The prevalence and risk factors of fear of childbirth among pregnant women: A cross‐sectional study in Ireland

There is growing evidence of the considerable impact of fear of childbirth on women's health and well‐being, but prevalence reports of high and severe fear of childbirth and reported risk factors have been inconsistent in various studies. Therefore, this study aimed to determine the prevalence of high and severe fear of childbirth, and to identify risk factors of childbirth fear.


| INTRODUC TI ON
Fear of childbirth (FOC) exists on a continuum from normal worries and fears, to severe fear, (tocophobia). [1][2][3] Although tocophobia is not clearly defined, the adverse impact of FOC on women's health and well-being in the perinatal period is well-established. 1,4,5 Previous studies have reported that sleep disturbances, nightmares, palpitations, stomach pains, panic attacks, flashbacks (after trauma) and a request for a cesarean are associated with FOC. 4,[6][7][8] Furthermore, the impact on emotional well-being may be long-term and powerful, affecting partner relationships 9 and breastfeeding. 10 Prevalence estimates from single country 11 and multi-country 7 studies differ (3.7%-43%), 2 due to poor consensus on definition and to various methods of measuring FOC. 2,12,13 A meta-analysis estimated the global pooled-prevalence at 14%, 2 noting increased prevalence in recent years, which may be attributable to increased awareness and reporting. 2 Notably, no Irish study on the prevalence of FOC was retrieved in the systematic literature search. 2 The metaanalysis found inconsistent evidence in relation to parity and FOC, with the majority of studies reporting higher prevalence in firsttime mothers 2 but with some studies reporting higher prevalence in parous women. 2,13,14 Previous research suggests an association between low perceived social support and FOC. 6,7 The Wijma Delivery Expectancy Questionnaire version A (W-DEQ A) 3 is the most commonly used tool to measure FOC severity 2 and is validated in many countries and languages. 2 The prevalence of severe FOC (defined as W-DEQ A ≥85) is reported to be between 5% and 21% 2 and high FOC (W-DEQ A ≥66) between 24% and 26%. 2,11,15 Researchers 7,12 suggest that the W-DEQ A consists of four subscales, 12 which may facilitate healthcare professionals' assessment of the nature of FOC, in addition to assessing severity, thereby facilitating a more personal approach to support offered to women. 12 There is limited evidence in relation to these subscales at present, thus assessing the subscales in various cultural settings was recommended. 12 Due to this knowledge gap, the primary objective of this study was to establish the prevalence of high and severe FOC in a sample of pregnant women in Ireland. Secondary objectives were to identify potential risk factors of high FOC and to elucidate the nature of FOC by applying W-DEQ A subscales in this study.

| Study design and population
We conducted a cross-sectional study between April 2015 and June 2016 in Cork, Ireland. A convenience sample of pregnant women attending routine antenatal appointments was recruited from public and private clinics. Recruitment took place over time periods when the researchers were available to recruit, rather than a consecutive period of time, since the study was carried out as part of part-time doctoral studies by the research midwife. The midwife trained the undergraduate students to recruit women to the study and either the midwife or research students invited pregnant women to participate.
All the participants were planning to give birth at Cork University Maternity Hospital. Participation in this study was short, requiring the completion of just one questionnaire. The follow up to the study for pregnancy outcomes was done using access to medical charts and the outcome data will be presented in a separate publication.
In Ireland, universal maternity benefits are available to all women, which means that free care is available during pregnancy and up to 6 weeks postpartum for those ordinarily resident. 16 The predominant model of care is obstetric-led, with combined care involving the woman's GP being provided under the HSE Maternity & Infant Care Scheme. 16 Women who choose shared care are seen by several different midwives and hospital doctors during their pregnancy, normal births would be facilitated by midwives and operative births by an obstetrician. Domiciliary Care In and Out of Hospital (DOMINO) is an option available in certain counties for women considered "low risk" and within a certain local radius of the hospital, allowing continuity of midwifery care and early discharge home. Private antenatal care led by one obstetric consultant is available for a fee. Private maternity care is available at all 19 maternity units in Ireland and there is also one fully private unit.
The study population included; women >18 years, able to complete the questionnaire in English, and between 12 and 24 weeks' gestation at the time of recruitment. Previous studies suggest that FOC is not a stable construct and FOC levels may increase in the third trimester. 11

