal 250
defect, hypoplastic left heart and transposition of the great vessels. Four congenital heart defect 251
subtypes were only reported by ≤2 studies and therefore not in the meta-analysis (Table II). Of these, 252
Page 18 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
12
two studies showed a consistent absence of association for tetralogy of Fallot, and a consistent and 253
high association for tricuspid atresia and stenosis (Botto 7.9, Oster 6.04). 254
Other anomalies 255
Limb reductions were associated with 1st
trimester influenza exposure by the meta-analysis (OR 2.03, 256
1.27-3.27) and this association is supported by several ecological studies (Table II). An association 257
with anophthalmia/microphthalmia is based on a single study (Table II). There was evidence that 258
there is no association with influenza for hypospadias (OR 1.02, 0.75-1.39) and clubfoot (OR 1.03, 259
0.83-1.27). 260
Discussion 261
This systematic review provides an overview of the published evidence on influenza exposure during 262
the 1st
trimester of pregnancy and CA. Meta-analysis revealed evidence for increases in a wide range 263
of major CA following 1st
trimester influenza exposure. The twofold increase in risk of non-264
chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT Central Registry 2013) to 265
3.6% of births among 1st
trimester influenza exposed pregnancies. 266
Exposure ascertainment 267
Case-control and cohort studies utilized serologic confirmation and maternal reports (prospective 268
and retrospective) for influenza exposure ascertainment. During influenza season the positive 269
predictive value of persons presenting with ILI for influenza is in the order of 66-77%.(Monto et al. 270
2000, Zambon et al. 2001) Serologic confirmation detects clinical and subclinical infections, which 271
might lead to different results since subclinical infections might affect the pregnant women 272
differently from clinical infections. Arguably, serologic confirmation of exposure is more reliable than 273
maternal reports. Five studies based on serologic confirmation were enrolled in the systematic 274
review (Elizan et al. 1969, Hardy et al. 1961, Walker and McKee 1959, Warrell et al. 1981, Wilson et al. 275
1959, Wilson and Stein 1969) of which three were included in the meta-analysis.(Hardy et al. 1961, 276
Warrell et al. 1981, Wilson et al. 1959, Wilson and Stein 1969) The two serologic studies limited to 277
Page 19 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
13
systematic review did not detect an association between influenza and neural tube defects (Elizan et 278
al. 1969) and did not detect an increased prevalence of CA among 1957 H2N2 Asian pandemic 279
influenza exposed pregnancies.(Walker and McKee 1959) The three serologic studies in the meta-280
analysis combined contributed 53 out of 27,584 CA, and it was therefore not possible to examine the 281
effect of exposure ascertainment method on effect size. One of these studies reported a possible 282
association between CA and 1957 H2N2 Asian pandemic influenza (Hardy et al. 1961) while a second 283
study did not.(Wilson et al. 1959, Wilson and Stein 1969) The third study did not detect an 284
association between CA and neural tube defects.(Warrell et al. 1981) Note that all studies using 285
serologic confirmation were limited to low numbers. 286
It is apparent that maternal reports lead to misclassification of exposure as women might not recall 287
infection, timing of infection relative to pregnancy or misdiagnose another infection for influenza. 288
However, as long as maternal reports are collected prospective to the mother being aware of the 289
malformation (e.g. from medical records or interviews during pregnancy), there is no reason to 290
believe misclassification of influenza exposure will differ between cases and controls. Therefore 291
prospective maternal reports will not lead to a spurious association, but rather bias the estimate 292
toward the null as cases and controls are subject to similar misclassification. Retrospective maternal 293
reports (e.g. interviews after birth), which are frequently utilized by case-control studies, are 294
susceptible to recall bias where mothers of cases have a different motivation to recall early 295
pregnancy exposure than mothers of non-cases. Mothers may differ not only in their tendency to 296
remember the infection, but in their tendency to misinterpret an illness as influenza.(MacKenzie and 297
Houghton 1974) Some of the studies included in the systematic review give an estimate of the 298
differential recall bias’ effect. For example, in the study by Lynberg, the OR for anencephaly after flu 299
with fever decreased from 3.1 (95% CI: 1.6-6.1) when compared with non-malformed controls to 1.4 300
(95% CI: 0.7-2.6) when compared with malformed controls.(Lynberg et al. 1994) Part of this decrease 301
might also have been related to CA in the control group being related to influenza, thus biasing the 302
OR towards 1. In our meta-analysis, the pooled estimate of OR for studies susceptible to recall bias 303
Page 20 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
14
(1.92) was slightly lower than the pooled estimate for other studies (2.00), contrary to expectation. 304
This is related to the low overall OR (1.11) reported by the large Baltimore-Washington Infant Study 305
(Oster et al. 2011), which was susceptible to differential recall bias and contributed over 20% to the 306
pooled estimate of susceptible studies. 307
Case-control, cohort and ecological study designs 308
Cohort studies failed to enrol large numbers of CA and this should not be surprising considering CA 309
only make up for 2-3% of births in a general population and short follow-up time after birth could 310
have led to underascertainment. Investigation of specific CA requires even higher numbers. Due to 311
the possibility of enrolling larger numbers, case-control studies seem better suited for addressing the 312
hypothesis of teratogenicity of influenza, although this comes at a cost as most case-control studies 313
ascertained exposure by retrospective maternal reports. 314
Ecological studies are generally considered a weaker study design than case-control or cohort studies 315
(Evans 2003) due to lack of individual exposure information. It cannot be verified that any excess CA 316
occurred among infected individuals. Correlation between influenza and confounding risk factors at 317
group level may lead to ecological fallacy, for example if influenza and nutritional deficiencies co-318
occur in winter. Ecological studies base exposure status on timing of pregnancy relative to influenza 319
season (or a proxy thereof) and therefore, the cohort defined as “exposed” is diluted by pregnancies 320
that did not have influenza. Population influenza exposure in the eight enrolled ecological studies 321
were derived from influenza incidences, counts or deaths (n=5) and sickness absenteeism rates or 322
claims (n=3). For this reason, the distinguishing power of ecological studies is highly dependent on 323
influenza attack rates and precision used to define influenza exposure. 324
The advantages of ecological studies are that large numbers of patients are enrolled easily and 325
ecological studies are not susceptible to the exposure misclassification or recall bias inherent in 326
individual level studies. Furthermore, they can be free of individual level confounding e.g. if those 327
most susceptible to influenza in the population have other risk factors for CA. Ecological studies 328
Page 21 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
15
therefore offer great value for addressing the hypothesis of teratogenicity of infectious diseases and 329
should not be discounted at the bottom of the evidence hierarchy in this area of research. 330
Consistency between study designs lends strength to a causal interpretation.(Hofler 2005) 331
Associations between 1st trimester influenza exposure and CA 332
One of the most striking results of the meta-analysis is the association between 1st
trimester 333
influenza exposure and neural tube defects. Neural tube defects are easily recognized at birth 334
eliminating susceptibility to underascertainment, while a previous meta-analysis reported an 335
association between neural tube defects and hyperthermia (OR 1.92, 95% CI: 1.62-2.29).(Moretti et 336
al. 2005) There also is evidence for neural tube defects following hyperthermia exposure in guinea 337
pigs.(Smith et al. 1992) In more recent human studies, underascertainment of neural tube defects 338
could have been a problem due to terminations of pregnancy for fetal anomaly. According to a 2004 339
EUROCAT analysis, 88% of neural tube defects are detected prenatally of which 88% are 340
aborted.(Boyd et al. 2008) Neural tube defects were more frequently studied than other CA, possibly 341
a result of interest in the 1950s studies by Coffey, suggesting an alarmingly strong link between 342
maternal influenza exposure and neural tube defects (data corresponds to OR 10.58, 4.30-26.02 in 343
the 1963 follow-up).(Coffey and Jessop 1963) This study utilized standardized questionnaires for 344
maternal interview after delivery and therefore was susceptible to differential recall bias, but this 345
limitation would not explain such a high OR. One of the ecological studies found an increase in neural 346
tube defects during the 1957 Asian influenza outbreak in Finland and concluded this might have been 347
caused either by influenza or influenza-related pharmaceuticals.(Hakosalo and Saxen 1971) The study 348
by Li et al had data available both on antiviral and antipyretic use (and other potential confounders, 349
see Table IV) and after adjusting for these co-exposures, OR for neural tube defects following 350
maternal influenza exposure dropped slightly but remained statistically significant (OR 3.93, 95% CI: 351
2.48-6.23). 352
Page 22 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
16
Hydrocephaly can sometimes be caused by spina bifida, and we could derive estimates for 353
hydrocephaly not associated with neural tube defects for two studies of the five studies on 354
hydrocephaly.(Coffey and Jessop 1959, Coffey and Jessop 1963, Hirvensalo and Kinnunen 1962) It is a 355
general problem in CA research that one baby may have more than one anomaly, and some of these 356
multiple malformed babies may have “sequences” (EUROCAT Central Registry 2009) that follow from 357
a primary anomaly. The data available for review cannot distinguish different types of diagnoses. 358
The evidence for an association between 1st
trimester influenza exposure and cleft lip with or 359
without palate is strong due to the lack of heterogeneity in the pooled data and consistent positive 360
associations detected across different study designs. It is well known that cleft lip +/- palate differs 361
aetiologically from cleft palate, so this difference is not surprising.(Mossey et al. 2009) The 362
associations detected for congenital heart defects and ventricular septal defects are somewhat 363
contradicted by the results from the BWIS.(Oster et al. 2011) On the other hand congenital heart 364
defects have been associated with hyperthermia in rats.(Cockroft and New 1975, Cockroft and New 365
1978) Club foot has been associated with hyperthermia in guinea pigs (Edwards 1971), but our meta-366
analysis did not find evidence for an association between club foot and influenza. A large number of 367
additional associations for other CA types were detected with more limited underlying evidence. 368
Pathways for mediation of hypothetical teratogenic effect of influenza 369
Influenza can mediate a possible teratogenic effect via multiple pathways and there is a risk of 370
confounding due to the intimate linkage between a disease and its cure. As well as antivirals, 371
antipyretics are also often used during influenza infection and the case-control study by Li et al 372
reported an AOR for antipyretic drugs and neural tube defects of 4.86 (95% CI: 1.33-17.78).(Li et al. 373
2007) Associations not adjusted for antivirals or antipyretics remain of importance since from a 374
vaccination policy perspective, it is less relevant whether the influenza virus or the antivirals are 375
causing any possible anomalies as vaccination will prevent both exposures. We recognize this puts 376
Page 23 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
17
limits on generalizability of study findings between populations with different use of 377
antipyretics/antivirals. 378
Direct pathways via which influenza infection can possibly lead to CA are toxic metabolites caused by 379
fever, hyperthermia and the influenza virus crossing the placenta. Hyperthermia has been associated 380
with causing neural tube defects as discussed above.(Moretti et al. 2005) It should be noted this 381
meta-analysis involved a large number of possible causes of hyperthermia, while influenza causes 382
high fever which might be different from general hyperthermia. This could explain the higher OR 383
reported in this meta-analysis for influenza exposure and neural tube defects (Table I). Several of the 384
included studies distinguished between 1st
trimester influenza and 1st
trimester fever.(Aro 1983, 385
Botto et al. 2001, Klemetti 1977, Lynberg et al. 1994, Oster et al. 2011, Saxen 1975, Saxen 1975) One 386
study found an association for influenza, but not for fever (Klemetti 1977), another found an 387
association for influenza with fever and neural tube defects (OR 1.7, 1.1-2.5), which was lowered for 388
influenza without fever (OR 1.3, 0.7-2.5).(Lynberg et al. 1994) A study on congenital heart defects 389
reported associations for fever (OR 1.8, 1.4-2.4) and influenza (OR 2.1, 0.8-5.5).(Botto et al. 2001) 390
while a second study on congenital heart defects reported very similar low and non-significant 391
excesses for fever (OR 1.14, 0.89-1.46) and influenza (OR 1.11, 0.91-1.35).(Oster et al. 2011) The two 392
studies by Saxen on orofacial clefts reported comparable associations for influenza (RR 2.00 for the 393
1st study) and fever (RR 1.96 for the 1
st study) (Saxen 1975, Saxen 1975), and the study by Aro on limb 394
reduction defects reported an OR of 1.6 for fever and 1.9 for influenza.(Aro 1983) It can be concluded 395
that included studies generally reported equivalent or stronger associations for influenza than for 396
fever with respect to CA following 1st
trimester exposure. 397
Another possible pathway by which the influenza virus can mediate a teratogenic effect is placental 398
transmission. Placental transmission has been documented but appears to be rare.(Gu et al. 2007, 399
McGregor et al. 1984) Toxic metabolites associated with fever as a cause of congenital anomalies 400
have also been suggested.(Edwards 2006) 401
Page 24 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
18
Limitations of the study 402
The systematic review results should be interpreted in the light of the findings that all of the included 403
studies are observational studies and many were susceptible to several types of bias and 404
ascertainment of exposure might not have been reliable. Although the funnel plots did not provide 405
evidence for publication bias, it should be noted that most of the included studies reported a wide 406
range of positive associations raising the question of whether studies reporting negative results 407
remained unpublished. Reference tracking identified 6 new studies and these studies were missed 408
because they were not indexed as influenza studies. This leaves the possibility open that some, 409
particularly negative, studies were missed due to poorly indexed terms. A possible reason for this is 410
that some case-control studies investigate a wide range of possible causes of CA and may tend to be 411
selectively indexed for the positive associations. For older studies, we could not be sure whether 412
chromosomal CA were excluded from analysis. 413
A weakness of the meta-analysis was that adjustment for confounders was not performed in most 414
included studies. Adjustment for confounders showed a moderate effect on OR within studies that 415
did report both crude OR and adjusted OR.(Acs et al. 2005, Granroth et al. 1978, Li et al. 2007) 416
Subgroup analysis comparing adjusted versus crude estimates did not detect differences, but this 417
could be related to the fact that 50% of the adjusted estimate was composed of neural tube defects 418
data since studies generally reported higher estimates for neural tube defects than for other CA. Due 419
to the limited amount of studies reporting adjusted OR, comparison of crude and adjusted OR for 420
subgroups of the same CA was not possible. Some very large datasets involved matched and/or 421
stratified controls (Botto et al. 2001, Czeizel et al. 2008), and most other datasets were matched by 422
one or more variables (Aro 1983, Granroth 1978, Granroth et al. 1978, Karkinen-Jaaskelainen and 423
Saxen 1974, Laurence et al. 1968, Li et al. 2007, Lynberg et al. 1994, Saxen 1975, Saxen 1975, Warrell 424
et al. 1981) lowering the impact of confounding on the meta-analysis. 425
Statistical limitations 426
Page 25 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
19
The study was susceptible to statistical limitations unique to meta-analysis of scarce events. In the 427
context of scarce events like CA and complicated exposure like 1st
trimester influenza, cohort studies 428
can enrol large numbers of exposed and unexposed, but have very few exposed CA outcomes. 429
Current statistical methods will assign these studies a lot of weight compared to case-control studies 430
with many cases and greater numbers of exposed cases. An example from this systematic review is 431
the cohort study by Pleydell, which reported 1 case of hydrocephaly among 12 1st
trimester influenza 432
exposed pregnancies and 1 case among 1071 unexposed pregnancies leading to an OR of 97.27. The 433
heterogeneity model favours outliers and smaller studies and provided this study with 20% weight in 434
the overall hydrocephaly estimate, compared to 43% weight for the case-control study by Czeizel 435
which enrolled 314 cases of hydrocephaly (16 exposed). The weight allocated by the heterogeneity 436
model to the Pleydell study is clearly disproportionate. 437
A second problem lies in the continuity correction which is used to address zero events in one of 438
both arms of a study when pooling odds ratios. For rare events like CA, zero events in one or both 439
arms are not uncommon. The standard value for continuity correction is 0.5, but this arbitrary value 440
causes problems in studies with uneven arms and can even dominate the other arm. We addressed 441
this problem as proposed by Sweeting (Sweeting et al. 2004) by letting the continuity correction 442
depend on the OR of (other) studies with >0 events in both arms and group ratio imbalance. 443
However, this still led to OR > 1 for studies reporting 0 events in the exposed arm such as the 444
hydrocephaly data by Hirvensalo. 445
Conclusions and implications for CA prevention 446
Given the risk of congenital anomaly associated with influenza we show here, prevention of influenza 447
by vaccinating women who are planning to get pregnant may reduce congenital anomaly risk. 448
However, before evidence based policy can be implemented, further safety data on use of influenza 449
vaccines in pregnancy with respect to congenital anomalies is required (Kallen and Olausson 2012, 450
Page 26 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
20
Pasternak et al. 2012). Other methods for preventing influenza in early pregnancy include improving 451
nutritional and general health status and adopting behaviours which prevent interpersonal spread. 452
In conclusion, prevention of influenza in pregnant women may reduce congenital anomaly risk, and 453
would be relevant to more than just neural tube defects. More research is needed to determine 454
whether influenza and/or its related medication is teratogenic, to determine the role of 455
hyperthermia in teratogenicity and the role of other environmental factors such as nutritional status 456
in determining susceptibility. 457
Declaration of authors roles 458
J.M.L. (corresponding author, first author) was involved in study design, data collection and synthesis, 459
manuscript preparation and revision, construction of tables and figures, statistical analysis and 460
submission of the manuscript. M.J.B. (co-author) was involved in data collection and manuscript 461
revision. H.D. (co-author) was involved in study design, writing and revising the manuscript. 462
Acknowledgements 463
We would like to thank Dr. I. Barisic for assistance with case classification and we would like to thank 464
Dr. Julia Métneki, Dr. Erzsébet Puhó, Professor Andrew Czeizel and Dr. Annukka Ritvanen for sending 465
additional data. We would like to thank Professor Lolkje de Jong-van den Berg and Dr. Gordon 466
Marnoch for their helpful comments. 467
Funding 468
Funded by the EC, under the framework of the EU Health Programme 2008-2013, Grant Agreement 469
2010 22 04 (Executive Agency for Health & Consumers). 470
Conflict of interest 471
No competing interests 472
Page 27 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
21
Figure 1: Flow chart of systematic review 473
[Figure 1] 474
Figure 2: Forest plot of non-chromosomal CA following 1st
trimester influenza exposure 475
[Figure 2] 476
Figure 3: Forest plot of neural tube defects following 1st trimester influenza exposure 477
[Figure 3] 478
Figure 4: Forest plot of anencephaly following 1st
trimester influenza exposure 479
[Figure 4] 480
Figure 5: Forest plot of encephalocele following 1st
trimester influenza exposure 481
[Figure 5] 482
Figure 6: Forest plot of spina bifida following 1st
trimester influenza exposure 483
[Figure 6] 484
Figure 7: Forest plot of hydrocephaly following 1st
trimester influenza exposure 485
[Figure 7] 486
Figure 8: Forest plot of congenital heart defects following 1st
trimester influenza exposure 487
[Figure 8] 488
Figure 9: Forest plot of orofacial clefts following 1st
trimester influenza exposure 489
[Figure 9] 490
Page 28 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
22
References 491
Abramowitz LJ. The effect of Asian influenza on pregnancy. S Afr Med J 1958:32:1155-1156. 492
Acs N, Banhidy F, Puho E and Czeizel AE. Maternal influenza during pregnancy and risk of 493
congenital abnormalities in offspring. Birth Defects Res A Clin Mol Teratol 2005:73:989-996. 494
Adams MM, Mulinare J and Dooley K. Risk factors for conotruncal cardiac defects in Atlanta. J Am 495
Coll Cardiol 1989:14:432-442. 496
Aro T. Maternal diseases, alcohol consumption and smoking during pregnancy associated with 497
reduction limb defects. Early Hum Dev 1983:9:49-57. 498
Botto LD, Lynberg MC and Erickson JD. Congenital heart defects, maternal febrile illness, and 499
multivitamin use: A population-based study. Epidemiology 2001:12:485-490. 500
Boyd PA, Devigan C, Khoshnood B, Loane M, Garne E, Dolk H and EUROCAT Working Group. 501
Survey of prenatal screening policies in Europe for structural malformations and chromosome 502
anomalies, and their impact on detection and termination rates for neural tube defects and Down's 503
syndrome. BJOG 2008:115:689-696. 504
Buck C. Exposure to virus diseases in early pregnancy and congenital malformations. Can Med Assoc 505
J 1955:72:744-746. 506
Busby A, Dolk H and Armstrong B. Eye anomalies: seasonal variation and maternal viral infections. 507
Epidemiology 2005:16:317-322. 508
Campbell WA. Influenza in early pregnancy; effects on the foetus. Lancet 1953:1:173-174. 509
Choi NW and Klaponski FA. On neural-tube defects: an epidemiological elicitation of etiological 510
factors. Neurology 1970:20:399-400. 511
Cockroft DL and New DA. Abnormalities induced in cultured rat embryos by hyperthermia. 512
Teratology 1978:17:277-283. 513
Cockroft DL and New DA. Effects of hyperthermia on rat embryos in culture. Nature 1975:258:604-514
606. 515
Coffey VP and Jessop WJ. Maternal influenza and congenital deformities. A follow-up study. Lancet 516
1963:1:748-751. 517
Coffey VP and Jessop WJ. Maternal influenza and congenital deformities: a prospective study. Lancet 518
1959:2:935-938. 519
Coffey VP and Jessop WJ. Congenital abnormalities. Ir J Med Sci 1955:6:30-48. 520
Csaky-Szunyogh M, Vereczkey A, Kosa Z, Urban R and Czeizel AE. Association of maternal diseases 521
during pregnancy with the risk of single ventricular septal defects in the offspring--a population-based 522
case-control study. J Matern Fetal Neonatal Med 2013:26:738-747. 523
Page 29 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
23
Czeizel AE, Puho EH, Acs N and Banhidy F. Use of specified critical periods of different congenital 524
abnormalities instead of the first trimester concept. Birth Defects Res A Clin Mol Teratol 2008:82:139-525
146. 526
Davies HT, Crombie IK and Tavakoli M. When can odds ratios mislead?. BMJ 1998:316:989-991. 527
DerSimonian R and Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986:7:177-188. 528
Dodds L, McNeil SA, Fell DB, Allen VM, Coombs A, Scott J and MacDonald N. Impact of influenza 529
exposure on rates of hospital admissions and physician visits because of respiratory illness among 530
pregnant women. CMAJ 2007:176:463-468. 531
Doll R, Hill AB and Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 532
1960:14:167-172. 533
Edwards MJ. Review: Hyperthermia and fever during pregnancy. Birth Defects Res A Clin Mol 534
Teratol 2006:76:507-516. 535
Edwards MJ. The experimental production of Arthrogryposis multiplex congenita in guinea-pigs by 536
maternal hyperthermia during gestation. J Pathol 1971:104:221-229. 537
Edwards MJ. The experimental production of clubfoot in guinea-pigs by maternal hyperthermia during 538
gestation. J Pathol 1971:103:49-53. 539
Edwards MJ. Congenital defects in guinea pigs: fetal resorptions, abortions, and malformations 540
following induced hyperthermia during early gestation. Teratology 1969:2:313-328. 541
Edwards MJ. Congenital defects in guinea pigs: prenatal retardation of brain growth of guinea pigs 542
following hyperthermia during gestation. Teratology 1969:2:329-336. 543
Elizan TS, Ajero-Froehlich L, Fabiyi A, Ley A and Sever JL. Viral infection in pregnancy and 544
congenital CNS malformations in man. Arch Neurol 1969:20:115-119. 545
EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013. 546
EUROCAT Central Registry. EUROCAT Guide 1.3 and reference documents. Instructions for the 547
Registration and Surveillance of Congenital Anomalies. Revised: August 2013. 548
Evans D. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare 549
interventions. J Clin Nurs 2003:12:77-84. 550
Germain MA, Webster WS and Edwards MJ. Hyperthermia as a teratogen: parameters determining 551
hyperthermia-induced head defects in the rat. Teratology 1985:31:265-272. 552
Granroth G. Defects of the central nervous system in Finland: III. Disease and drugs in pregnancy. 553
Early Hum Dev 1978:2:147-162. 554
Granroth G, Haapakoski J and Saxen L. Defects of the central nervous system in Finland: V. 555
Multivariate analysis of risk indicators. Int J Epidemiol 1978:7:301-308. 556
Gu J, Xie Z, Gao Z, Liu J, Korteweg C, Ye J, Lau LT, Lu J, Gao Z, Zhang B et al. H5N1 infection of 557
the respiratory tract and beyond: a molecular pathology study. Lancet 2007:370:1137-1145. 558
Page 30 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
24
Hakosalo J and Saxen L. Influenza epidemic and congenital defects. Lancet 1971:2:1346-1347. 559
Hardy JM, Azarowicz EN, Mannini A, Medearis DN,Jr and Cooke RE. The effect of Asian influenza 560
on the outcome of pregnancy, Baltimore, 1957-1958. Am J Public Health Nations Health 561
1961:51:1182-1188. 562
Harris J. Influenza occurring in pregnant women: a statistical study of thirteen hundred and fifty cases. 563
JAMA 1919;72:978–80 1919:. 564
Hirvensalo M and Kinnunen O. Influenza and pregnancy. Duodecim 1962:78:740-747. 565
Hofler M. The Bradford Hill considerations on causality: a counterfactual perspective. Emerg Themes 566
Epidemiol 2005:2:11. 567
Jamieson DJ, Theiler RN and Rasmussen SA. Emerging infections and pregnancy. Emerg Infect Dis 568
2006:12:1638-1643. 569
Kallen B and Olausson PO. Vaccination against H1N1 influenza with Pandemrix((R)) during 570
pregnancy and delivery outcome: a Swedish register study. BJOG 2012:119:1583-1590. 571
Karkinen-Jaaskelainen and Saxen L. Maternal influenza, drug consumption, and congenital defects of 572
the central nervous system. Am J Obstet Gynecol 1974:118:815-818. 573
Kelly S, Keuhl K and Leffredo C. Tricupsid Atresia: an epidemiologic investigation in the Baltimore-574
Washington area. American Journal of Epidemiology 2012:175:S59. 575
Klemetti A. Definition of congenital malformations and detection of associations with maternal 576
factors. Early Hum Dev 1977:1:117-123. 577
Laurence KM, Carter CO and David PA. Major central nervous system malformations in South Wales. 578
II. Pregnancy factors, seasonal variation, and social class effects. Br J Prev Soc Med 1968:22:212-222. 579
Leck I. Further tests of the hypothesis that influenza in pregnancy causes malformations. HSMHA 580
Health Rep 1971:86:265-269. 581
Leck I. Examination of the Incidence of Malformations for Evidence of Drug Teratogenesis. Br J Prev 582
Soc Med 1964:18:196-201. 583
Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963:17:70-80. 584
Leck I, Hay S, Witte JJ and Greene JC. Malformations recorded on birth certificates following A2 585
influenza epidemics. Public Health Rep 1969:84:971-979. 586
Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural 587
tube defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 588
Teratol 2007:79:295-300. 589
Luteijn JM, Dolk H and Marnoch GJ. Differences in pandemic influenza vaccination policies for 590
pregnant women in Europe. BMC Public Health 2011:11:819. 591
Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: 592
a population-based case-control study. Am J Epidemiol 1994:140:244-255. 593
Page 31 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
25
MacKenzie JS and Houghton M. Influenza infections during pregnancy: association with congenital 594
malformations and with subsequent neoplasms in children, and potential hazards of live virus 595
vaccines. Bacteriol Rev 1974:38:356-370. 596
Mak TK, Mangtani P, Leese J, Watson JM and Pfeifer D. Influenza vaccination in pregnancy: current 597
evidence and selected national policies. Lancet Infect Dis 2008:8:44-52. 598
McGregor JA, Burns JC, Levin MJ, Burlington B and Meiklejohn G. Transplacental passage of 599
influenza A/Bangkok (H3N2) mimicking amniotic fluid infection syndrome. Am J Obstet Gynecol 600
1984:149:856-859. 601
Medveczky E, Puho E and Czeizel AE. An evaluation of maternal illnesses in the origin of neural-tube 602
defects. Arch Gynecol Obstet 2004:270:244-251. 603
Mereckiene J, Cotter S, D'Ancona F, Giambi C, Nicoll A, Levy-Bruhl D, Lopalco PL, Weber JT, 604
Johansen K, Dematte L et al. Differences in national influenza vaccination policies across the 605
European Union, Norway and Iceland 2008-2009. Euro Surveill 2010:15:19700. 606
Metneki J, Puho E and Czeizel AE. Maternal diseases and isolated orofacial clefts in Hungary. Birth 607
Defects Res A Clin Mol Teratol 2005:73:617-623. 608
Moher D, Liberati A, Tetzlaff J, Altman DG and PRISMA Group. Preferred reporting items for 609
systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009:339:b2535. 610
Monto AS, Gravenstein S, Elliott M, Colopy M and Schweinle J. Clinical signs and symptoms 611
predicting influenza infection. Arch Intern Med 2000:160:3243-3247. 612
Moretti ME, Bar-Oz B, Fried S and Koren G. Maternal hyperthermia and the risk for neural tube 613
defects in offspring: systematic review and meta-analysis. Epidemiology 2005:16:216-219. 614
Mossey PA, Little J, Munger RG, Dixon MJ and Shaw WC. Cleft lip and palate. Lancet 615
2009:374:1773-1785. 616
Neuzil KM, Reed GW, Mitchel EF, Simonsen L and Griffin MR. Impact of influenza on acute 617
cardiopulmonary hospitalizations in pregnant women. Am J Epidemiol 1998:148:1094-1102. 618
Oster ME, Riehle-Colarusso T, Alverson CJ and Correa A. Associations between maternal fever and 619
influenza and congenital heart defects. J Pediatr 2011:158:990-995. 620
Park CH, Stewart W, Khoury MJ and Mulinare J. Is there etiologic heterogeneity between upper and 621
lower neural tube defects?. Am J Epidemiol 1992:136:1493-1501. 622
Pasternak B, Svanstrom H, Molgaard-Nielsen D, Krause TG, Emborg HD, Melbye M and Hviid A. 623
Risk of adverse fetal outcomes following administration of a pandemic influenza A(H1N1) vaccine 624
during pregnancy. JAMA 2012:308:165-174. 625
Pleydell MJ. Anencephaly and other congenital abnormalities. An epidemiological study in 626
Northamptonshire. Br Med J 1960:1:309-315. 627
Record RG. Anencephalus in Scotland. Br J Prev Soc Med 1961:15:93-105. 628
Page 32 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
26
Rockenbauer M, Olsen J, Czeizel AE, Pedersen L, Sorensen HT and EuroMAP Group. Recall bias in a 629
case-control surveillance system on the use of medicine during pregnancy. Epidemiology 630
2001:12:461-466. 631
Saxen I. The association between maternal influenza, drug consumption and oral clefts. Acta Odontol 632
Scand 1975:33:259-267. 633
Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev 634
Soc Med 1975:29:103-110. 635
Saxen L, Hjelt L, Sjostedt JE, Hakosalo J and Hakosalo H. Asian influenza during pregnancy and 636
congenital malformations. Acta Pathol Microbiol Scand 1960:49:114-126. 637
Saxen L, Holmberg PC, Kurppa K, Kuosma E and Pyhala R. Influenza epidemics and anencephaly. 638
Am J Public Health 1990:80:473-475. 639
Saxen L, Klemetti A and Haro AS. A matched-pair register for studies of selected congenital defects. 640
Am J Epidemiol 1974:100:297-306. 641
Siston AM, Rasmussen SA, Honein MA, Fry AM, Seib K, Callaghan WM, Louie J, Doyle TJ, 642
Crockett M, Lynfield R et al. Pandemic 2009 influenza A(H1N1) virus illness among pregnant women 643
in the United States. JAMA 2010:303:1517-1525. 644
Smith MS, Upfold JB, Edwards MJ, Shiota K and Cawdell-Smith J. The induction of neural tube 645
defects by maternal hyperthermia: a comparison of the guinea-pig and human. Neuropathol Appl 646
Neurobiol 1992:18:71-80. 647
Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA 648
and Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. 649
Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000:283:2008-650
2012. 651
Sweeting MJ, Sutton AJ and Lambert PC. What to add to nothing? Use and avoidance of continuity 652
corrections in meta-analysis of sparse data. Stat Med 2004:23:1351-1375. 653
Walker WM and McKee AP. Asian influenza in pregnancy; relationship to fetal anomalies. Obstet 654
Gynecol 1959:13:394-398. 655
Warrell MJ, Tobin JO and Wald NJ. Examination for influenza IgA and IgM antibodies in pregnancies 656
associated with fetal neural-tube defects. J Med Microbiol 1981:14:159-162. 657
Wells GA, Shea B and O'Connell D. The Newcastle-Ottawa Scale (NOS) for assessing the quality of 658
nonrandomised studies in meta-analyses 2004:. 659
Wilson MG, Heins HL, Imagawa DT and Adams JM. Teratogenic effects of Asian influenza. J Am 660
Med Assoc 1959:171:638-641. 661
Wilson MG and Stein AM. Teratogenic effects of asian influenza. A n extended study. JAMA 662
1969:210:336-337. 663
Zambon M, Hays J, Webster A, Newman R and Keene O. Diagnosis of influenza in the community: 664
relationship of clinical diagnosis to confirmed virological, serologic, or molecular detection of 665
influenza. Arch Intern Med 2001:161:2116-2122. 666
Page 33 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Table I: First trimester maternal influenza exposure and risk of congenital anomalies: studies, total 1
number of cases of congenital anomaly, pooled odds ratio and heterogeneity. 2
Group Participating
studies (n)
I2
Statistic for
heterogeneity (%)
Pooled OR (95%
CI)
Total
number of
CA (n)
Any congenital anomaly 22 64 2.00 (1.62-2.48) 29,945c
-Susceptible to differential
recall bias
9 65 1.92 (1.35-2.72) 5,426
-Not susceptible to
differential recall bias
13 64 2.12 (1.54-2.91) 24,519c
-Case-control studies 13 60 1.84 (1.49-2.27) 29,542c
-Cohort studies 9 62 2.12 (1.21-3.72) 403
-Any type, published 1955-
1969
11 58 2.47 (1.50-4.70) 1,171
-Any type, Published 1975-
2011
11 55 1.71 (1.41-2.08) 28,774
-Adjusted estimates only a 4 87 2.15 (1.05-4.42) 3,865
-Crude estimates only a 21 61 2.22 (1.78-2.77) 27,584
Neural Tube Defects 11 50 3.33 (2.05-5.40) 2,500
-Anencephaly 10 44 3.52 (1.69-7.32) 608
-Encephalocele 4 63 2.95 (0.78-11.13) 225
-Spina Bifida 7 0 2.20 (1.48-3.28) 1,093
Hydrocephaly 5 45 5.74 (1.10-30.00) 323
Congenital heart defects 10 41 1.56 (1.13-2.14) 7,715
-Aortic Valve 3 31 2.59 (1.21-5.54) 167
Page 34 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
a Three studies (Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994) were able to provide estimates 3
both for adjusted and crude OR.
4
b Note that a single study contributed over 90% of the total weight to this pooled estimate
5
c Note that for some studies such as Czeizel et al 2008, the “any congenital anomaly” group also 6
included CA which were not included in any other analysis.
7
Atresia/Stenosis
-Atrial Septal Defect 3 0 0.82 (0.45-1.51) 429
-Hypoplastic Left Heart 3 0 1.58 (0.94-2.64) 203
-Transposition of the Great
Vessels
3 0 1.40 (0.90-2.17) 321
-Ventricular Septal Defect 4 0 1.59 (1.24-2.04) 1,434
Orofacial Clefts 10 37 1.96 (1.33-2.91) 2,773d
-Cleft Lip +- Palate b 7 0 3.12 (2.20-4.42) 1,404
-Cleft Palate b 3 0 1.05 (0.60-1.84) 584
Digestive System b 4 0 1.71 (1.09-2.69) 1,195
Urinary 5 0 1.45 (0.90-2.34) 48
Hypospadias b 4 0 1.02 (0.75-1.39) 3,041
Limb Reduction 3 0 2.03 (1.27-3.27) 1,002
Club Foot b 4 0 1.11 (0.93-1.34) 2,430
Hip Dislocation/Dysplasia 3 0 0.31 (0.00-37.62) 37
Polydactyly b 4 0 1.72 (0.85-3.48) 1,094
Syndactyly 3 71 1.98 (0.19-
20.563)
662
Musculo-Skeletal 3 0 1.05 (0.16-6.97) 776
Page 35 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
d Note that 591 orofacial clefts from the Saxen 1975a study (Saxen 1975) and 194 orofacial clefts 8
from the Saxen 1975b study (Saxen 1975) were not specified and therefore solely included in the 9
overall orofacial clefts analysis. 10
Page 36 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Table II: Evidence included in the systematic review relating to 1st
trimester influenza exposure and 1
CA not included in meta-analysis. 2
Anomaly Evidence reported with respect to 1st
trimester influenza
exposure
Any anomaly Visual inspection: no convincing evidence that infant death rates
from CA are associated with 1st
trimester influenza
exposure(Buck 1955); RR 1.10(Leck 1963); single defect RR 1.00,
multiple defects RR 0.9(Leck 1964); single defect RR 1.03,
multiple defects RR 1.07(Leck et al. 1969)
Central Nervous System defects 0.81% in influenza exposed group vs. 0.30% and 0.32% in control
groups(Hakosalo and Saxen 1971)
Neural Tube Defects p=0.03, no data(Choi and Klaponski 1970)
Anencephaly RR 0.82(Leck 1963); sRR 1.32(Leck et al. 1969); RR 1.0, 0.8-
1.3(Saxen et al. 1990) No rise in anencephaly rates following
influenza epidemics detected(Record 1961)
Spina Bifida + Encephalocele RR 0.99(Leck 1963), sRR 1.13(Leck et al. 1969)
Hydrocephaly sRR 0.93(Leck et al. 1969)
Microcephaly OR 1.1, 0.3-4.1(Czeizel et al. 2008); sRR 0.24(Leck et al. 1969)
Anophthalmos/Microphthalmos OR 1.26, 1.02-1.57(Busby et al. 2005)
Congenital Cataract OR 0.8, 0.3-2.2(Czeizel et al. 2008)
Congenital Glucaoma OR 2.0, 0.4-10.9(Czeizel et al. 2008)
Congenital heart defects (suggestive but not significant; 0.93% in influenza exposed
group vs. 0.47% and 0.73% in control groups)(Hakosalo and
Saxen 1971)
Page 37 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
Atrioventricular Septal Defect OR 2.0, 0.3-15.3(Botto et al. 2001); AOR 1.29, 0.82-2.01(Oster et
al. 2011)
Tetralogy of Fallot OR 0.5, 0.1-3.6(Botto et al. 2001); AOR 0.78, 0.37-1.62(Oster et
al. 2011)
Triscuspid atresia and stenosis AOR 7.9, 0.3-29.6(Botto et al. 2001); AOR 6.04, 2.36-15.42(Oster
et al. 2011)
Ebstein's anomaly OR 3.0, 0.4-23.9(Botto et al. 2001)
Pulmonary valve stenosis AOR 1.21, 0.71-2.04(Oster et al. 2011)
Pulmonary valve atresia AOR 2.71, 1.16-6.32(Oster et al. 2011)
Coarctation of aorta AOR 3.8, 1.6-8.8(Botto et al. 2001); AOR 0.41, 0.13-1.33(Oster et
al. 2011)
Total anomalous pulm venous OR 2.2, 0.3-16.9(Botto et al. 2001)
Cleft Lip RR 1.55, p<0.05(Leck 1963); isolated RR 1.4, with other defects
RR 6.3(Leck 1964); cleft lip without cleft palate sRR 1.47,
p<0.05(Leck et al. 1969); cleft lip without cleft palate RR 1.64,
p<0.05(Leck 1971); cleft lip without cleft palate sRR 1.08, cleft
lip with cleft palate sRR 1.01(Leck et al. 1969); cleft lip with cleft
palate sRR 1.10(Leck et al. 1969), cleft lip with cleft palate RR
1.35(Leck 1971)
Cleft Palate RR 0.81(Leck 1963); sRR 0.94(Leck et al. 1969); sRR 0.93(Leck et
al. 1969)
Page 38 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
Digestive System 1 in 63 exposed during 1st
trimester of pregnancy and 4 in 1106
non-exposed, RR 4.4(Coffey and Jessop 1959, Coffey and Jessop
1963); 1 in 171 exposed during 1st
trimester of pregnancy and 13
in 6720 non-exposed, RR 3.0(Hirvensalo and Kinnunen 1962)
Oesophageal atresia
with/without
tracheo-oesophageal fistula
OR 1.6, 0.2-11.7(Czeizel et al. 2008); RR 2.41, p<0.01(Leck 1963);
isolated RR 1.5, with other defects RR 4.8(Leck 1964); sRR
1.12(Leck et al. 1969)
Atresia/Stenosis of the small
intestine
OR 2.1, 0.5-8.0(Czeizel et al. 2008)
Ano-rectal atresia and stenosis OR 1.1, 0.4-3.0(Czeizel et al. 2008); RR 2.12, p<0.05(Leck 1963);
isolated RR 0.8, with other defects RR 3.6(Leck 1964); sRR
0.80(Leck et al. 1969)
Hirschsprung's disease OR 0.7, 0.2-3.0(Czeizel et al. 2008)
Diaphragmatic Hernia OR 3.2, 1.2-8.8(Czeizel et al. 2008); 0 in 171 exposed during 1st
trimester of pregnancy and 3 in 6720 non-exposed(Hirvensalo
and Kinnunen 1962); RR 1.05(Leck 1963); sRR 1.01(Leck et al.