| Variables
We developed a questionnaire package based on the literature to meet the aims of the study. Demographic information collected included: age (by category), marital status, country of birth, education, smoking, weight, height and employment status. Women were asked to rate their general health from 1 to 5 (poor to very good). Obstetric Isolation"-containing questions relating to perceived social support (comprising four items: 3, 7, 11, 15); and (4) "Moment of Birth"-containing questions relating to how the woman imagines she will feel during birth (comprising three items: 28,29,30). 12 Using a cut-off ≥2.5 (the midpoint) was recommended for comparison purposes. 12 The EPDS is a widely used and well-validated self-report screening tool for recognizing women at risk of perinatal depression. 17,18 Negative questions are reverse-scored and a total score calculated, with scores of 0-30 possible. A systematic review of studies validating the use of EPDS in antenatal and postpartum women, recommended using a cut-off of 9 or 10 for very likely risk of depression. 18 Therefore, a cut-off ≥10 was used in this study. 18 The Perinatal Infant Care Social Support Scale (PICSS) 19 was used to measure maternal social support by investigating functional social support using four domains-informational, instrumental, emotional and appraisal support, 20 and structural social support or people available in a person's social networks (formal and informal). 19,20 An individual score was calculated for each domain. For informational and instrumental support domains, low support was defined as a score ≤20. 19 For emotional and appraisal support domains, low support was defined as a score <12. Structural social support was measured by asking which individuals from the participant's social network (ie, formal, such as health professionals, and informal, such as family/friends) would be available to provide the four types of functional support.
Formal or informal structural support was considered available if any type of support was available from at least one source. 19 The questionnaire was piloted for ease of use with the first 100 women; the font size was increased on the PICSS, 19 as it was deemed unclear in the initial format.

| Statistical analyses
IBM SPSS Version 22.0 statistical software program (IBM Corp., Chicago, IL, USA) was used for all statistical analyses. When determining sample size, the literature was examined and a sample of 1000 women deemed adequate on the basis of findings of previously published prevalence studies. 11 Descriptive statistics were calculated for all variables and presented as n (%) or mean with standard deviation as appropriate. Following this, scores were calculated for each standardized measure (EPDS and W-DEQ A).
FOC prevalence was estimated using the whole study population and subsequently in subgroups according to a priori chosen variables: parity, marital status and history of pregnancy loss (history of miscarriage or stillbirth). Student's t-tests were used for continuous variables and chi-square tests for categorical variables as appropriate. P < 0.05 was considered statistically significant. The W-DEQ A was treated as a categorical variable (0-37, 38-65, 66-165) for this analysis because the study was not adequately powered for the risk factor analysis of the W-DEQ A ≥85 category. When there were ≤6 missing items, each item was replaced by the series mean for each participant 7 and total score calculated (n = 44). Participants with >7 W-DEQ A items were excluded. Internal consistency in each scale used was determined using Cronbach's co-efficient alpha 21  Relative risk ratio (RRR) and 95% confidence intervals (CI) were calculated to examine the association between each selected variable and risk of high FOC, using univariate multinomial logistic regression analysis, followed by multivariate multinomial logistic regression.
Variables with a P value < 0.15 in the univariate models were included in the multivariate models (maternal age, marital status, employment, smoking, body mass index [BMI], living with partner, EPDS history of anxiety with treatment, history of depression, history of postnatal depression, low formal and informal social support).

| Ethical approval
This study obtained full ethical approval from the Cork Research Ethics Committee for the Teaching and Learning Hospitals ECM 4 (06/01/15) and ECM 3 (03/03/15). The study was explained using the information leaflet (explaining the voluntary nature of the study), eligibility clarified and written consent (separate from the questionnaire) obtained.
Participants returned completed questionnaires to a sealed box.

| RE SULTS
A total of 1180 women were invited to participate; 1056 women consented and were given questionnaires. Of these, 1001 pregnant women self-completed and returned questionnaires (Figure 1). Data on demographics for 69 women who refused to participate are not available but reasons for declining are outlined in Figure 1. Fifty-five women were ineligible. A further 21 (2%), returned ineligible questionnaires ( Figure 1).
Although there was a high overall response rate (n = 980 [85%]), outcome data were available for 882 (75%) of these women. Participants had a median gestational age of 20 weeks and interquartile range 15-21.