1969)
Abdominal Wall Defects OR 2.8, 1.1-6.9(Czeizel et al. 2008)
Omphalocele RR 1.92, <0.05(Leck 1963); isolated RR 1.3, with other defects RR
2.3(Leck 1964); sRR 1.16(Leck et al. 1969)
Urinary 1 in 171 exposed during 1st trimester of pregnancy and 42 in
6720 non-exposed(Hirvensalo and Kinnunen 1962)
Bilateral renal agenesis including
Potter syndrome
OR 0.8, 0.2-2.8(Czeizel et al. 2008); RR 1.38(Leck 1963)
Page 39 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Congenital Hydronephrosis RR 1.63, p=0.05(Leck 1963)
Hypospadias RR 1.51(Leck 1963); sRR 0.93(Leck et al. 1969)
Limb sRR 0.75(Leck et al. 1969)
Limb Reduction Limited to thumbs or radii RR 2.20(Leck 1963); both arms or legs
RR 1.30(Leck 1963); one arm or leg RR 1.37(Leck 1963); sRR
1.32(Leck et al. 1969); RR 1.22(Leck 1971)
Upper Limb Reductions sRR 1.91, p<0.01(Leck et al. 1969)
Lower Limb Reductions sRR 1.23(Leck et al. 1969)
Club Foot sRR 0.89(Leck et al. 1969)
Hip Dislocation and/or Dysplasia sRR 0.77(Leck et al. 1969)
Disorders of skin 0 in 171 exposed during 1st trimester of pregnancy and 3 in 6720
non-exposed(Hirvensalo and Kinnunen 1962)
Down’s Syndrome sRR 1.16(Leck et al. 1969)
Solely data involving >2 CA cases was included and some reported CA could not be translated to 3
EUROCAT defined subgroups. We did not distinguish between “positive” and “negative” findings 4
since some of the studies were severely underpowered. Note for some CA, ≥3 estimates were 5
available from studies that could not be included in the meta-analysis. 6
Page 40 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4,018 potentially relevant articles identified by Embase (2,649) and PubMed® (1,369)
database search
6 potentially relevant articles identified by
reference tracking
2,897 unique articles -1,121 duplicates within and across databases
2,897 unique articles screened
2,615 articles excluded -No abstract. Title, MeSH, publication
type ruled out relevance 843 - Title, MeSH and abstract ruled out
relevance 1,772
282 + 6 articles selected for full review
242 articles excluded -Relevance unknown due to foreign
language 39 - Non relevant original research 134
-No original data in article 50 -Exposure timing outside 1st trimester/non specified 13
-Exposure not sufficiently specified 6
46 articles included in systematic review
Screening
Eligibility
Included
32 articles (22 datasets) included in meta-analysis
13 articles excluded -Ecological studies 9
- Cohort studies without controls 1 -Studies with insufficient information 4
Page 41 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
239x202mm (72 x 72 DPI)
Page 42 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
311x225mm (72 x 72 DPI)
Page 43 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
283x202mm (72 x 72 DPI)
Page 44 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
306x202mm (72 x 72 DPI)
Page 45 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
298x202mm (72 x 72 DPI)
Page 46 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
306x202mm (72 x 72 DPI)
Page 47 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
285x202mm (72 x 72 DPI)
Page 48 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
280x202mm (72 x 72 DPI)
Page 49 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
eAppendix 1
Quality assessment of case-control and cohort studies enrolled in meta-analysis 2
For quality assessment of studies included in the meta-analysis, we used a slightly modified version 3
of the Newcastle-Ottawa scale (NOS).(Wells et al. 2004) We did not distinguish between community-4
based or hospital-based controls since most studies would recruit from antenatal clinics it is safe to 5
assume that this population is similar to a community-recruited population. In addition, the risk of 6
referrals would not have been present in studies from before the late 1980s before prenatal 7
screening was widely available. There is some risk of bias introduced from high risk pregnancies 8
being more likely to book at specific hospitals. The following NOS items were not assessed: due to 9
the unique nature of CA we did not require case-control studies to demonstrate that controls did not 10
have a history of the outcome of interest and we did not require cohort studies to provide 11
demonstration that the outcome of interest was not present at start of study. Quality scores were 12
not used since weighting and calculating overall quality scores is arbitrary and the importance of 13
individual items and direction of bias depends on the context in which they are applied. 14
Case-control studies 15
Case definition was considered low risk of bias if the source reported diagnosis by healthcare 16
professional, hospital or primary records, or utilized independent validation. Representativeness of 17
cases were considered low risk of bias if cases were drawn from a clearly defined period of time and 18
catchment area, or consisted of an appropriate random sample of cases. We considered selection of 19
controls low risk of bias if controls could have been in the case group had the outcome of interest 20
been present as opposed to controls from different time periods or locations than cases. For 21
assessing risk of confounding (comparability in NOS) we considered any comparative study which did 22
match or adjust for confounding by at least maternal age and socioeconomic class low risk of bias. 23
For socioeconomic class, several proxies such as malnutrition for older studies and race for American 24
studies were also considered satisfactory. Fever and antiviral use were not considered confounders 25
Page 50 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
for reasons stated in the introduction. For ascertainment of exposure, we considered serological 26
confirmation, diagnosis by healthcare professionals and prospective interviews low risk of bias. Post 27
delivery maternal interviews were considered high risk of bias since mothers of infants born with 28
anomalies might make a more determined effort to recall illnesses during pregnancy than other 29
mothers leading to differential recall bias. In case-control designs, recall bias has been estimated to 30
bias OR up to a factor of 1.9.(Rockenbauer et al. 2001) We considered attrition or non-response 31
rates ≥20% for cases or controls high risk of bias. 32
Cohort studies 33
Representativeness of the exposed cohort was considered low risk of bias if the exposed cohort was 34
drawn from a clearly defined period of time and catchment area, or consisted of an appropriate 35
random sample of eligible cohort subjects. If ≤80% of the identified individuals in this place and time 36
failed to enrol, representativeness was also considered high risk of bias. We considered selection of 37
the unexposed cohort low risk of bias if individuals in the unexposed cohort could have been in the 38
exposed cohort had they been exposed to 1st
trimester influenza infection. Ascertainment of 39
exposure was handled the same way as for the case-control studies (see above). Length of follow-up 40
is only relevant for certain congenital anomalies such as congenital heart defects and we considered 41
a length of follow up of 1 year satisfactory. With respect to adequacy of follow up, we considered 42
≤20% attrition low risk of bias. 43
Figure 10: Risk of bias assessment of included case-control studies. 44
[Figure 10] 45
Green represents low risk of bias, red represents high risk of bias, yellow represents unknown risk of 46
bias. 47
Figure 11: Risk of bias assessment of included cohort studies. 48
[Figure 11] 49
Page 51 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
Green represents low risk of bias, red represents high risk of bias, yellow represents unknown risk of 50
bias. 51
Supplementary Forest Plots 52
Figure 12: Forest plot of aortic valve atresia/stenosis following 1st
trimester influenza exposure 53
[Figure 12] 54
Figure 13: Forest plot of atrial septal defect following 1st
trimester influenza exposure 55
[Figure 13] 56
Figure 14: Forest plot of hypoplastic left heart following 1st
trimester influenza exposure 57
[Figure 14] 58
Figure 15: Forest plot of transposition of the great vessels following 1st
trimester influenza exposure 59
[Figure 15] 60
Figure 16: Forest plot of ventricular septal defects following 1st
trimester influenza exposure 61
[Figure 16] 62
Figure 17: Forest plot of cleft lip ± cleft palate following 1st
trimester influenza exposure 63
[Figure 17] 64
Figure 18: Forest plot of cleft palate following 1st trimester influenza exposure 65
[Figure 18] 66
Figure 19: Forest plot of digestive system defects following 1st
trimester influenza exposure 67
[Figure 19] 68
Figure 20: Forest plot of urinary defects following 1st
trimester influenza exposure 69
[Figure 20] 70
Figure 21: Forest plot of hypospadias following 1st
trimester influenza exposure 71
[Figure 21] 72
Figure 22 Forest plot of limb reduction defects following 1st
trimester influenza exposure 73
[Figure 22] 74
Page 52 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Figure 23: Forest plot of club foot following 1st trimester influenza exposure 75
[Figure 23] 76
Figure 24: Forest plot of hip dislocation and/or dysplasia following 1st
trimester influenza exposure 77
[Figure 25] 78
Figure 25: Forest plot of polydactyly following 1st
trimester influenza exposure 79
[Figure 25] 80
Figure 26: Forest plot of syndactyly following 1st trimester influenza exposure 81
[Figure 26] 82
Figure 27: Forest plot of musculo-skeletal defects following 1st
trimester influenza exposure 83
[Figure 27] 84
Supplementary Funnel Plots 85
Funnel plot asymmetry is indicative for publication bias as for example small studies reporting 86
negative findings might not be published. Visual inspection of funnel plots did not provide evidence 87
for publication bias. 88
Figure 28: Funnel plot of non-chromosomal CA following 1st
trimester influenza exposure 89
[Figure 28] 90
Figure 29: Funnel plot of neural tube defectsfollowing 1st trimester influenza exposure 91
[Figure 29] 92
Figure 30: Funnel plot of anencephaly following 1st
trimester influenza exposure 93
[Figure 30] 94
Figure 31: Funnel plot of encephalocele following 1st
trimester influenza exposure 95
[Figure 31] 96
Figure 32: Funnel plot of spina bifida following 1st
trimester influenza exposure 97
[Figure 32] 98
Figure 33: Funnel plot of hydrocephaly following 1st
trimester influenza exposure 99
Page 53 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
[Figure 33] 100
Figure 34: Funnel plot of congenital heart defects following 1st trimester influenza exposure 101
[Figure 34] 102
Figure 35: Funnel plot of orofacial clefts following 1st
trimester influenza exposure 103
[Figure 35] 104
Figure 36: Funnel plot of aortic valve atresia/stenosis following 1st
trimester influenza exposure 105
[Figure 36] 106
Figure 37: Funnel plot of atrial septal defect following 1st
trimester influenza exposure 107
[Figure 37] 108
Figure 38: Funnel plot of hypoplastic left heart following 1st
trimester influenza exposure 109
[Figure 38] 110
Figure 39: Funnel plot of transposition of the great vessels following 1st
trimester influenza exposure 111
[Figure 39] 112
Figure 40: Funnel plot of ventricular septal defects following 1st
trimester influenza exposure 113
[Figure 40] 114
Figure 41: Funnel plot of cleft lip ± cleft palate following 1st
trimester influenza exposure 115
[Figure 41] 116
Figure 42: Funnel plot of cleft palate following 1st
trimester influenza exposure 117
[Figure 42] 118
Figure 43: Funnel plot of digestive system defects following 1st
trimester influenza exposure 119
[Figure 43] 120
Figure 44: Funnel plot of urinary defects following 1st
trimester influenza exposure 121
[Figure 44] 122
Figure 45: Funnel plot of hypospadias following 1st
trimester influenza exposure 123
[Figure 45] 124
Figure 46: Funnel plot of limb reduction defects following 1st trimester influenza exposure 125
Page 54 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
6
[Figure 46] 126
Figure 47: Funnel plot of club foot following 1st trimester influenza exposure 127
[Figure 47] 128
Figure 48: Funnel plot of hip dislocation and/or dysplasia defects following 1st
trimester influenza 129
exposure 130
[Figure 48] 131
Figure 49: Funnel plot of polydactyly following 1st trimester influenza exposure 132
[Figure 49] 133
Figure 50: Funnel plot of syndactyly following 1st trimester influenza exposure 134
[Figure 50] 135
Figure 51: Funnel plot of musculo-skeletal defects following 1st
trimester influenza exposure 136
[Figure 51] 137
Page 55 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Table III: Overview of 9 papers from 8 ecological studies included in the systematic review. 1
Authors Publication
year (study
years) and
location
Study outcome Influenza
exposure and
critical period
No. of cases and
methodology
Buck
(Buck
1955)
1955
(Jan 1944-Jul
1951), Canada
Infant death due to CA Influenza
incidence per
calendar month
coinciding with
1st trimester of
pregnancy
Unclear population
size, infant deaths
only. Analysis by visual
inspection.
Busby et
al (Busby
et al.
2005)
2005 (1988-
1994), England
Anophthalmos and
microphthalmos
Reported
weekly
influenza
infection counts
coinciding with
6th-10th
week
of gestation.
275 malformed infants,
live births and
stillbirths. Analysis by
Poisson regression
comparing cases to all
births.
Hakosalo
and Saxen
(Hakosalo
and Saxen
1971)
1971 (Jan 15th
-
Oct 31st, 1958),
Finland
Anencephaly and CA of
the CNS, circulatory
system and urogenital
system.
Sickness
absenteeism
rates coinciding
with 5th
-11th
week of
gestation.
90 malformed infants,
live births and
stillbirths. Analysis by
χ2 test, comparing high
risk versus low risk
pregnancies. High risk
pregnancies were
Page 56 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
exposed to Pandemic
Asian Flu in 5th
-11th
week of gestation. Low
risk pregnancies were
conceived earlier
(Group 1) or conceived
later (Group 2)
Leck (Leck
1963, Leck
1964)
1963 and 1964
(1957-
1961(Leck
1963), 1957-
1963(Leck
1964)),
Birmingham,
UK
Anencephaly, Spina
Bifida, Cleft Lip, Cleft
Palate, Oesophageal
Atresia, Anal Atresia,
Renal Agenesis or
Hypoplasia,
Hydronephrosis,
Hypospadias, Limb
Reductions (Thumb or
Radii, Bilateral arms or
legs, single arm or leg),
Exomphalos and
Diaphragmatic
Hernia.(Leck 1963)
Weekly new
claims to
sickness benefit
coinciding with
26 to 40 weeks
before delivery.
939 malformed infants,
live births and
stillbirths.(Leck 1963)
15 additional 1957-
1961 cases and
unknown amount of
1962-1963 cases.(Leck
1964)
Analysis by incidence
ratio’s of high risk
versus low risk
pregnancies.
Leck et al
(Leck et al.
1969)
1969 (1956-
1965), Several
USA states
Orofacial Clefts (1956-
1965) and any CA (1961-
196%). Anencephalus,
Spina Bifida,
Weekly number
of deaths
attributable to
influenza
For 1956-1961: 1,496
isolated cleft palate,
1,418 isolated cleft lip,
2,298 cleft lip with
Page 57 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
Hydrocephalus,
Microcephalus, Cleft
Palate, Cleft Lip, Cleft Lip
with Cleft Palate,
Esophageal Defects,
Anoreactal Defects,
Hypospadias, Clubfoot,
Reduction Defects
(Upper Limb, Lower
Limb, Upper and Lower
Limb, Limbs
Unspecified), Congenital
Hip Dislocation,
Diaphragmatic Hernia,
Down’s Syndrome and
Exomphalos.
coinciding with
26 to 40 weeks
before delivery.
cleft palate. For 1962-
1965: 9,980 CAs. Live
births only. Analysis by
crude and
standardized (by
season and year)
incidence ratios of high
risk versus low risk
pregnancies and χ2
test.
Leck (Leck
1971)
1971
(Birmingham:
1954-1965,
England and
Wales 1964-
1968, USA:
1955-1965).
Birmingham: Cleft Lip,
Cleft Lip with Cleft Palate
and Limb Reductions.
England and Wales:
Anencephalus, Spina
Bifida, Encephalocele,
Cleft Lip, Cleft Lip with
Cleft Palate, Esophageal
Atresia and Stenosis,
Weekly new
claims to
sickness benefit
coinciding with
26 to 40 weeks
before delivery.
Birmingham: 106 cleft
lip, 200 cleft lip with
cleft palate and 136
limb reductions, live
births and stillbirths.
England and Wales:
13,707 CAs, live births
and stillbirths. USA:
Unknown amount of
Page 58 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Anorectal Atresia and
Stenosis, Limb
Reduction, Exomphalos.
USA: Cleft Lip and Limb
Reductions.
CAs, live births only.
Analysis by crude and
standardized (by
season and year)
incidence ratios of high
risk versus low risk
pregnancies and χ2
test.
Record
(Record
1961)
1961 (1938-
1958),
Scotland.
Anencephaly Winters (Oct-
Mar) with
influenza
deaths >500
were
considered
epidemic
winters and
linked with
anencephaly
incidences the
next May-Oct.
5,039 Anencephaly
cases. Analysis by
visual inspection.
Saxen et
al (Saxen
et al.
1990)
1990 (1968-
1982), Finland
Anencephaly High risk
months were
identified via
the Central
Public Health
248 Anencephaly
cases. Analysis by rate
ratio’s between high
risk and low risk
pregnancies.
Page 59 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
Laboratory and
coinciding
pregnancies in
the 1st
trimester
were
considered high
risk.
Results from these studies are included in Table II. 2
Page 60 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Table IV: Overview of 25 papers from 15 case-control studies sorted by year. 1
1st
Author Publication
year (study
years) and
location
Exposure
ascertainment
Birth types Controls, matching
and adjustment (if
any)
Coffey (Coffey and
Jessop 1955)
1955 (1953-
1954),
Ireland
Exposure was based on
maternal reports,
retrospective to
delivery.
Live and
stillbirths.
Random sample of
non-malformed
controls.
Laurence (Laurence
et al. 1968)
1968 (1956-
1962), UK
The study suggests
exposure was based on
maternal reports,
retrospective to
delivery for 1956-1960
and mainly prospective
to delivery after 1960.
Live and
stillbirths.
Controls without
neural tube defects.
Matched by date of
birth, place of
residence, sex of
infant.
Elizan (Elizan et al.
1969)
1969 (1959-
1964), USA
Exposure was based on
complement fixation
test.
Autopsied
stillbirths.
Controls without
central nervous
system defects.
Matched by
gravidity, institution,
last menstrual
period, maternal age
and race.
Choi (Choi and 1970 (1963- Exposure was based on It is Controls without
Page 61 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
Klaponski 1970) 1968),
Canada
maternal interviews
after birth
supplemented by
medical records.
unknown
which birth
types were
included.
neural tube defects.
Matched by date of
birth, hospital, sex of
infant.
Saxen (Saxen 1975) 1975a
(1967-1971),
Finland
Finnish registry of
congenital
malformations (FRCM).
Exposure was based on
maternal reports,
retrospective to
delivery.
Live and
stillbirths.
Controls without
orofacial clefts.
Matched by place of
residence and time
of birth.
Saxen (Saxen 1975) 1975b
(1972-1973),
Finland
Finnish registry of
congenital
malformations (FRCM).
Exposure was based on
maternal reports,
retrospective to
delivery.
Live and
stillbirths.
Controls without
orofacial clefts.
Matched by place of
residence and time
of birth.
Klemetti (Klemetti
1977)
1977 (1963-
1965),
Finland
Finnish registry of
congenital
malformations (FRCM).
Exposure was based on
maternal reports,
prospective to delivery.
Live and
stillbirths.
Non-malformed
controls and cases
originated from a
series of consecutive
pregnancies
registered in the
same county.
Page 62 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
Granroth (Granroth
1978, Granroth et al.
1978) (+earlier
papers by Karkinen-
Jääskeläinen
(Karkinen-
Jaaskelainen and
Saxen 1974) and
Saxen (Saxen et al.
1974)).
1978 (1965-
1973),
Finland
Finnish registry of
congenital
malformations (FRCM).
Exposure was based on
maternal reports,
retrospective to
delivery and maternity
centres, prospective to
delivery.
Live and
stillbirths.
Non-malformed
controls. Matched
by date of delivery
and Maternity
Welfare District.
Warrell (Warrell et
al. 1981)
1981 (1972-
1976), UK
Exposure was based on
fluorescent-antibody
technique.
It is
unknown
which types
of births
were
included.
Non-malformed
controls.
Matched (as far as
possible) by data of
blood serum
collection,
gestational age and
parity.
Aro (Aro 1983) 1983 (1964-
1977),
Finland
Finnish registry of
congenital
malformations (FRCM).
Exposure was mainly
based antenatal
records, collected
prospective to delivery.
Live and
stillbirths.
Non-malformed
control.
Matched by date of
delivery and place.
Page 63 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Lynberg (Lynberg et
al. 1994) (+ earlier
paper by Park (Park
et al. 1992))
1994 (1968-
1980), USA
Atlanta Birth Defects
Case Control Study
(ABDCCS). Exposure
was based on maternal
reported episodes of
>2 days of flu and fever
from 1 month before
to 3 months after
conception,
retrospective to
delivery.
Live and
stillbirths.
Non-malformed and
malformed controls
(with CA “outside
the hyperthermia
spectrum”).
Matched by hospital
of birth, race, year
and quarter of birth.
Adjusted for alcohol
use, education,
maternal age,
multivitamin use
and smoking.
Botto (Botto et al.
2001) (+ earlier
paper by Adams
(Adams et al. 1989))
2001 (1968-
1980), USA
Atlanta Birth Defects
Case Control Study
(ABDCCS). Exposure
was based on maternal
reports, retrospective
to delivery.
Live and
stillbirths.
Non-malformed
controls.
Matched by
hospital, race and
time of birth.
Adjusted for alcohol
use, chronic illness,
education, race,
multivitamin use,
period of birth and
smoking.
Page 64 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
Li(Li et al. 2007) 2006 (2003-
2005), China
Exposure was based on
maternal reports,
retrospective to
delivery.
Live and
stillbirths.
Non-malformed
controls.
Matched by county,
estimated date of
conception, ethnic
group and sex.
Adjusted for
antibiotics,
antipyretics, diet,
folic acid, history of
major congenital
anomaly affected
pregnancy, maternal
education and
infectious disease
other than influenza.
Czeizel (+ earlier
papers by Acs,
Medveczky and
Metneki) (Acs et al.
2005, Czeizel et al.
2008, Medveczky et
al. 2004, Metneki et
al. 2005) (+ later
paper by Csáky-
2007 (1980-
1996),
Hungary
Hungarian Case-
Control Surveillance of
Congenital Anomalies
(HCCSCA). Exposure
was based on maternal
reports, retrospective
to delivery.
Live and
stillbirths.
Non-malformed
controls.
Matched by district
of residence, sex,
year and week of
birth.
Page 65 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
6
Szunyogh (Csaky-
Szunyogh et al.
2013))
Oster (Oster et al.
2011) (+ later paper
by Kelly (Kelly et al.
2012))
2011 (1981-
1989), USA
Baltimore-Washington
Infant Study (BWIS).
Exposure was based on
maternal reports,
retrospective to
delivery.
Live births. Random sample of
live-bortn controls
without congenital
heart disease.
Frequency matched
on age at interview,
month, year and
hospital of birth.
Adjusted for alcohol
use, BMI, family
history of congenital
heart disease,
gestational diabetes,
race, sex and
smoking.
Results from these studies are included in the meta-analysis (Table I) and Table II. 2
Page 66 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Table V: Overview of 12 papers from 10 cohort studies sorted by year. 1
1st
Author Publication
year (study
years) and
location
Recruitment Exposure ascertainment Matching
additional to
location and
adjustment (if
any)
Campbell
(Campbell
1953)
1953 (1950-
1951), UK
Mothers attending
antenatal clinics
were enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Abramowitz
(Abramowitz
1958)
1958 (1957),
South Africa
Women attending
gynaecology wards
were enrolled.
Exposure ascertainment
was not described.
None
Walker (Walker
and McKee
1959)
1959 (1957),
USA
Ward patients who
delivered at
University hospitals
were enrolled.
Exposure was based on
hemagglutination
inhibition and maternal
reports, retrospective to
delivery.
None
Doll (Doll et al.
1960)
1960 (1957-
1958), UK
Mothers attending
an antenatal clinic
for the first time
during the study
period were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery
and some of these were
confirmed with
professional diagnosis.
None
Pleydell 1960 (1957),
UK
Expecting mothers
reported to suffer
Exposure was based on
maternal reports,
None
Page 67 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
(Pleydell 1960) from ILI were
included and
exposed and non-
exposed were
recruited via local
midwives.
prospective to delivery.
Saxen (Saxen et
al. 1960)
1960 (1958),
Finland
Women delivering
in the University
Central Hospital of
Helsinki were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Hardy (Hardy
et al. 1961)
1961 (1957-
1958), USA
Patients visiting an
obstetrical prenatal
clinic were enrolled.
Exposure was based on
complement fixation,
hemagglutinin inhibition
and maternal reports,
prospective to delivery.
None
Hirvensalo
(Hirvensalo and
Kinnunen 1962)
1962 (1957-
1959),
Finland
Women attending
the Maternity
Health Centres were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Coffey (Coffey
and Jessop
1959, Coffey
and Jessop
1963)
1959 and
1963 follow-
up (1957-
1958), Ireland
Women attending
antenatal clinics
were enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
Stage of
pregnancy.
Page 68 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
Wilson (Wilson
et al. 1959,
Wilson and
Stein 1969)
1959 and
1969 follow-
up (1957),
USA
Prenatal registrants
were enrolled.
Exposure was based on
hemagglutinin inhibition.
None
Results from these studies are included in the meta-analysis (Table I) and Table II. 2
Page 69 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
77x213mm (96 x 96 DPI)
Page 70 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
87x165mm (96 x 96 DPI)
Page 71 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
1
Manuscript title 1
Influenza and congenital anomalies: a systematic review and meta-analysis 2
Suggestion for a running title 3
Influenza and congenital anomalies 4
Authors full names 5
JM Luteijn. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 6
Ulster, Jordanstown Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 7
MJ Brown. Institute of Nursing Research/School of Nursing, University of Ulster, Jordanstown 8
Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 9
H Dolk. EUROCAT Central Registry, Institute of Nursing Research/School of Nursing, University of 10
Ulster, Jordanstown Campus, Shore Road, Newtownabbey, BT37 0QB, United Kingdom 11
Page 72 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
2
Abstract 12
Study question: Does 1st
trimester maternal influenza infection increase the risk of non-13
chromosomal congenital anomalies? 14
Summary answer: 1st trimester maternal influenza exposure is associated with raised risk of a 15
number of non-chromosomal congenital anomalies, including neural tube defects, hydrocephaly, 16
congenital heart defects, , cleft lip, digestive system defects and limb reduction defects. 17
What is known already: Hyperthermia is a well-established risk factor for neural tube defects. 18
Previous studies suggest influenza may be a risk factor not only for neural tube defects, but also 19
other congenital anomalies. No systematic review has previously been undertaken. 20
Study design, size, duration: Systematic review and meta-analysis. A search of EMBASE and PUBMED 21
was performed for English and Dutch studies published up to July 2013. A total of 33 studies (15 22
case-control, 10 cohort and 8 ecological) were included in the systematic review of which 2122 23
studies were included in the meta-analysis. 24
Participants/materials, settings, methods: A total of 27,18129,542 babies with congenital anomaly 25
(9191,112 exposed) from case-control studies and 1,608 exposed pregnancies resulting in 56 babies 26
with congenital anomaly from cohort studies were included in the meta-analysis: 919 from case-27
control studies and 56 from cohort studies. Maternal influenza exposure was defined as any reported 28
influenza, influenza-like illness, or fever with flu, with or without serological or clinical confirmation 29
during the 1st
trimester of pregnancy. Data for 2024 (sub)groups of congenital anomaly available 30
from ≥ 3 studies were analysed using the DerSimonian-Laird random effects model. The hypothesis 31
of publication bias was assessed using funnel plots and risk of bias of included studies was assessed 32
using a slightly modified version of the Newcastle-Ottawa Scale. 33
Main results and the role of chance: 1st trimester maternal influenza exposure was associated with 34
an increased risk of any congenital anomaly (adjusted odds ratio 2.252.00, 95% CI: 1.77-2.851.62-35
Page 73 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
3
2.48), neural tube defects (OR 3.613.33, 2.23-5.822.05-5.40), hydrocephaly (5.775.74, 1.10-36
30.291.10-30.00), congenital heart defects (1.841.56, 1.50-2.261.13-2.14), aortic valve 37
atresia/stenosis (AOR 2.59, 1.21-5.54), ventricular septal defect (AOR 1.59, 1.24-2.14), cleft lip 38
(3.003.12, 2.12-4.252.20-4.42), digestive system (1.731.72, 1.11-2.691.09-2.68) and limb reduction 39
defects (2.062.03, 1.29-3.291.27-3.27). The increased risk for cleft lip (but not for cleft palate), was 40
also reported by ecological studies not included in the meta-analysis. Study outcomes reported for 41
31 27 subgroups of congenital anomaly could not be included in the meta-analysis. Visual inspection 42
of funnel plots did not suggest evidence for publication bias. 43
Limitations, reasons for caution: This study enrolled observational studies which can be subject to 44
limitations such as confounding, retrospective maternal exposure reports and non-response of 45
intended participants. Influenza exposed pregnancies can also have been exposed to influenza 46
related medication. 47
Wider implications of the findings: Prevention of influenza in pregnant women may reduce 48
congenital anomaly risk, and would be relevant to more than just neural tube defects. More research 49
is needed to determine whether influenza and/or its related medication is teratogenic, to determine 50
the role of hyperthermia in teratogenicity and the role of other environmental factors such as 51
nutritional status in determining susceptibility. 52
Study funding/ competing interests: Funded by the EC, under the framework of the EU Health 53
Programme 2008-2013, Grant Agreement 2010 22 04 (Executive Agency for Health & Consumers). No 54
competing interests. 55
Page 74 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
4
Introduction 56
Both during seasonal influenza and pandemic influenza outbreaks, pregnant women have been at 57
risk of increased morbidity and mortality from influenza infection compared to the general 58
population.(Dodds et al. 2007, Harris 1919, Neuzil et al. 1998, Siston et al. 2010) Women in the later 59
stages of pregnancy are particularly vulnerable to adverse health outcomes after influenza infection 60
(Mak et al. 2008), perhaps because of immunological changes that take place during 61
pregnancy.(Jamieson et al. 2006) 62
Unravelling the question of teratogenicity of influenza is complex. In observational studies, influenza 63
exposure can affect the foetus not only via viral infection of the foetus, influenza-induced 64
hyperthermia and toxic metabolites associated with fever (Edwards 2006), but also via antiviral and 65
antipyretic use. A recent systematic review found strong evidence of an association between 66
maternal hyperthermia and neural tube defects.(Moretti et al. 2005) Evidence on other anomalies 67
with respect to hyperthermia or fever is scarce. and aAnimal models have associated maternal 68
hyperthermia with arthrogryposis (Edwards 1971), congenital heart defects (Cockroft and New 1975, 69
Cockroft and New 1978), club foot(Edwards 1971), microcephaly (Edwards 1969, Edwards 1969), 70
microphthalmos (Germain et al. 1985) and others.(Edwards 2006) Evidence on other anomalies with 71
respect to hyperthermia or fever is scarce. 72
The primary method of protecting pregnant women and their unborn child against influenza 73
infection is vaccination. In an increasing number of countries, pregnant women are advised to be 74
vaccinated against seasonal influenza infection.(Mak et al. 2008, Mereckiene et al. 2010) Despite the 75
health benefits of vaccination by protecting pregnant women against influenza infectionHowever, 76
vaccination policies in European countries vary both for seasonal influenza vaccination and during 77
the 2009 H1N1 influenza pandemic, especially with respect to trimesters eligible for 78
vaccination.(Luteijn et al. 2011, Mereckiene et al. 2010) In the absence of consensus on whether or 79
not to vaccinate 1st trimester pregnant women, the hypothesis of a causal relationship between 80
Page 75 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
5
congenital anomalies (CA) and influenza virus deserves renewed attention. A better understanding of 81
the possible relationship between influenza and congenital anomaliesCA will allow for better 82
understanding of the benefit-risk balance of vaccinating 1st trimester pregnant women and women of 83
childbearing age against influenza. 84
The objective of our review is to identify and summarize the available epidemiologic evidence 85
regarding the risk of congenital anomalyCA associated with 1st trimester exposure to maternal 86
influenza. 87
Materials and methods 88
Search strategy 89
This systematic review was informed by PRISMA guidelines (Moher et al. 2009) and the MOOSE 90
group guidelines.(Stroup et al. 2000) Two of the authors (JML and MJB) conducted the various steps 91
of the review and resolved any disagreements by discussion and consensus. The PubMed® and 92
Embase® databases were searched using the MeSH terms ("Influenza, Human") AND (pregnancy OR 93
congenital abnormality) and (Influenza) AND (pregnancy OR congenital abnormality), respectively on 94
July 1st, 2013. No publication or date restrictions were set. Where the papers’ abstract, title or 95
indexed MeSH terms suggested the possibility of reporting any fetal outcomes after maternal 96
exposure to influenza, the full paper was obtained. Reference lists of enrolled papers were reviewed. 97
Eligibility criteria 98
Case-control, cohort and ecological studies investigating congenital anomalyCA outcomes following 99
maternal exposure to influenza were eligible for inclusion. No quality criteria were set for inclusion, 100
although risk of bias analysis was performed (eAppendix). Influenza was defined as any reported 101
influenza, influenza-like illness, or fever with flu, with or without serological or clinical confirmation. 102
Solely studies reporting influenza exposures during the 1st trimester of pregnancy were included in 103
the systematic review and meta-analysis. In order to be included in the meta-analysis, case-control 104
and cohort studies needed to allow for the calculation or ofreport odds ratios (OR) or relative risks 105
Page 76 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
6
(RR). For financial reasons, only English and Dutch language papers were eligible for inclusion. No 106
Dutch papers were included. 107
Data extraction 108
Study characteristics were extracted by JML and MJB (eAppendix, Table III-V). Data Crude and 109
adjusted ORs and RRs and 2x2 tables relevant for meta-analysis waswere extracted by JML and MJB 110
from cohort and case-control studies in 2x2 tables. We contacted three authors of studies published 111
≤10 years ago(Czeizel et al. 2008, Kelly et al. 2012, Oster et al. 2011) to obtain core unavailable data 112
in order to create aggregate groups such as orofacial clefts and for crude OR calculation and received 113
the complete dataset for one study.(Czeizel et al. 2008) In case of studies distinguishing between flu 114
with fever and flu without fever where it was impossible to combine the two, such as the stratified 115
study by Lynberg (Lynberg et al. 1994), the flu with fever dataset was extracted. We extracted data 116
regarding malformed controls, if available, rather than non-malformed controls since use of 117
malformed controls reduces the impact of differential recall bias. We recognize this can lead to 118
underestimation of effect size if CA in the control group are related to influenza exposure. In the 119
meta-analysis, influenza exposures outside of the 1st
trimester were added to the non-exposed 120
cohort and for one cohort study without controls (Doll et al. 1960), this allowed us to form a control 121
group. 122
Congenital anomaliesCA were classified into European surveillance of Congenital Anomalies 123
(EUROCAT) defined subgroups, excluding minor anomalies as specified by EUROCAT.(EUROCAT 124
Central Registry 2009) EUROCAT is a network of population-based congenital anomaly registries 125
which surveys over 1.7 million births annually in 23 European countries. CA were classified down to 126
the greatest level of precision possible. For example, a study that reported anomalies only as neural 127
tube defects without further specification contributed data to the analysis of neural tube defects but 128
could not contribute data to an analysis of anencephaly or spina bifida specifically. We excluded 129
chromosomal syndromes since their etiology is not related to maternal exposures. 130
Page 77 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
7
Risk of bias assessment 131
In order to assess the validity of included studies’ findings we assessed the risk of bias of case-control 132
and cohort studies included in the meta-analysis using a slightly modified version of the Newcastle-133
Ottawa scale (NOS).(Wells et al. 2004) Judgement criteria and deviations from the NOS are 134
summarized in the eAppendix. 135
Quantitative data summary and synthesis 136
Meta-analysis was performed combining adjusted ORs (AOR) and RRs (ARR) i.e. adjusted for 137
confounders). Only four case-control studies provided AORs (Table IV) and none of the cohort studies 138
made statistical adjustments (Table V). Where adjusted estimates were not available, crude 139
estimates were used. 140
We assumed similarity between OR and RR because congenital anomaliesCA are rare events.(Davies 141
et al. 1998) Statistical analysis was performed using Stata version 9.2 [StataCorp, College Station, TX]. 142
Meta-analysis was performed on all anomalies CA combined and, for EUROCAT defined subgroups of 143
anomalies CA (if n studies ≥ 3). The DerSimonian & Laird random effects model (DerSimonian and 144
Laird 1986) was used since the studies in this meta-analysis involved varying countries, time periods 145
and influenza strains. Subgroup analysis was performed based on study type, publication date, and 146
risk of differential recall bias and adjustment for confounders in order to assess the impact of these 147
variables on study outcome. Subgroup analyses combined all studies in the relevant categories, using 148
the estimate for all non-chromosomal CA combined where available and if not available, the 149
estimate for the specific CA subgroup studied. 150
Due to scarce numbers and imbalance between some study arms, we used an alternative continuity 151
correction based on the OR of (other) studies with >0 events in both arms and group ratio imbalance 152
as discussed by Sweeting et al.(Sweeting et al. 2004) Heterogeneity between studies was assessed 153