| Demographic and maternal characteristics
Demographic characteristics (n = 882) are summarized in Table 1. The majority of women were aged 31-35 years (44%, n = 388) and married (62.2%, n = 549). Most women were Irish (78%, n = 688), employed (64.5%, n = 569) and had a university education (39.9%, n = 352). Sixtyeight women (7.7%) were self-reported smokers. Of the total sample, 298 (33.8%) were nulliparous and 581 (65.9%) women were multiparous and three women did not report parity. Three women who reported the current pregnancy as their first, stated they had at least one child at home; this may be a partner's child or adopted child (Table 1). At least one pregnancy loss was reported by 174 (19.7%) women (Table 1).  (Table 2), but there was no statistically significant difference when compared (P < 0.07). The prevalence of high FOC (W-DEQ A ≥66) was 43% (n = 128) in nulliparous women, and 33.6% (n = 195) in multiparous women ( Table 2); when compared, the difference was statistically significant (P < 0.005). The prevalence of severe FOC among women who reported at least one pregnancy loss was 4.3%, and 5.2% among women who reported no pregnancy loss; this difference was not statistically significant (P = 0.34) and was not significant for high fear (P = 0.38). The minimum W-DEQ A score reported was 1 and the maximum 128.

| Risk factors of fear of childbirth
The analysis of demographic factors revealed that high FOC was significantly more common among women who identified themselves as single (P < 0.008) when compared with married or cohabiting women, but there was no difference at the severe level of fear (P = 0.13). Adjusted results from the multivariate analysis are presented in Table 3. In terms of psychological factors, a history of depression or current depressive symptoms per the EPDS were identified as statistically significant factors associated with F I G U R E 1 Flow chart of study recruitment process TA B L E 1 Demographic and maternal characteristics of participants   In this study, high FOC was associated with first-time mothers (P < 0.005). These findings are in line with similar findings in at least nine studies, 2 but one study 14 found that FOC was more common in multiparous women, and two studies showed no association between FOC and parity. 23,24 In the present study, we found no significant difference in prevalence in women with and without previous pregnancy loss, which was associated with FOC in a large epidemiological study. 17

| D ISCUSS I ON
Our finding that women who identified themselves as single were more likely to have high FOC is aligned with findings from previous research. 17  When considering generalizability and external validity of the study, the convenience sample which was recruited from a single site and not selected randomly must be taken into account. This may limit the degree to which results are generalizable to the Irish pregnant population.
However, this is one of the largest maternity hospitals in Europe with approximately 8000 births annually and our sample would appear to be comparable with national averages 32 (Table 5), apart from the figure for smoking, which was lower than the most recent national figure of smoking during pregnancy; 33 , 39.9% among the study participants compared with 33.5% at the national level. The small difference in higher education rate could be due to the large university based in Cork or because the only available figure is for all females rather than mothers. We cannot rule out, however, that this difference is due to other reasons. With these limitations in mind, findings from this study may still be considered useful.
An important limitation in this study is that multiparous women were not asked about previous birth mode, since women who report a previous negative birth or FOC in one pregnancy are more likely to report FOC in a subsequent pregnancy. 7 Another weakness in the study is that it was primarily designed to estimate the prevalence of FOC but was not powered for the risk factor analysis, which may have led to several associations with moderate to large RRR that were not statistically significant, which could be due to small numbers within categories. The women completed questionnaires only in the second trimester, this is acknowledged as a study limitation. There were missing BMI data for 124 (14.0%) women. BMI was self-reported, with 72 (8.1%) women having missing weight and the rest having missing height and weight.
Forty-one (4.6%) women did not complete the EPDS questionnaire. The mean W-DEQ score for women who completed the EPDS was 57.08 and the mean W-DEQ score of women who did not complete the EPDS was 62.66. The data were incomplete for the whole EPDS questionnaire in the 41 participants who did not complete, rather than missing certain questions.

| CON CLUS ION
This study found a similar prevalence of severe FOC but a higher prevalence of high FOC when compared with reported international prevalence. High FOC was associated with depression, being a firsttime mother and low perceived informational support; assessing social support, antenatal education provision and high quality information are therefore essential in pregnancy. This study adds to our limited understanding of FOC by using subscales to explore the nature of, as well as the severity of FOC. More investigation of other possible risk factors is recommended for future research.

ACK N OWLED G M ENTS
We are grateful to the following students from the Department of This figure is for all females, not mothers. c As reported in the Growing Up in Ireland Survey. TA B L E 5 A comparison of national statistics with study results