using the I-squared statistic. Values of I-squared equal to 25%, 50% and 75% were considered to 154
Page 78 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
8
represent low, moderate and high levels of heterogeneity, respectively. The hypothesis of 155
publication bias was assessed using funnel plots (eAppendix, Figures 2428-4351). 156
Results 157
Selection flow 158
The PubMed® database search yielded 1369 papers and the Embase® database search yielded 2649 159
papers (Figure 1). After removing 1121 duplicates, a total of 2897 potentially relevant papers were 160
identified by the literature search. After screening by MeSH terms, titles and abstracts, 2615 papers 161
were excluded and full papers were retrieved for the remaining 282 papers. Of these, a total of 40 162
papers covering 27 studies met the inclusion criteria and were included in the systematic review. 6 163
additional eligible papers were detected by reference tracking, leading to a grand total of 46 included 164
papers covering 33 studies. 165
Study characteristics 166
The 46 enrolled papers were classified as 25 papers covering 15 case-control studies, 12 papers 167
covering 10 cohort studies and 9 papers covering 8 ecological studies. Enrolled studies are 168
summarized by study type in eAppendix, Table III-V and evidence provided by enrolled studies that 169
was not included in the meta-analysis is summarized in Table II. For one case-control study 170
information was limited to a conference abstract.(Choi and Klaponski 1970) Included papers were 171
published between 1953 and 2013, with the median year of publication 1971. 172
Risk of bias assessment 173
Visual inspection of the funnel plots did not suggest evidence for publication bias (eAppendix, 174
Figures 2428-4351). 175
Of the 15 case-control studies (25 papers), 10 studies did not take into account possible confounding 176
by maternal age, socioeconomic class or both. In the majority of these 10 studies some form of 177
matching between cases and controls (usually maternal ward, sex and day of birth) took place. Ten 178
papers relied on retrospective maternal reported influenza episodes (or timing of maternal 179
Page 79 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
9
interviews was unknown), making these studies susceptible to differential recall bias. Of the 180
remaining 5 studies, 2 used serologic confirmation and 3 used prospectively collected antenatal 181
records. The last notable source of possible bias was that in 5 case-control studies over 20% of the 182
cases intended for inclusion were not enrolled, making these studies susceptible to non-response 183
bias (eAppendix). 184
For the 10 cohort studies (12 papers), none took into account possible confounding by maternal age, 185
socioeconomic class or both. For 9 cohort studies, infants were not followed up for at least a year (or 186
unclear), making these studies susceptible to misclassification bias for some congenital anomaliesCA 187
not apparent at birth. 6 of the studies used prospectively collected maternal reports for exposure, 3 188
serologic confirmation, for 1 study exposure ascertainment was not described. The last notable 189
source of possible bias was that in 5 cohort studies the exposed cohort was not drawn from a clearly 190
defined place and time, or failed to enrol ≥80% of the population identified in specified place and 191
time, raising questions over representativeness of the enrolled exposed cohort. 192
Quantitative data summary and synthesis 193
Meta-analysis was possible for data from 21 22 studies, forming groups of ≥3 independent studies for 194
20 24 (sub)groups of EUROCAT defined major CA: any non-chromosomal major CA, neural tube 195
defects, anencephaly, encephalocele, spina bifida, hydrocephaly, congenital heart defects, orofacial 196
clefts (Figures 2-9) and twelve sixteen other CA subgroups (eAppendix, Figures 12-2327). 197
Overall, our meta-analysis involved 27,18129,542 CA cases of which 9191,112 were exposed to 198
influenza in the first trimester of pregnancy and 49,65453,089 controls of which 11211,382 were 199
exposed to influenza in the first trimester of pregnancy from case-control studies. From cohort 200
studies, 1,608 exposed pregnancies resulting in 56 CA plus 14,613 non-exposed pregnancies resulting 201
in 347 CA were enrolled. The enrolled cohort studies were relatively small with only the Coffey and 202
Jessop study reaching over 50 congenital anomaliesCA (including minor anomalies) following 203
maternal influenza exposure. Case-control studies more readily enrol the large number required for 204
Page 80 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
10
research on CA and 6 out of 15 case-control studies managed to enrolled over 500 cases.(Botto et al. 205
2001, Czeizel et al. 2008, Granroth et al. 1978, Laurence et al. 1968, Oster et al. 2011, Saxen 1975) 206
The larger numbers come at a cost and 11 out of 15 case-control studies gathered exposure data by 207
retrospective maternal reports. 208
Meta-analysis discovered statistically significant associations between 1st
trimester influenza 209
exposure and a large number of CA subgroups (Table I) including all non-chromosomal CA combined 210
(OR 2.252.00, 95% CI: 1.77-2.851.62-2.48). Medium heterogeneity was detected for the pooled 211
estimate of all non-chromosomal CA (70.364%). Subgroup analysis showed similar outcomeslower 212
odds ratios for pooled case-control study outcomes (OR 2.251.84, 95% CI: 1.721.49-2.932.27), and 213
than for pooled cohort study outcomes (OR 2.12, 95% CI: 1.20-3.75) while pre-1970 studies reported 214
slightly higher odds ratios (OR 2.47, 95% CI: 1.50-4.70) than studies published after 1970 (OR 215
2.111.71, 95% CI: 1.62-2.761.41-2.08). Overall, studies susceptible to differential recall bias reported 216
a greater lower risk (OR 2.421.92, 95% CI: 1.71-3.411.35-2.72) than studies not susceptible to 217
differential recall bias (OR 2.12, 95% CI: 1.54-2.92). No differences were detected between pooled 218
adjusted and pooled crude estimates (2.15, 1.05-4.42 versus 2.22, 1.78-2.77). 219
Central nervous system defects 220
Associations were detected found for all neural tube defects (OR 3.613.33, 2.23-5.822.05-5.40) and 221
the NTD neural tube defect subgroups anencephaly (OR 4.083.52, 2.14-7.081.69-7.32) and spina 222
bifida (OR 2.682.20, 1.79-4.021.48-3.28). The majority of the 2,500 neural tube defects were 223
reported by Czeizel (n=1,202, AOR 2.492.40, 1.361.30-4.574.40), Li (344, AOR 4.463.06, 3.061.40-224
6.506.67), Laurence (n=551, OR 3.93, 1.37-11.27) and Lynberg (331, AOR 1.721.70, 1.151.10-225
2.572.50). The lower OR reported by Lynberg is related to our preference for malformed controls 226
over healthy controls for the study by Lynberg, which lead to more conservative estimates. The study 227
by Lynberg reports higher AOR when using healthy controls for neural tube defects (AOR 3.43.0, 2.1-228
5.51.9-4.7). We discovered significant heterogeneity in the aggregate groups and NTDneural tube 229
Page 81 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
11
defects, anencephaly, encephalocele and hydrocephaly (Table I). The heterogeneity for neural tube 230
defects, anencephaly, encephalocele and hydrocephaly seems to be driven by the Coffey, Hirvensalo, 231
Pleydell, Saxen (1960) and Wilson studies which all contributed OR of >10 in at least one CA subgroup. 232
The ecological study on 1957 pandemic influenza by Hakosalo, enrolling 27 NTD neural tube defects 233
reported suggestive evidence for a relationship between influenza and NTD neural tube defects, but 234
two larger ecological studies by Leck (n=2,484 (Leck et al. 1969) and n=162 (Leck 1963)) did not find 235
such evidence for any neural tube defectsNTD subgroup. The study by Hakosalo calculated cases 236
back to last menstrual period, while none of the studies by Leck had access to gestational length, 237
making these studies susceptible to misclassification of exposure introduced by assumptions around 238
gestational age. Of all 8 ecological studies included, only three corrected for gestational age.(Busby 239
et al. 2005, Hakosalo and Saxen 1971, Saxen et al. 1990) 240
Orofacial clefts 241
Our meta-analysis detected found an association between orofacial clefts and 1st
trimester influenza 242
exposure (OR 2.151.96, 95% CI: 1.711.33-2.702.91). There was a significant association for cleft lip 243
with or without palate (OR 3.003.12, 2.122.20-4.254.42), but not for cleft palate (OR 1.101.05, 244
0.630.60-.1921.84) and no heterogeneity was detected in these pooled groups (Table II). The 245
majority of the orofacial clefts (n=2,773) were reported by Czeizel (n=1,956) and Saxen (1975a n=591 246
and 1975b, n=194) and these studies report ORs between 1.90 and 2.32. Four of the ecological 247
studies, all by Leck, also reported the association between orofacial clefts and influenza and two of 248
these four studies reported associations for cleft lip ± cleft palate, but not for isolated cleft 249
palate.(Leck 1963, Leck et al. 1969) 250
Congenital heart defects 251
Our meta-analysis detected found an association between congenital heart defectsCHD and 1st
252
trimester influenza exposure (OR 1.841.56, 95% CI: 1.501.13-2.262.14), with very low heterogeneity. 253
The vast majority of the congenital heart defectsCHD reported in the meta-analysis were reported 254
Page 82 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
12
by Botto (n=829, AOR 2.1, 0.8-5.5), Czeizel (n=4,479, OR 1.6, 1.3-1.9) and Oster (n=2,361, AOR 1.11, 255
0.91-1.35). Data reported by Botto (AOR 1.8, 1.1-2.9) and Czeizel (OR 2.0, 1.3-3.2) were suggestive of 256
an association between ventricular septal defects and influenza. It should be noted the study by 257
Botto suffered from a 2-12 year delay between delivery and maternal interview. These findings are 258
contradicted by the Baltimore-Washington Infant Study (BWIS), for which we were unable to obtain 259
core data in 2x2 tables (and therefore could not be included in the meta-analysis, Table II). The BWIS 260
did not find an association with CHD (AOR 1.11, 0.91-1.35), nor for VSD (AOR 1.30, 0.92-1.83). while 261
The BWIS did detect associations for tricuspid atresia (AOR 6.04, 2.36-15.42) and pulmonary valve 262
atresia (AOR 2.71, 1.16-6.32). It should be noted influenza exposure in the BWIS studythe study by 263
Oster could occur from 3 months before pregnancy to the 3rd
month of pregnancy. 264
With respect to specific types of congenital heart defects (Table I), meta-analysis showed aortic valve 265
atresia/stenosis and ventricular septal defect to be associated with 1st
trimester influenza exposure 266
(OR 2.59, 1.21-5.54 and OR 1.59, 1.24-2.04, respectively). No associations were found for atrial septal 267
defect, hypoplastic left heart and transposition of the great vessels. Four congenital heart defect 268
subtypes were only reported by ≤2 studies and therefore not in the meta-analysis (Table II). Of these, 269
two studies showed a consistent absence of association for tetralogy of Fallot, and a consistent and 270
high association for tricuspid atresia and stenosis (Botto 7.9, Oster 6.04). 271
Other anomalies 272
Limb reductions were associated with 1st
trimester influenza exposure by the meta-analysis (OR 273
2.062.03, 1.291.27-3.293.27) and this association is supported by several ecological studies (Table II). 274
An association with anophthalmia/microphthalmia is based on a single study (Table II). There was 275
evidence that there is no association with influenza for hypospadias (OR 0.951.02, 0.700.75-1.281.39) 276
and clubfoot (OR 1.03, 0.83-1.27). 277
Page 83 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
13
Discussion 278
This systematic review provides an overview of the published evidence on influenza exposure during 279
the 1st
trimester of pregnancy and congenital anomaliesCA. Meta-analysis revealed evidence for 280
increases in a wide range of major CA following 1st trimester influenza exposure. The twofold 281
increase in risk of non-chromosomal CA represents an increase in prevalence from 1.8% (EUROCAT 282
Central Registry 2013) to 3.6% of births among 1st
trimester influenza exposed pregnancies. 283
Exposure ascertainment 284
Case-control and cohort studies utilized serologic confirmation and maternal reports (prospective 285
and retrospective) for influenza exposure ascertainment. During influenza season the positive 286
predictive value of persons presenting with ILI for influenza is in the order of 66-77%.(Monto et al. 287
2000, Zambon et al. 2001) Serologic confirmation detects clinical and subclinical infections, which 288
might lead to different results since subclinical infections might affect the pregnant women 289
differently from clinical infections. Arguably, serologic confirmation of exposure is more reliable than 290
maternal reports. Five studies based on serologic confirmation were enrolled in the systematic 291
review (Elizan et al. 1969, Hardy et al. 1961, Walker and McKee 1959, Warrell et al. 1981, Wilson et al. 292
1959, Wilson and Stein 1969) of which three were included in the meta-analysis.(Hardy et al. 1961, 293
Warrell et al. 1981, Wilson et al. 1959, Wilson and Stein 1969) The two serologic studies limited to 294
systematic review did not detect an association between influenza and neural tube defectsNTD 295
(Elizan et al. 1969) and did not detect an increased prevalence of CA among 1957 H2N2 Asian 296
pandemic influenza exposed pregnancies.(Walker and McKee 1959) The three serologic studies in the 297
meta-analysis combined contributed 53 out of 27,584 CA, and it was therefore not possible to 298
examine the effect of exposure ascertainment method on effect size. One of these studies reported a 299
possible association between CA and 1957 H2N2 Asian pandemic influenza (Hardy et al. 1961) while 300
a second study did not.(Wilson et al. 1959, Wilson and Stein 1969) The third study did not detect an 301
association between CA and neural tube defectsNTD.(Warrell et al. 1981) Note that all studies using 302
serologic confirmation were limited to low numbers. 303
Page 84 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
14
It is apparent that maternal reports lead to misclassification of exposure as women might not recall 304
infection, timing of infection relative to pregnancy or misdiagnose another infection for influenza. 305
However, as long as maternal reports are collected prospectivelyprospective to the mother being 306
aware of the malformation (e.g. from medical records or interviews during pregnancy), there is no 307
reason to believe misclassification of influenza exposure will differ between armscases and controls. , 308
and tTherefore prospective maternal reports will not lead to a spurious association, but rather bias 309
the estimate toward the null as cases and controls are subject to similar misclassification. 310
Retrospective maternal reports (e.g. interviews after birth), which are frequently utilized by case-311
control studies, are susceptible to recall bias where mothers of cases have a different motivation to 312
recall early pregnancy exposure than mothers of non-cases. Mothers may differ not only in their 313
tendency to remember the infection, but in their tendency to misinterpret an illness as 314
influenza.(MacKenzie and Houghton 1974) Some of the studies included in the systematic review give 315
an estimate of the differential recall bias’ effect. For example, in the study by Lynberg, the OR for 316
anencephaly after flu with fever decreased from 3.1 (95% CI: 1.6-6.1) when compared with non-317
malformed controls to 1.4 (95% CI: 0.7-2.6) when compared with malformed controls.(Lynberg et al. 318
1994) Part of this decrease might also have been related to CA in the control group being related to 319
influenza, thus biasing the OR towards 1. In our meta-analysis, the pooled estimate of OR for studies 320
susceptible to recall bias (2.421.92) was slightly higher lower than the pooled estimate for other 321
studies (2.122.00), contrary to expectation. This is related to the low overall OR (1.11) reported by 322
the large Baltimore-Washington Infant Study (Oster et al. 2011), which was susceptible to differential 323
recall bias and contributed over 20% to the pooled estimate of susceptible studies. 324
Case-control, cohort and ecological study designs 325
Cohort studies failed to enrol large numbers of CA and this should not be surprising considering CA 326
only make up for 2-3% of births in a general population and short follow-up time after birth could 327
have led to underascertainment. Investigation of specific CA requires even higher numbers. Due to 328
the possibility of enrolling larger numbers, case-control studies seem better suited for addressing the 329
Page 85 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
15
hypothesis of teratogenicity of influenza, although this comes at a cost as most case-control studies 330
ascertained exposure by retrospective maternal reports. 331
Ecological studies are generally considered a weaker study design than case-control or cohort studies 332
(Evans 2003) due to lack of individual exposure information. It cannot be verified that theany excess 333
congenital anomaliesCA occurred among infected individuals. Correlation between influenza and 334
confounding risk factors at group level may lead to ecological fallacy, for example if influenza and 335
nutritional deficiencies co-occur in winter. Ecological studies base exposure status on timing of 336
pregnancy relative to influenza season (or a proxy thereof) and therefore, the cohort defined as 337
“exposed” is diluted by pregnancies that did not have influenza. Population influenza exposure in the 338
eight enrolled ecological studies were derived from influenza incidences, counts or deaths (n=5) and 339
sickness absenteeism rates or claims (n=3). For this reason, the distinguishing power of ecological 340
studies is highly dependent on influenza attack rates and precision used to define influenza 341
seasonexposure. 342
The advantages of ecological studies are that large numbers of patients are enrolled easily and 343
ecological studies are not susceptible to the exposure misclassification or recall bias inherent in 344
individual level studies. Furthermore, they can be free of individual level confounding e.g. if those 345
most susceptible to influenza in the population have other risk factors for congenital anomaliesCA. 346
Ecological studies therefore offer great value for addressing the hypothesis of teratogenicity of 347
infectious diseases and should not be discounted at the bottom of the evidence hierarchy in this area 348
of research. Consistency between study designs lends strength to a causal interpretation.(Hofler 349
2005) 350
Associations between 1st trimester influenza exposure and CA 351
One of the most striking results of the meta-analysis is the association between 1st
trimester 352
influenza exposure and neural tube defectsNTD. NTD Neural tube defects are easily recognized at 353
birth eliminating susceptibility to underascertainment, while a previous meta-analysis reported an 354
Page 86 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
16
association between neural tube defectsNTD and hyperthermia (OR 1.92, 95% CI: 1.62-355
2.29).(Moretti et al. 2005) There also is evidence for neural tube defectsNTD following hyperthermia 356
exposure in guinea pigs.(Smith et al. 1992) In more recent human studies, underascertainment of 357
neural tube defectsNTD could have been a problem due to terminations of pregnancy for fetal 358
anomaly. According to a 2004 EUROCAT analysis, 88% of neural tube defectsNTD are detected 359
prenatally of which 88% are aborted.(Boyd et al. 2008) Neural tube defectsNTD were more 360
frequently studied than other congenital anomaliesCA, possibly a result of interest in the 1950s 361
studies by Coffey, suggesting an alarmingly strong link between maternal influenza exposure and 362
neural tube defects NTD (data corresponds to OR 10.58, 4.30-26.02 in the 1963 follow-up).(Coffey 363
and Jessop 1963) This study utilized standardized questionnaires for maternal interview after 364
delivery and therefore was susceptible to differential recall bias, but this limitation would not explain 365
such a high OR. One of the ecological studies found an increase in neural tube defectsNTD during the 366
1957 Asian influenza outbreak in Finland and concluded this might have been caused either by 367
influenza or influenza-related pharmaceuticals.(Hakosalo and Saxen 1971) The study by Li et al had 368
data available both on antiviral and antipyretic use (and other potential confounders, see Table IV) 369
and after adjustmenting for these co-exposures, OR for neural tube defectsNTD following maternal 370
influenza exposure dropped slightly but remained statistically significant (OR 3.93, 95% CI: 2.48-6.23). 371
Hydrocephaly can sometimes be caused by spina bifida, and we could derive estimates for 372
hydrocephaly not associated with neural tube defectsNTD for two studies of the five studies on 373
hydrocephaly.(Coffey and Jessop 1959, Coffey and Jessop 1963, Hirvensalo and Kinnunen 1962) It is a 374
general problem in congenital anomalyCA research that one baby may have more than one anomaly, 375
and some of these multiple malformed babies may have “sequences” (EUROCAT Central Registry 376
2009) that follow from a primary anomaly. The data available for review cannot distinguish different 377
types of diagnoses. 378
Page 87 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
17
The evidence for an association between 1st
trimester influenza exposure and cleft lip with or 379
without palate is strong due to the lack of heterogeneity in the pooled data and consistent positive 380
associations detected between across different study designs. It is well known that cleft lip +/- palate 381
differs aetiologically from cleft palate, so this difference is not surprising.(Mossey et al. 2009) The 382
associations detected for congenital heart defectsCHD and VSD ventricular septal defects are 383
somewhat contradicted by the results from the BWIS.(Oster et al. 2011) On the other hand 384
congenital heart defectsCHD have been associated with hyperthermia in rats.(Cockroft and New 385
1975, Cockroft and New 1978) Club foot has been associated with hyperthermia in guinea pigs 386
(Edwards 1971), but our meta-analysis did not find evidence for an association between club foot 387
and influenza. A large number of additional associations for other congenital anomalyCA types were 388
detected with more limited underlying evidence. 389
Pathways for mediation of hypothetical teratogenic effect of influenza 390
Influenza can mediate a possible teratogenic effect via multiple pathways and there is a risk of 391
confounding due to the intimate linkage between a disease and its cure. As well as antivirals, 392
antipyretics are also often used during influenza infection and the case-control study by Li et al 393
reported an AOR for antipyretic drugs and neural tube defectsNTD of 4.86 (95% CI: 1.33-17.78).(Li et 394
al. 2007) Associations not adjusted for antivirals or antipyretics remain of importance since from a 395
vaccination policy perspective, it is less relevant whether the influenza virus or the antivirals are 396
causing any possible anomalies as vaccination will prevent both exposures. We recognize this puts 397
limits on generalizability of study findings between populations with different use of 398
antipyretics/antivirals. 399
Direct pathways via which influenza infection can possibly lead to CA are toxic metabolites caused by 400
fever, hyperthermia and the influenza virus crossing the placenta. Hyperthermia has been associated 401
with causing neural tube defectsNTD as discussed above.(Moretti et al. 2005) It should be noted this 402
meta-analysis involved a large number of possible causes of hyperthermia, while influenza causes 403
Page 88 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
18
high fever which might be different from general hyperthermia. This could explain the higher OR 404
reported in this meta-analysis for influenza exposure and neural tube defects NTD (Table I). Several 405
of the included studies distinguished between 1st
trimester influenza and 1st
trimester fever.(Aro 406
1983, Botto et al. 2001, Klemetti 1977, Lynberg et al. 1994, Oster et al. 2011, Saxen 1975, Saxen 1975) 407
One study found an association for influenza, but not for fever (Klemetti 1977), another found an 408
association for influenza with fever and neural tube defects NTD (OR 1.7, 1.1-2.5), which was lowered 409
for influenza without fever (OR 1.3, 0.7-2.5).(Lynberg et al. 1994) A study on congenital heart 410
defectsCHD reported associations for fever (OR 1.8, 1.4-2.4) and influenza (OR 2.1, 0.8-5.5).(Botto et 411
al. 2001) while a second study on CHDcongenital heart defects reported very similar low and non-412
significant excesses for fever (OR 1.14, 0.89-1.46) and influenza (OR 1.11, 0.91-1.35).(Oster et al. 413
2011) The two studies by Saxen on orofacial clefts reported comparable associations for influenza 414
(RR 2.00 for the 1st study) and fever (RR 1.96 for the 1
st study) (Saxen 1975, Saxen 1975), and the 415
study by Aro on limb reduction defects reported an OR of 1.6 for fever and 1.9 for influenza.(Aro 416
1983) It can be concluded that included studies generally reported equivalent or stronger 417
associations for influenza than for fever with respect to CA following 1st
trimester exposure. 418
Another possible pathway by which the influenza virus can mediate a teratogenic effect is placental 419
transmission. Placental transmission has been documented but appears to be rare.(Gu et al. 2007, 420
McGregor et al. 1984) Toxic metabolites associated with fever as a cause of congenital anomalies 421
have also been suggested.(Edwards 2006) 422
Limitations of the study 423
The systematic review results should be interpreted in the light of the findings that all of the included 424
studies are observational studies and many were susceptible to several types of bias and 425
ascertainment of exposure might not have been reliable. Although the funnel plots did not provide 426
evidence for publication bias, it should be noted that most of the included studies reported a wide 427
range of positive associations raising the question of whether studies reporting negative results 428
Page 89 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
19
remained unpublished. Reference tracking identified 6 new studies and these studies were missed 429
because they were not indexed as influenza studies. This leaves the possibility open that some, 430
particularly negative, studies were missed due to poorly indexed terms. A possible reason for this is 431
that some case-control studies investigate a wide range of possible causes of CA and may tend to be 432
selectively indexed for the positive associations. For older studies, we could not be sure whether 433
chromosomal CA were excluded from analysis. 434
A weakness of the meta-analysis was that adjustment for confounders was not possible since some 435
studies did not adjust for confounders and confounder data could not be extracted from papers that 436
did adjustnot performed in most included studies. Adjustment for confounders showed a moderate 437
effect on OR within studies that did report both crude OR and adjusted OR.(Acs et al. 2005, Granroth 438
et al. 1978, Li et al. 2007) Subgroup analysis comparing adjusted versus crude estimates did not 439
detect differences, but this could be related to the fact that 50% of the adjusted estimate was 440
composed of neural tube defects data since studies generally reported higher estimates for neural 441
tube defects than for other CA. Due to the limited amount of studies reporting adjusted OR, 442
comparison of crude and adjusted OR for subgroups of the same CA was not possible. Some very 443
large datasets involved matched and/or stratified controls (Botto et al. 2001, Czeizel et al. 2008), and 444
most other datasets were matched by one or more variables (Aro 1983, Granroth 1978, Granroth et 445
al. 1978, Karkinen-Jaaskelainen and Saxen 1974, Laurence et al. 1968, Li et al. 2007, Lynberg et al. 446
1994, Saxen 1975, Saxen 1975, Warrell et al. 1981) lowering the impact of confounding on the meta-447
analysis. 448
Statistical limitations 449
The study was susceptible to statistical limitations unique to meta-analysis of scarce events. In the 450
context of scarce events like CA and complicated exposure like 1st
trimester influenza, cohort studies 451
can enrol large numbers of exposed and unexposed, but have very few exposed congenital 452
anomalyCA outcomes. Current statistical methods will assign these studies a lot of weight compared 453
Page 90 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
20
to case-control studies with many cases and greater numbers of exposed cases. An example from this 454
systematic review is the cohort study by Pleydell, which reported 1 case of hydrocephaly among 12 455
1st trimester influenza exposed pregnancies and 1 case among 1071 unexposed pregnancies leading 456
to an OR of 97.27. The heterogeneity model favours outliers and smaller studies and provided this 457
study with 20% weight in the overall hydrocephaly estimate, compared to 43% weight for the case-458
control study by Czeizel which enrolled 314 cases of hydrocephaly (16 exposed). The weight allocated 459
by the heterogeneity model to the Pleydell study is clearly disproportionate. 460
A second problem lies in the continuity correction which is used to address zero events in one of 461
both arms of a study when pooling odds ratios. For rare events like CA, zero events in one or both 462
arms are not uncommon. The standard value for continuity correction is 0.5, but this arbitrary value 463
causes problems in studies with uneven arms and can even dominate the other arm. We addressed 464
this problem as proposed by Sweeting (Sweeting et al. 2004) by letting the continuity correction 465
depend on the OR of (other) studies with >0 events in both arms and group ratio imbalance. 466
However, this still led to OR > 1 for studies reporting 0 events in the exposed arm such as the 467
hydrocephaly data by Hirvensalo. 468
Conclusions and implications for CA prevention 469
Given the risk of congenital anomaly associated with influenza we show here, prevention of influenza 470
by vaccinating women who are planning to get pregnant may reduce congenital anomaly risk. 471
However, before evidence based policy can be implemented, further safety data on use of influenza 472
vaccines in pregnancy with respect to congenital anomalies is required (Kallen and Olausson 2012, 473
Pasternak et al. 2012). Other methods for preventing influenza in early pregnancy include improving 474
nutritional and general health status and adopting behaviours which prevent interpersonal spread. 475
In conclusion, prevention of influenza in pregnant women may reduce congenital anomaly risk, and 476
would be relevant to more than just neural tube defects. More research is needed to determine 477
whether influenza and/or its related medication is teratogenic, to determine the role of 478
Page 91 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
21
hyperthermia in teratogenicity and the role of other environmental factors such as nutritional status 479
in determining susceptibility. 480
Prevention of 1st
trimester pregnant women and women planning to get pregnant in influenza season 481
can possibly prevent occurrence of CA, most notably NTD and cleft lip, induced either by influenza or 482
related medication. Before evidence based policy can be implemented, safety data on use of 483
influenza vaccines in pregnancy with respect to congenital anomalies is required and as argued by 484
Parisi et al a new attitude towards research, medication and pregnancy might be required. 485
Declaration of authors roles 486
J.M.L. (corresponding author, first author) was involved in study design, data collection and synthesis, 487
manuscript preparation and revision, construction of tables and figures, statistical analysis and 488
submission of the manuscript. M.J.B. (co-author) was involved in data collection and manuscript 489
revision. H.D. (co-author) was involved in study design, writing and revising the manuscript. 490
Acknowledgements 491
We would like to thank Dr. I. Barisic for assistance with case classification and we would like to thank 492
Dr. Julia Métneki, Dr. Erzsébet Puhó, Professor Andrew Czeizel and Dr. Annukka Ritvanen for sending 493
additional data. We would like to thank Professor Lolkje de Jong-van den Berg enand Dr. Gordon 494
Marnoch for their helpful comments. 495
Funding 496
Funded by the EC, under the framework of the EU Health Programme 2008-2013, Grant Agreement 497
2010 22 04 (Executive Agency for Health & Consumers). 498
Conflict of interest 499
No competing interests 500
Page 92 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
22
Table I: Heterogeneity and number of cases for 1st
First trimester maternal influenza exposure and 501
risk of congenital anomalies.: studies, total number of cases of congenital anomaly, pooled odds ratio 502
and heterogeneity. 503
Group Participating
studies (n)
I2
Statistic for
heterogeneity (%)
Pooled OR (95%
CI)
Total number
of CA (n)
Any congenital anomaly 2122 7064 2.25 (1.77-
2.85)2.00 (1.62-
2.48)
27,58429,945bc
-Susceptible to
differential recall bias
89 6565 2.421.92
(1.711.35-
3.412.72)
3,0655,426
-Not susceptible to
differential recall bias
13 64 2.12 (1.54-2.91) 24,519bc
-Case-control studies 1213 7560 2.251.84
(1.721.49-
2.932.27)
27,18129,542bc
-Cohort studies 9 62 2.12 (1.21-3.72) 403
-Any type, published
1955-1969
11 58 2.47 (1.50-4.70) 1,171
-Any type, Published
19745-200711
1011 7655 2.111.71
(1.621.41-
2.762.08)
26,41328,774
-Adjusted estimates only a 4 87 2.15 (1.05-4.42) 3,865
-Crude estimates only a 21 61 2.22 (1.78-2.77) 27,584
Neural Tube Defects 11 6050 3.613.33
(2.232.05-
2,500
Page 93 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
23
5.825.40)
-Anencephaly 10 4744 4.083.52
(2.141.69-
7.807.32)
608
-Encephalocele 4 6463 3.022.95
(0.790.78-
11.4811.13)
225
-Spina Bifida 7 190 2.682.20
(1.791.48-
4.023.28)
1,093
Hydrocephaly 5 45 5.775.74 (1.10-
30.2930.00)
323
Congenital heart
defectsCHD
910 541 1.841.56
(1.501.13-
2.262.14)
5,3547,715
-Aortic Valve
Atresia/Stenosis
3 31 2.59 (1.21-5.54) 167
-Atrial Septal Defect 3 0 0.82 (0.45-1.51) 429
-Hypoplastic Left Heart 3 0 1.58 (0.94-2.64) 203
-Transposition of the
Great Vessels
3 0 1.40 (0.90-2.17) 321
-Ventricular Septal Defect 34 0 2.061.59
(1.451.24-
2.922.04)
9551,434
Orofacial Clefts 10 037 2.151.96
(1.711.33-
2,773cd
Page 94 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
24
2.702.91)
-Cleft Lip +- Palate ab
7 0 3.003.12
(2.122.20-
4.254.42)
1,404
-Cleft Palate ab
3 0 1.101.05
(0.630.60-
1.921.84)
584
Digestive System ab
4 0 1.731.71
(1.111.09-2.69)
1,195
Urinary 5 30 1.441.45
(0.910.90-
2.292.34)
48
Hypospadias ab
4 0 0.951.02
(0.700.75-
1.281.39)
3,041
Limb Reduction 3 0 2.062.03
(1.291.27-
3.293.27)
1,002
Club Foot ab
4 0 1.031.11
(0.830.93-
1.271.34)
2,430
Hip Dislocation/Dysplasia 3 0 0.300.31 (0.00-
37.0837.62)
37
Polydactyly ab
4 0 1.681.72
(0.840.85-
3.383.48)
1,094
Page 95 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
25
a Three studies (Botto et al. 2001, Li et al. 2007, Lynberg et al. 1994) were able to provide estimates 504
both for adjusted and crude OR.
505
ab Note that a single study contributed over 90% of the total weight to this pooled estimate
506
bc Note that for some studies such as Czeizel et al 2008, the “any congenital anomaly” group also 507
included CA which were not included in any other analysis.
508
cd Note that 591 orofacial clefts from the Saxen 1975a study (Saxen 1975) and 194 orofacial clefts 509
from the Saxen 1975b study (Saxen 1975) were not specified and therefore solely included in the 510
overall orofacial clefts analysis. 511
Syndactyly 3 6871 1.961.98
(0.180.19-
21.0320.563)
662
Musculo-Skeletal 3 0 1.031.05
(0.130.16-
8.106.97)
776
Page 96 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
26
Table II: Evidence included in the systematic review relating to 1st trimester influenza exposure and 512
CA not included in meta-analysis. 513
Anomaly Evidence reported with respect to 1st trimester influenza
exposure
Any anomaly Visual inspection: no convincing evidence that infant death rates
from CA are associated with 1st trimester influenza
exposure(Buck 1955); RR 1.10(Leck 1963); single defect RR 1.00,
multiple defects RR 0.9(Leck 1964); single defect RR 1.03,
multiple defects RR 1.07(Leck et al. 1969)
CNS Central Nervous System
defects
0.81% in influenza exposed group vs. 0.30% and 0.32% in control
groups(Hakosalo and Saxen 1971)
NTDNeural Tube Defects p=0.03, no data(Choi and Klaponski 1970)
Anencephaly RR 0.82(Leck 1963); sRR 1.32(Leck et al. 1969); RR 1.0, 0.8-
1.3(Saxen et al. 1990) No rise in anencephaly rates following
influenza epidemics detected(Record 1961)
Spina Bifida + Encephalocele RR 0.99(Leck 1963), sRR 1.13(Leck et al. 1969)
Hydrocephaly sRR 0.93(Leck et al. 1969)
Microcephaly OR 1.1, 0.3-4.1(Czeizel et al. 2008); sRR 0.24(Leck et al. 1969)
Anophthalmos/Microphthalmos OR 1.26, 1.02-1.57(Busby et al. 2005)
Congenital Cataract OR 0.8, 0.3-2.2(Czeizel et al. 2008)
Congenital Glucaoma OR 2.0, 0.4-10.9(Czeizel et al. 2008)
Congenital heart defectsCHD (suggestive but not significant; 0.93% in influenza exposed group
vs. 0.47% and 0.73% in control groups)(Hakosalo and Saxen
1971); AOR 1.11, 0.91-1.35(Oster et al. 2011)
Transposition of great vessels AOR 2.1, 0.8-5.5(Botto et al. 2001); OR 2.2, 0.3-15.8(Czeizel et al.
2008); AOR 1.21, 0.73-2.02(Oster et al. 2011)
Page 97 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
27
Ventricular Septal Defect AOR 2.0, 1.1-3.6(Botto et al. 2001); AOR 1.30, 0.92-1.83(Oster et
al. 2011)
Atrial Septal Defect OR 1.0, 0.1-7.4(Botto et al. 2001); OR 3.0, 0.3-35.8(Czeizel et al.
2008); AOR 0.73, 0.38-1.42(Oster et al. 2011)
Atrioventricular Septal Defect OR 2.0, 0.3-15.3(Botto et al. 2001); AOR 1.29, 0.82-2.01(Oster et
al. 2011)
Tetralogy of Fallot OR 0.5, 0.1-3.6(Botto et al. 2001); AOR 0.78, 0.37-1.62(Oster et
al. 2011)
Triscuspid atresia and stenosis AOR 7.9, 0.3-29.6(Botto et al. 2001); AOR 6.04, 2.36-15.42(Oster
et al. 2011)
Ebstein's anomaly OR 3.0, 0.4-23.9(Botto et al. 2001)
Pulmonary valve stenosis AOR 1.21, 0.71-2.04(Oster et al. 2011)
Pulmonary valve atresia AOR 2.71, 1.16-6.32(Oster et al. 2011)
Aortic valve atresia/stenosis OR 4.0, 0.9-17.9(Botto et al. 2001); OR 4.6, 1.4-14.9(Czeizel et al.
2008); AOR 1.53, 0.70-3.34(Oster et al. 2011)
Hypoplastic left heart OR 1.6, 0.4-6.7(Botto et al. 2001);OR 1.6, 0.4-6.1(Czeizel et al.
2008); AOR 1.57, 0.86-2.88(Oster et al. 2011)
Coarctation of aorta AOR 3.8, 1.6-8.8(Botto et al. 2001); AOR 0.41, 0.13-1.33(Oster et
al. 2011)
Total anomalous pulm venous OR 2.2, 0.3-16.9(Botto et al. 2001)
Cleft Lip RR 1.55, p<0.05(Leck 1963); isolated RR 1.4, with other defects
RR 6.3(Leck 1964); cleft lip without cleft palate sRR 1.47,
p<0.05(Leck et al. 1969); cleft lip without cleft palate RR 1.64,
p<0.05(Leck 1971); cleft lip without cleft palate sRR 1.08, cleft lip
with cleft palate sRR 1.01(Leck et al. 1969); cleft lip with cleft
palate sRR 1.10(Leck et al. 1969), cleft lip with cleft palate RR
Page 98 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
28
1.35(Leck 1971)
Cleft Palate RR 0.81(Leck 1963); sRR 0.94(Leck et al. 1969); sRR 0.93(Leck et
al. 1969)
Digestive System 1 in 63 exposed during 1st
trimester of pregnancy and 4 in 1106
non-exposed, RR 4.4(Coffey and Jessop 1959, Coffey and Jessop
1963); 1 in 171 exposed during 1st
trimester of pregnancy and 13
in 6720 non-exposed, RR 3.0(Hirvensalo and Kinnunen 1962)
Oesophageal atresia
with/without
tracheo-oesophageal fistula
OR 1.6, 0.2-11.7(Czeizel et al. 2008); RR 2.41, p<0.01(Leck 1963);
isolated RR 1.5, with other defects RR 4.8(Leck 1964); sRR
1.12(Leck et al. 1969)
Atresia/Stenosis of the small
intestine
OR 2.1, 0.5-8.0(Czeizel et al. 2008)
Ano-rectal atresia and stenosis OR 1.1, 0.4-3.0(Czeizel et al. 2008); RR 2.12, p<0.05(Leck 1963);
isolated RR 0.8, with other defects RR 3.6(Leck 1964); sRR
0.80(Leck et al. 1969)
Hirschsprung's disease OR 0.7, 0.2-3.0(Czeizel et al. 2008)
Diaphragmatic Hernia OR 3.2, 1.2-8.8(Czeizel et al. 2008); 0 in 171 exposed during 1st
trimester of pregnancy and 3 in 6720 non-exposed(Hirvensalo
and Kinnunen 1962); RR 1.05(Leck 1963); sRR 1.01(Leck et al.
1969)
Abdominal Wall Defects OR 2.8, 1.1-6.9(Czeizel et al. 2008)
Omphalocele RR 1.92, <0.05(Leck 1963); isolated RR 1.3, with other defects RR
2.3(Leck 1964); sRR 1.16(Leck et al. 1969)
Urinary 1 in 171 exposed during 1st
trimester of pregnancy and 42 in
6720 non-exposed(Hirvensalo and Kinnunen 1962)
Bilateral renal agenesis including OR 0.8, 0.2-2.8(Czeizel et al. 2008); RR 1.38(Leck 1963)
Page 99 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
29
Potter syndrome
Congenital Hydronephrosis RR 1.63, p=0.05(Leck 1963)
Hypospadias RR 1.51(Leck 1963); sRR 0.93(Leck et al. 1969)
Limb sRR 0.75(Leck et al. 1969)
Limb Reduction Limited to thumbs or radii RR 2.20(Leck 1963); both arms or legs
RR 1.30(Leck 1963); one arm or leg RR 1.37(Leck 1963); sRR
1.32(Leck et al. 1969); RR 1.22(Leck 1971)
Upper Limb Reductions sRR 1.91, p<0.01(Leck et al. 1969)
Lower Limb Reductions sRR 1.23(Leck et al. 1969)
Club Foot sRR 0.89(Leck et al. 1969)
Hip Dislocation and/or Dysplasia sRR 0.77(Leck et al. 1969)
Disorders of skin 0 in 171 exposed during 1st
trimester of pregnancy and 3 in 6720
non-exposed(Hirvensalo and Kinnunen 1962)
Down’s Syndrome sRR 1.16(Leck et al. 1969)
Solely data involving >2 CA cases was included and some reported CA could not be translated to 514
EUROCAT defined subgroups. We did not distinguish between “positive” and “negative” findings 515
since some of the studies were severely underpowered. Note for some CA, ≥3 estimates were 516
available from studies that could not be included in the meta-analysis. 517
Page 100 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
30
Figure 1: Flow chart of systematic review 518
[Figure 1] 519
Figure 2: Forest plot of non-chromosomal CA following 1st
trimester influenza exposure 520
[Figure 2] 521
Figure 3: Forest plot of neural tube defects following 1st trimester influenza exposure 522
[Figure 3] 523
Figure 4: Forest plot of anencephaly following 1st
trimester influenza exposure 524
[Figure 4] 525
Figure 5: Forest plot of encephalocele following 1st
trimester influenza exposure 526
[Figure 5] 527
Figure 6: Forest plot of spina bifida following 1st
trimester influenza exposure 528
[Figure 6] 529
Figure 7: Forest plot of hydrocephaly following 1st
trimester influenza exposure 530
[Figure 7] 531
Figure 8: Forest plot of congenital heart defects following 1st
trimester influenza exposure 532
[Figure 8] 533
Figure 9: Forest plot of orofacial clefts following 1st
trimester influenza exposure 534
[Figure 9] 535
Page 101 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
31
eAppendix 536
Quality assessment of case-control and cohort studies enrolled in meta-analysis 537
For quality assessment of studies included in the meta-analysis, we used a slightly modified version 538
of the Newcastle-Ottawa scale (NOS).(Wells et al. 2004) We did not distinguish between community-539
based or hospital-based controls since most studies would recruit from antenatal clinics it is safe to 540
assume that this population is similar to a community-recruited population. In addition, the risk of 541
referrals would not have been present in studies from before the late 1980s before prenatal 542
screening was widely available. There is some risk of bias introduced from high risk pregnancies 543
being more likely to book at specific hospitals. The following NOS items were not assessed: due to 544
the unique nature of CA we did not require case-control studies to demonstrate that controls did not 545
have a history of the outcome of interest and we did not require cohort studies to provide 546
demonstration that the outcome of interest was not present at start of study. Quality scores were 547
not used since weighting and calculating overall quality scores is arbitrary and the importance of 548
individual items and direction of bias depends on the context in which they are applied. 549
Case-control studies 550
Case definition was considered low risk of bias if the source reported diagnosis by healthcare 551
professional, hospital or primary records, or utilized independent validation. Representativeness of 552
cases were considered low risk of bias if cases were drawn from a clearly defined period of time and 553
catchment area, or consisted of an appropriate random sample of cases. We considered selection of 554
controls low risk of bias if controls could have been in the case group had the outcome of interest 555
been present as opposed to controls from different time periods or locations than cases. For 556
assessing risk of confounding (comparability in NOS) we considered any comparative study which did 557
match or adjust for confounding by at least maternal age and socioeconomic class low risk of bias. 558
For socioeconomic class, several proxies such as malnutrition for older studies and race for American 559
studies were also considered satisfactory. Fever and antiviral use were not considered confounders 560
Page 102 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
32
for reasons stated in the introduction. For ascertainment of exposure, we considered serological 561
confirmation, diagnosis by healthcare professionals and prospective interviews low risk of bias. Post 562
delivery maternal interviews were considered high risk of bias since mothers of infants born with 563
anomalies might make a more determined effort to recall illnesses during pregnancy than other 564
mothers leading to differential recall bias. In case-control designs, recall bias has been estimated to 565
bias OR up to a factor of 1.9.(Rockenbauer et al. 2001) We considered attrition or non-response rates 566
≥20% for cases or controls high risk of bias. 567
Cohort studies 568
Representativeness of the exposed cohort was considered low risk of bias if the exposed cohort was 569
drawn from a clearly defined period of time and catchment area, or consisted of an appropriate 570
random sample of eligible cohort subjects. If ≤80% of the identified individuals in this place and time 571
failed to enrol, representativeness was also considered high risk of bias. We considered selection of 572
the unexposed cohort low risk of bias if individuals in the unexposed cohort could have been in the 573
exposed cohort had they been exposed to 1st
trimester influenza infection. Ascertainment of 574
exposure was handled the same way as for the case-control studies (see above). Length of follow-up 575
is only relevant for certain congenital anomalies such as congenital heart defects and we considered 576
a length of follow up of 1 year satisfactory. With respect to adequacy of follow up, we considered 577
≤20% attrition low risk of bias. 578
Table III: Overview of 9 papers from 8 ecological studies included in the systematic review. 579
Authors Publication
year (study
years) and
location
Study outcome Influenza
exposure and
critical period
No. of cases and
methodology
Buck
(Buck
1955
(Jan 1944-Jul
Infant death due to CA Influenza
incidence per
Unclear population
size, infant deaths only.
Page 103 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
33
1955) 1951), Canada calendar month
coinciding with
1st trimester of
pregnancy
Analysis by visual
inspection.
Busby et
al (Busby
et al.
2005)
2005 (1988-
1994), England
Anophthalmos and
microphthalmos
Reported
weekly
influenza
infection counts
coinciding with
6th-10th
week
of gestation.
275 malformed infants,
live births and
stillbirths. Analysis by
Poisson regression
comparing cases to all
births.
Hakosalo
and Saxen
(Hakosalo
and Saxen
1971)
1971 (Jan 15th
-
Oct 31st
, 1958),
Finland
Anencephaly and CA of
the CNS, circulatory
system and urogenital
system.
Sickness
absenteeism
rates coinciding
with 5th
-11th
week of
gestation.
90 malformed infants,
live births and
stillbirths. Analysis by
χ2 test, comparing high
risk versus low risk
pregnancies. High risk
pregnancies were
exposed to Pandemic
Asian Flu in 5th
-11th
week of gestation. Low
risk pregnancies were
conceived earlier
(Group 1) or conceived
later (Group 2)
Leck (Leck 1963 and 1964 Anencephaly, Spina Weekly new 939 malformed infants,
Page 104 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
34
1963, Leck
1964)
(1957-
1961(Leck
1963), 1957-
1963(Leck
1964)),
Birmingham,
UK
Bifida, Cleft Lip, Cleft
Palate, Oesophageal
Atresia, Anal Atresia,
Renal Agenesis or
Hypoplasia,
Hydronephrosis,
Hypospadias, Limb
Reductions (Thumb or
Radii, Bilateral arms or
legs, single arm or leg),
Exomphalos and
Diaphragmatic
Hernia.(Leck 1963)
claims to
sickness benefit
coinciding with
26 to 40 weeks
before delivery.
live births and
stillbirths.(Leck 1963)
15 additional 1957-
1961 cases and
unknown amount of
1962-1963 cases.(Leck
1964)
Analysis by incidence
ratio’s of high risk
versus low risk
pregnancies.
Leck et al
(Leck et al.
1969)
1969 (1956-
1965), Several
USA states
Orofacial Clefts (1956-
1965) and any CA (1961-
196%). Anencephalus,
Spina Bifida,
Hydrocephalus,
Microcephalus, Cleft
Palate, Cleft Lip, Cleft Lip
with Cleft Palate,
Esophageal Defects,
Anoreactal Defects,
Hypospadias, Clubfoot,
Reduction Defects
(Upper Limb, Lower
Weekly number
of deaths
attributable to
influenza
coinciding with
26 to 40 weeks
before delivery.
For 1956-1961: 1,496
isolated cleft palate,
1,418 isolated cleft lip,
2,298 cleft lip with cleft
palate. For 1962-1965:
9,980 CAs. Live births
only. Analysis by crude
and standardized (by
season and year)
incidence ratios of high
risk versus low risk
pregnancies and χ2
test.
Page 105 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
35
Limb, Upper and Lower
Limb, Limbs
Unspecified), Congenital
Hip Dislocation,
Diaphragmatic Hernia,
Down’s Syndrome and
Exomphalos.
Leck (Leck
1971)
1971
(Birmingham:
1954-1965,
England and
Wales 1964-
1968, USA:
1955-1965).
Birmingham: Cleft Lip,
Cleft Lip with Cleft Palate
and Limb Reductions.
England and Wales:
Anencephalus, Spina
Bifida, Encephalocele,
Cleft Lip, Cleft Lip with
Cleft Palate, Esophageal
Atresia and Stenosis,
Anorectal Atresia and
Stenosis, Limb
Reduction, Exomphalos.
USA: Cleft Lip and Limb
Reductions.
Weekly new
claims to
sickness benefit
coinciding with
26 to 40 weeks
before delivery.
Birmingham: 106 cleft
lip, 200 cleft lip with
cleft palate and 136
limb reductions, live
births and stillbirths.
England and Wales:
13,707 CAs, live births
and stillbirths. USA:
Unknown amount of
CAs, live births only.
Analysis by crude and
standardized (by
season and year)
incidence ratios of high
risk versus low risk
pregnancies and χ2
test.
Record
(Record
1961 (1938-
1958),
Anencephaly Winters (Oct-
Mar) with
5,039 Anencephaly
cases. Analysis by
Page 106 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
36
1961) Scotland. influenza
deaths >500
were
considered
epidemic
winters and
linked with
anencephaly
incidences the
next May-Oct.
visual inspection.
Saxen et
al (Saxen
et al.
1990)
1990 (1968-
1982), Finland
Anencephaly High risk
months were
identified via
the Central
Public Health
Laboratory and
coinciding
pregnancies in
the 1st
trimester
were
considered high
risk.
248 Anencephaly
cases. Analysis by rate
ratio’s between high
risk and low risk
pregnancies.
Results from these studies are included in Table II. 580
Page 107 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
37
Table IV: Overview of 25 papers from 15 case-control studies sorted by year. 581
1st
Author Publication
year (study
years) and
location
Exposure ascertainment Birth types Controls, matching and
adjustment (if any)
Coffey (Coffey and Jessop 1955) 1955 (1953-
1954), Ireland
Exposure was based on maternal
reports, retrospective to delivery.
Live and
stillbirths were
included.
Random sample of non-malformed
controls.
Laurence (Laurence et al. 1968) 1968 (1956-
1962), UK
The study suggests exposure was based
on maternal reports, retrospective to
delivery for 1956-1960 and mainly
prospective to delivery after 1960.
Live and
stillbirths were
included.
Controls without neural tube
defects.
Matched by date of birth, place of
residence, sex of infant.
Elizan (Elizan et al. 1969) 1969 (1959-
1964), USA
Exposure was based on complement
fixation test.
Autopsied
stillbirths were
included.
Controls without central nervous
system defects.
Matched by gravidity, institution,
last menstrual period, maternal
Page 108 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
38
age and race.
Choi (Choi and Klaponski 1970) 1970 (1963-
1968), Canada
Exposure was based on maternal
interviews after birth supplemented by
medical records.
It is unknown
which birth
types were
included.
Controls without neural tube
defects.
Matched by date of birth, hospital,
sex of infant.
Saxen (Saxen 1975) 1975a (1967-
1971), Finland
Finnish registry of congenital
malformations (FRCM). Exposure was
based on maternal reports, retrospective
to delivery.
Live and
stillbirths were
included.
Controls without orofacial clefts.
Matched by place of residence and
time of birth.
Saxen (Saxen 1975) 1975b (1972-
1973), Finland
Finnish registry of congenital
malformations (FRCM). Exposure was
based on maternal reports, retrospective
to delivery.
Live and
stillbirths were
included.
Controls without orofacial clefts.
Matched by place of residence and
time of birth.
Klemetti (Klemetti 1977) 1977 (1963-
1965), Finland
Finnish registry of congenital
malformations (FRCM). Exposure was
based on maternal reports, prospective
Live and
stillbirths were
included.
Non-malformed controls and cases
originated from a series of
consecutive pregnancies registered
Page 109 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
39
to delivery. in the same county.
Granroth (Granroth 1978, Granroth et
al. 1978) (+earlier papers by
Karkinen-Jääskeläinen (Karkinen-
Jaaskelainen and Saxen 1974) and
Saxen (Saxen et al. 1974)).
1978 (1965-
1973), Finland
Finnish registry of congenital
malformations (FRCM). Exposure was
based on maternal reports, retrospective
to delivery and maternity centres,
prospective to delivery.
Live and
stillbirths were
included.
Non-malformed controls. Matched
by date of delivery and Maternity
Welfare District.
Warrell (Warrell et al. 1981) 1981 (1972-
1976), UK
Exposure was based on fluorescent-
antibody technique.
It is unknown
which types of
births were
included.
Non-malformed controls.
Matched (as far as possible) by
data of blood serum collection,
gestational age and parity.
Aro (Aro 1983) 1983 (1964-
1977), Finland
Finnish registry of congenital
malformations (FRCM). Exposure was
mainly based antenatal records,
collected prospective to delivery.
Live and
stillbirths were
included.
Non-malformed control.
Matched by date of delivery and
place.
Lynberg (Lynberg et al. 1994) (+
earlier paper by Park (Park et al.
1994 (1968-
1980), USA
Atlanta Birth Defects Case Control Study
(ABDCCS). Exposure was based on
Live and
stillbirths were
Non-malformed and malformed
controls (with CA “outside the
Page 110 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
40
1992)) maternal reported episodes of >2 days
of flu and fever from 1 month before to
3 months after conception, retrospective
to delivery.
included. hyperthermia spectrum”).
Matched by hospital of birth, race,
year and quarter of birth.
Adjusted for alcohol use,
education, maternal age,
multivitamin use and smoking.
Botto (Botto et al. 2001) (+ earlier
paper by Adams (Adams et al. 1989))
2001 (1968-
1980), USA
Atlanta Birth Defects Case Control Study
(ABDCCS). Exposure was based on
maternal reports, retrospective to
delivery.
Live and
stillbirths were
included.
Non-malformed controls.
Matched by hospital, race and time
of birth.
Adjusted for alcohol use, chronic
illness, education, race,
multivitamin use, period of birth
and smoking.
Li(Li et al. 2007) 2006 (2003-
2005), China
Exposure was based on maternal
reports, retrospective to delivery.
Live and
stillbirths were
Non-malformed controls.
Matched by county, estimated
Page 111 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
41
included. date of conception, ethnic group
and sex.
Adjusted for antibiotics,
antipyretics, diet, folic acid, history
of major congenital anomaly
affected pregnancy, maternal
education and infectious disease
other than influenza.
Czeizel (+ earlier papers by Acs,
Medveczky and Metneki) (Acs et al.
2005, Czeizel et al. 2008, Medveczky
et al. 2004, Metneki et al. 2005) (+
later paper by Csáky-Szunyogh
(Csaky-Szunyogh et al. 2013))
2007 (1980-
1996), Hungary
Hungarian Case-Control Surveillance of
Congenital Anomalies (HCCSCA).
Exposure was based on maternal
reports, retrospective to delivery.
Live and
stillbirths were
included.
Non-malformed controls.
Matched by district of residence,
sex, year and week of birth.
Oster (Oster et al. 2011) (+ later paper
by Kelly (Kelly et al. 2012))
2011 (1981-
1989), USA
Baltimore-Washington Infant Study
(BWIS). Exposure was based on maternal
Live births
were included.
Random sample of live-bortn
controls without congenital heart
Page 112 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
42
reports, retrospective to delivery. disease.
Frequency matched on age at
interview, month, year and
hospital of birth.
Adjusted for alcohol use, BMI,
family history of congenital heart
disease, gestational diabetes, race,
sex and smoking.
Results from these studies are included in the meta-analysis (Table I) and Table II. 582
Page 113 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
43
Table V: Overview of 12 papers from 10 cohort studies sorted by year. 583
1st
Author Publication
year (study
years) and
location
Recruitment Exposure ascertainment Matching
additional to
location and
adjustment (if
any)
Campbell
(Campbell
1953)
1953 (1950-
1951), UK
Mothers attending
antenatal clinics
were enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Abramowitz
(Abramowitz
1958)
1958 (1957),
South Africa
Women attending
gynaecology wards
were enrolled.
Exposure ascertainment
was not described.
None
Walker (Walker
and McKee
1959)
1959 (1957),
USA
Ward patients who
delivered at
University hospitals
were enrolled.
Exposure was based on
hemagglutination
inhibition and maternal
reports, retrospective to
delivery.
None
Doll (Doll et al.
1960)
1960 (1957-
1958), UK
Mothers attending
an antenatal clinic
for the first time
during the study
period were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery
and some of these were
confirmed with
professional diagnosis.
None
Pleydell
(Pleydell 1960)
1960 (1957),
UK
Expecting mothers
reported to suffer
from ILI were
Exposure was based on
maternal reports,
prospective to delivery.
None
Page 114 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
44
included and
exposed and non-
exposed were
recruited via local
midwives.
Saxen (Saxen et
al. 1960)
1960 (1958),
Finland
Women delivering in
the University
Central Hospital of
Helsinki were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Hardy (Hardy et
al. 1961)
1961 (1957-
1958), USA
Patients visiting an
obstetrical prenatal
clinic were enrolled.
Exposure was based on
complement fixation,
hemagglutinin inhibition
and maternal reports,
prospective to delivery.
None
Hirvensalo
(Hirvensalo and
Kinnunen 1962)
1962 (1957-
1959),
Finland
Women attending
the Maternity
Health Centres were
enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
None
Coffey (Coffey
and Jessop
1959, Coffey
and Jessop
1963)
1959 and
1963 follow-
up (1957-
1958), Ireland
Women attending
antenatal clinics
were enrolled.
Exposure was based on
maternal reports,
prospective to delivery.
Stage of
pregnancy.
Wilson (Wilson
et al. 1959,
1959 and
1969 follow-
Prenatal registrants
were enrolled.
Exposure was based on
hemagglutinin inhibition.
None
Page 115 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
45
Wilson and
Stein 1969)
up (1957),
USA
Results from these studies are included in the meta-analysis (Table I) and Table II. 584
Figure 10: Risk of bias assessment of included case-control studies. 585
[Figure 10] 586
Green represents low risk of bias, red represents high risk of bias, yellow represents unknown risk of 587
bias. 588
Figure 11: Risk of bias assessment of included cohort studies. 589
[Figure 11] 590
Green represents low risk of bias, red represents high risk of bias, yellow represents unknown risk of 591
bias. 592
Supplementary Forest Plots 593
Figure 12: Forest plot of aortic valve atresia/stenosis following 1st
trimester influenza exposure 594
[Figure 12] 595
Figure 13: Forest plot of atrial septal defect following 1st trimester influenza exposure 596
[Figure 13] 597
Figure 14: Forest plot of hypoplastic left heart following 1st
trimester influenza exposure 598
[Figure 14] 599
Figure 15: Forest plot of transposition of the great vessels following 1st trimester influenza exposure 600
[Figure 15] 601
Figure 126: Forest plot of ventricular septal defects following 1st
trimester influenza exposure 602
[Figure 126] 603
Figure 137: Forest plot of cleft lip ± cleft palate following 1st
trimester influenza exposure 604
[Figure 137] 605
Figure 148: Forest plot of cleft palate following 1st
trimester influenza exposure 606
Page 116 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
46
[Figure 148] 607
Figure 159: Forest plot of digestive system defects following 1st
trimester influenza exposure 608
[Figure 159] 609
Figure 1620: Forest plot of urinary defects following 1st trimester influenza exposure 610
[Figure 1620] 611
Figure 1721: Forest plot of hypospadias following 1st
trimester influenza exposure 612
[Figure 1721] 613
Figure 1822 Forest plot of limb reduction defects following 1st
trimester influenza exposure 614
[Figure 1822] 615
Figure 1923: Forest plot of club foot following 1st trimester influenza exposure 616
[Figure 1923] 617
Figure 2024: Forest plot of hip dislocation and/or dysplasia following 1st
trimester influenza exposure 618
[Figure 2025] 619
Figure 2125: Forest plot of polydactyly following 1st trimester influenza exposure 620
[Figure 2125] 621
Figure 2226: Forest plot of syndactyly following 1st
trimester influenza exposure 622
[Figure 2226] 623
Figure 2327: Forest plot of musculo-skeletal defects following 1st
trimester influenza exposure 624
[Figure 2327] 625
Supplementary Funnel Plots 626
Funnel plot asymmetry is indicative for publication bias as for example small studies reporting 627
negative findings might not be published. Visual inspection of funnel plots did not provide evidence 628
for publication bias. 629
Figure 248: Funnel plot of non-chromosomal CA following 1st
trimester influenza exposure 630
[Figure 2428] 631
Page 117 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
47
Figure 259: Funnel plot of neural tube defectsNTD following 1st
trimester influenza exposure 632
[Figure 2529] 633
Figure 2630: Funnel plot of anencephaly following 1st trimester influenza exposure 634
[Figure 2630] 635
Figure 2731: Funnel plot of encephalocele following 1st trimester influenza exposure 636
[Figure 2731] 637
Figure 2832: Funnel plot of spina bifida following 1st
trimester influenza exposure 638
[Figure 2832] 639
Figure 2933: Funnel plot of hydrocephaly following 1st
trimester influenza exposure 640
[Figure 2933] 641
Figure 3034: Funnel plot of congenital heart defects following 1st
trimester influenza exposure 642
[Figure 3034] 643
Figure 3235: Funnel plot of orofacial clefts following 1st
trimester influenza exposure 644
[Figure35] 645
Figure 36: Funnel plot of aortic valve atresia/stenosis following 1st
trimester influenza exposure 646
[Figure 36] 647
Figure 37: Funnel plot of atrial septal defect following 1st trimester influenza exposure 648
[Figure 37] 649
Figure 38: Funnel plot of hypoplastic left heart following 1st
trimester influenza exposure 650
[Figure 38] 651
Figure 39: Funnel plot of transposition of the great vessels following 1st
trimester influenza exposure 652
[Figure 39] 653
Figure 3140: Funnel plot of ventricular septal defects following 1st
trimester influenza exposure 654
[Figure 3140] 655
Figure 32: Funnel plot of orofacial clefts following 1st trimester influenza exposure 656
[Figure 32] 657
Page 118 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
48
Figure 3341: Funnel plot of cleft lip ± cleft palate following 1st
trimester influenza exposure 658
[Figure 3341] 659
Figure 3442: Funnel plot of cleft palate following 1st
trimester influenza exposure 660
[Figure 3442] 661
Figure 3543: Funnel plot of digestive system defects following 1st
trimester influenza exposure 662
[Figure 3543] 663
Figure 3644: Funnel plot of urinary defects following 1st
trimester influenza exposure 664
[Figure 3644] 665
Figure 3745: Funnel plot of hypospadias following 1st
trimester influenza exposure 666
[Figure 3745] 667
Figure 3846: Funnel plot of limb reduction defects following 1st
trimester influenza exposure 668
[Figure 3846] 669
Figure 3947: Funnel plot of club foot following 1st trimester influenza exposure 670
[Figure 3947] 671
Figure 4048: Funnel plot of hip dislocation and/or dysplasia defects following 1st
trimester influenza 672
exposure 673
[Figure 4048] 674
Figure 4149: Funnel plot of polydactyly following 1st
trimester influenza exposure 675
[Figure 4149] 676
Figure 4250: Funnel plot of syndactyly following 1st trimester influenza exposure 677
[Figure 4250] 678
Figure 4351: Funnel plot of musculo-skeletal defects following 1st
trimester influenza exposure 679
[Figure 4351] 680
Page 119 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
49
References 681
Abramowitz LJ. The effect of Asian influenza on pregnancy. S Afr Med J 1958:32:1155-1156. 682
Acs N, Banhidy F, Puho E and Czeizel AE. Maternal influenza during pregnancy and risk of congenital 683
abnormalities in offspring. Birth Defects Res A Clin Mol Teratol 2005:73:989-996. 684
Adams MM, Mulinare J and Dooley K. Risk factors for conotruncal cardiac defects in Atlanta. J Am 685
Coll Cardiol 1989:14:432-442. 686
Aro T. Maternal diseases, alcohol consumption and smoking during pregnancy associated with 687
reduction limb defects. Early Hum Dev 1983:9:49-57. 688
Botto LD, Lynberg MC and Erickson JD. Congenital heart defects, maternal febrile illness, and 689
multivitamin use: A population-based study. Epidemiology 2001:12:485-490. 690
Boyd PA, Devigan C, Khoshnood B, Loane M, Garne E, Dolk H and EUROCAT Working Group. Survey of 691
prenatal screening policies in Europe for structural malformations and chromosome anomalies, and 692
their impact on detection and termination rates for neural tube defects and Down's syndrome. BJOG 693
2008:115:689-696. 694
Buck C. Exposure to virus diseases in early pregnancy and congenital malformations. Can Med Assoc J 695
1955:72:744-746. 696
Busby A, Dolk H and Armstrong B. Eye anomalies: seasonal variation and maternal viral infections. 697
Epidemiology 2005:16:317-322. 698
Campbell WA. Influenza in early pregnancy; effects on the foetus. Lancet 1953:1:173-174. 699
Choi NW and Klaponski FA. On neural-tube defects: an epidemiological elicitation of etiological 700
factors. Neurology 1970:20:399-400. 701
Page 120 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
50
Cockroft DL and New DA. Abnormalities induced in cultured rat embryos by hyperthermia. 702
Teratology 1978:17:277-283. 703
Cockroft DL and New DA. Effects of hyperthermia on rat embryos in culture. Nature 1975:258:604-704
606. 705
Coffey VP and Jessop WJ. Maternal influenza and congenital deformities. A follow-up study. Lancet 706
1963:1:748-751. 707
Coffey VP and Jessop WJ. Maternal influenza and congenital deformities: a prospective study. Lancet 708
1959:2:935-938. 709
Coffey VP and Jessop WJ. Congenital abnormalities. Ir J Med Sci 1955:6:30-48. 710
Csaky-Szunyogh M, Vereczkey A, Kosa Z, Urban R and Czeizel AE. Association of maternal diseases 711
during pregnancy with the risk of single ventricular septal defects in the offspring--a population-712
based case-control study. J Matern Fetal Neonatal Med 2013:26:738-747. 713
Czeizel AE, Puho EH, Acs N and Banhidy F. Use of specified critical periods of different congenital 714
abnormalities instead of the first trimester concept. Birth Defects Res A Clin Mol Teratol 2008:82:139-715
146. 716
Davies HT, Crombie IK and Tavakoli M. When can odds ratios mislead?. BMJ 1998:316:989-991. 717
DerSimonian R and Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986:7:177-188. 718
Dodds L, McNeil SA, Fell DB, Allen VM, Coombs A, Scott J and MacDonald N. Impact of influenza 719
exposure on rates of hospital admissions and physician visits because of respiratory illness among 720
pregnant women. CMAJ 2007:176:463-468. 721
Page 121 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
51
Doll R, Hill AB and Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 722
1960:14:167-172. 723
Edwards MJ. Review: Hyperthermia and fever during pregnancy. Birth Defects Res A Clin Mol Teratol 724
2006:76:507-516. 725
Edwards MJ. The experimental production of Arthrogryposis multiplex congenita in guinea-pigs by 726
maternal hyperthermia during gestation. J Pathol 1971:104:221-229. 727
Edwards MJ. The experimental production of clubfoot in guinea-pigs by maternal hyperthermia 728
during gestation. J Pathol 1971:103:49-53. 729
Edwards MJ. Congenital defects in guinea pigs: fetal resorptions, abortions, and malformations 730
following induced hyperthermia during early gestation. Teratology 1969:2:313-328. 731
Edwards MJ. Congenital defects in guinea pigs: prenatal retardation of brain growth of guinea pigs 732
following hyperthermia during gestation. Teratology 1969:2:329-336. 733
Elizan TS, Ajero-Froehlich L, Fabiyi A, Ley A and Sever JL. Viral infection in pregnancy and congenital 734
CNS malformations in man. Arch Neurol 1969:20:115-119. 735
EUROCAT Central Registry. EUROCAT Prevalence Tables 2007-2011 2013:2013. 736
EUROCAT Central Registry. EUROCAT Guide 1.3 and reference documents. Instructions for the 737
Registration and Surveillance of Congenital Anomalities. Revised: May 2009 2009:. 738
Evans D. Hierarchy of evidence: a framework for ranking evidence evaluating healthcare 739
interventions. J Clin Nurs 2003:12:77-84. 740
Germain MA, Webster WS and Edwards MJ. Hyperthermia as a teratogen: parameters determining 741
hyperthermia-induced head defects in the rat. Teratology 1985:31:265-272. 742
Page 122 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
52
Granroth G. Defects of the central nervous system in Finland: III. Disease and drugs in pregnancy. 743
Early Hum Dev 1978:2:147-162. 744
Granroth G, Haapakoski J and Saxen L. Defects of the central nervous system in Finland: V. 745
Multivariate analysis of risk indicators. Int J Epidemiol 1978:7:301-308. 746
Gu J, Xie Z, Gao Z, Liu J, Korteweg C, Ye J, Lau LT, Lu J, Gao Z, Zhang B et al. H5N1 infection of the 747
respiratory tract and beyond: a molecular pathology study. Lancet 2007:370:1137-1145. 748
Hakosalo J and Saxen L. Influenza epidemic and congenital defects. Lancet 1971:2:1346-1347. 749
Hardy JM, Azarowicz EN, Mannini A, Medearis DN,Jr and Cooke RE. The effect of Asian influenza on 750
the outcome of pregnancy, Baltimore, 1957-1958. Am J Public Health Nations Health 1961:51:1182-751
1188. 752
Harris J. Influenza occurring in pregnant women: a statistical study of thirteen hundred and fifty 753
cases. JAMA 1919;72:978–80 1919:. 754
Hirvensalo M and Kinnunen O. Influenza and pregnancy. Duodecim 1962:78:740-747. 755
Hofler M. The Bradford Hill considerations on causality: a counterfactual perspective. Emerg Themes 756
Epidemiol 2005:2:11. 757
Jamieson DJ, Theiler RN and Rasmussen SA. Emerging infections and pregnancy. Emerg Infect Dis 758
2006:12:1638-1643. 759
Kallen B and Olausson PO. Vaccination against H1N1 influenza with Pandemrix((R)) during pregnancy 760
and delivery outcome: a Swedish register study. BJOG 2012:119:1583-1590. 761
Karkinen-Jaaskelainen and Saxen L. Maternal influenza, drug consumption, and congenital defects of 762
the central nervous system. Am J Obstet Gynecol 1974:118:815-818. 763
Page 123 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
53
Kelly S, Keuhl K and Leffredo C. Tricupsid Atresia: an epidemiologic investigation in the Baltimore-764
Washington area. American Journal of Epidemiology 2012:175:S59. 765
Klemetti A. Definition of congenital malformations and detection of associations with maternal 766
factors. Early Hum Dev 1977:1:117-123. 767
Laurence KM, Carter CO and David PA. Major central nervous system malformations in South Wales. 768
II. Pregnancy factors, seasonal variation, and social class effects. Br J Prev Soc Med 1968:22:212-222. 769
Leck I. Further tests of the hypothesis that influenza in pregnancy causes malformations. HSMHA 770
Health Rep 1971:86:265-269. 771
Leck I. Examination of the Incidence of Malformations for Evidence of Drug Teratogenesis. Br J Prev 772
Soc Med 1964:18:196-201. 773
Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963:17:70-80. 774
Leck I, Hay S, Witte JJ and Greene JC. Malformations recorded on birth certificates following A2 775
influenza epidemics. Public Health Rep 1969:84:971-979. 776
Li Z, Ren A, Liu J, Pei L, Zhang L, Guo Z and Li Z. Maternal flu or fever, medication use, and neural tube 777
defects: a population-based case-control study in Northern China. Birth Defects Res A Clin Mol 778
Teratol 2007:79:295-300. 779
Luteijn JM, Dolk H and Marnoch GJ. Differences in pandemic influenza vaccination policies for 780
pregnant women in Europe. BMC Public Health 2011:11:819. 781
Lynberg MC, Khoury MJ, Lu X and Cocian T. Maternal flu, fever, and the risk of neural tube defects: a 782
population-based case-control study. Am J Epidemiol 1994:140:244-255. 783
Page 124 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
54
MacKenzie JS and Houghton M. Influenza infections during pregnancy: association with congenital 784
malformations and with subsequent neoplasms in children, and potential hazards of live virus 785
vaccines. Bacteriol Rev 1974:38:356-370. 786
Mak TK, Mangtani P, Leese J, Watson JM and Pfeifer D. Influenza vaccination in pregnancy: current 787
evidence and selected national policies. Lancet Infect Dis 2008:8:44-52. 788
McGregor JA, Burns JC, Levin MJ, Burlington B and Meiklejohn G. Transplacental passage of influenza 789
A/Bangkok (H3N2) mimicking amniotic fluid infection syndrome. Am J Obstet Gynecol 1984:149:856-790
859. 791
Medveczky E, Puho E and Czeizel AE. An evaluation of maternal illnesses in the origin of neural-tube 792
defects. Arch Gynecol Obstet 2004:270:244-251. 793
Mereckiene J, Cotter S, D'Ancona F, Giambi C, Nicoll A, Levy-Bruhl D, Lopalco PL, Weber JT, Johansen 794
K, Dematte L et al. Differences in national influenza vaccination policies across the European Union, 795
Norway and Iceland 2008-2009. Euro Surveill 2010:15:19700. 796
Metneki J, Puho E and Czeizel AE. Maternal diseases and isolated orofacial clefts in Hungary. Birth 797
Defects Res A Clin Mol Teratol 2005:73:617-623. 798
Moher D, Liberati A, Tetzlaff J, Altman DG and PRISMA Group. Preferred reporting items for 799
systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009:339:b2535. 800
Monto AS, Gravenstein S, Elliott M, Colopy M and Schweinle J. Clinical signs and symptoms predicting 801
influenza infection. Arch Intern Med 2000:160:3243-3247. 802
Moretti ME, Bar-Oz B, Fried S and Koren G. Maternal hyperthermia and the risk for neural tube 803
defects in offspring: systematic review and meta-analysis. Epidemiology 2005:16:216-219. 804
Page 125 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
55
Mossey PA, Little J, Munger RG, Dixon MJ and Shaw WC. Cleft lip and palate. Lancet 2009:374:1773-805
1785. 806
Neuzil KM, Reed GW, Mitchel EF, Simonsen L and Griffin MR. Impact of influenza on acute 807
cardiopulmonary hospitalizations in pregnant women. Am J Epidemiol 1998:148:1094-1102. 808
Oster ME, Riehle-Colarusso T, Alverson CJ and Correa A. Associations between maternal fever and 809
influenza and congenital heart defects. J Pediatr 2011:158:990-995. 810
Park CH, Stewart W, Khoury MJ and Mulinare J. Is there etiologic heterogeneity between upper and 811
lower neural tube defects?. Am J Epidemiol 1992:136:1493-1501. 812
Pasternak B, Svanstrom H, Molgaard-Nielsen D, Krause TG, Emborg HD, Melbye M and Hviid A. Risk 813
of adverse fetal outcomes following administration of a pandemic influenza A(H1N1) vaccine during 814
pregnancy. JAMA 2012:308:165-174. 815
Pleydell MJ. Anencephaly and other congenital abnormalities. An epidemiological study in 816
Northamptonshire. Br Med J 1960:1:309-315. 817
Record RG. Anencephalus in Scotland. Br J Prev Soc Med 1961:15:93-105. 818
Rockenbauer M, Olsen J, Czeizel AE, Pedersen L, Sorensen HT and EuroMAP Group. Recall bias in a 819
case-control surveillance system on the use of medicine during pregnancy. Epidemiology 820
2001:12:461-466. 821
Saxen I. The association between maternal influenza, drug consumption and oral clefts. Acta Odontol 822
Scand 1975:33:259-267. 823
Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev Soc 824
Med 1975:29:103-110. 825
Page 126 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
56
Saxen I. Epidemiology of cleft lip and palate. An attempt to rule out chance correlations. Br J Prev Soc 826
Med 1975:29:103-110. 827
Saxen L, Hjelt L, Sjostedt JE, Hakosalo J and Hakosalo H. Asian influenza during pregnancy and 828
congenital malformations. Acta Pathol Microbiol Scand 1960:49:114-126. 829
Saxen L, Holmberg PC, Kurppa K, Kuosma E and Pyhala R. Influenza epidemics and anencephaly. Am J 830
Public Health 1990:80:473-475. 831
Saxen L, Klemetti A and Haro AS. A matched-pair register for studies of selected congenital defects. 832
Am J Epidemiol 1974:100:297-306. 833
Siston AM, Rasmussen SA, Honein MA, Fry AM, Seib K, Callaghan WM, Louie J, Doyle TJ, Crockett M, 834
Lynfield R et al. Pandemic 2009 influenza A(H1N1) virus illness among pregnant women in the United 835
States. JAMA 2010:303:1517-1525. 836
Smith MS, Upfold JB, Edwards MJ, Shiota K and Cawdell-Smith J. The induction of neural tube defects 837
by maternal hyperthermia: a comparison of the guinea-pig and human. Neuropathol Appl Neurobiol 838
1992:18:71-80. 839
Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA and 840
Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-841
analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000:283:2008-2012. 842
Sweeting MJ, Sutton AJ and Lambert PC. What to add to nothing? Use and avoidance of continuity 843
corrections in meta-analysis of sparse data. Stat Med 2004:23:1351-1375. 844
Walker WM and McKee AP. Asian influenza in pregnancy; relationship to fetal anomalies. Obstet 845
Gynecol 1959:13:394-398. 846
Page 127 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
57
Warrell MJ, Tobin JO and Wald NJ. Examination for influenza IgA and IgM antibodies in pregnancies 847
associated with fetal neural-tube defects. J Med Microbiol 1981:14:159-162. 848
Wells GA, Shea B and O'Connell D. The Newcastle-Ottawa Scale (NOS) for assessing the quality of 849
nonrandomised studies in meta-analyses 2004:. 850
Wilson MG, Heins HL, Imagawa DT and Adams JM. Teratogenic effects of Asian influenza. J Am Med 851
Assoc 1959:171:638-641. 852
Wilson MG and Stein AM. Teratogenic effects of asian influenza. A n extended study. JAMA 853
1969:210:336-337. 854
Zambon M, Hays J, Webster A, Newman R and Keene O. Diagnosis of influenza in the community: 855
relationship of clinical diagnosis to confirmed virological, serologic, or molecular detection of 856
influenza. Arch Intern Med 2001:161:2116-2122. 857
Page 128 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
Section/topic # Checklist item Reported
on page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 8 (of
133).
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria,
participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and
implications of key findings; systematic review registration number.
9-10
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 11
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons,
outcomes, and study design (PICOS).
12
METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide
registration information including registration number.
NA
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered,
language, publication status) used as criteria for eligibility, giving rationale.
12
Page 129 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify
additional studies) in the search and date last searched.
12
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be
repeated.
12
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable,
included in the meta-analysis).
12
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes
for obtaining and confirming data from investigators.
13
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and
simplifications made.
13
Risk of bias in individual
studies
12 Describe methods used for assessing risk of bias of individual studies (including specification of whether this was
done at the study or outcome level), and how this information is to be used in any data synthesis.
13-14
(full
summary
100-101)
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 14
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency
(e.g., I2) for each meta-analysis.
14
Page 1 of 2
Page 130 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
Section/topic # Checklist item Reported
on page #
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective
reporting within studies).
Publication
Bias page
14.
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done,
indicating which were pre-specified.
Subgroup
analysis
page 14.
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions
at each stage, ideally with a flow diagram.
15 (+Flow
Chart page
41).
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period)
and provide the citations.
Ecological
Studies 56-
60.
Case-
Control
Page 131 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
studies 61-
66.
Cohort
studies 67-
69.
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). 15-17, 70-
71.
Results of individual studies 20 For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each
intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.
37-40 for
studies not
enrolled in
meta-
analysis.
42-49 (plus
figures 12-
27 of
eAppendix)
for studies
Page 132 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
enrolled in
meta-
analysis.
Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency. 34-36
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). 34
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]). 34
DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to
key groups (e.g., healthcare providers, users, and policy makers).
15-17.
Relevance
key groups
26-27.
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of
identified research, reporting bias).
18-19.
(limitations
and
statistical
limitations
discussed
Page 133 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF
PRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 ChecklistPRISMA 2009 Checklist
in separate
paragraphs)
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research. 19-20
FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for
the systematic review.
20
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097
For more information, visit: www.prisma-statement.org.
Page 2 of 2
Page 134 of 134
http://humrep.oupjournals.org
Draft Manuscript Submitted to Human Reproduction for Peer Review
Ref: 20102204 D07-06 OTH UK PS.PDF