Paternal bonding is influenced by prenatal paternal depression and trait-anxiety

ABSTRACT

Background

Even though the development of an emotional bond to the child involves both parents, studies on the development of paternal bonding and the influencing factors are scarce. This pilot study examines the quality of paternal postnatal bonding in association with paternal depressive and anxiety symptoms before and after birth. Methods: Expecting parents (n = 81) were recruited from maternity services in Frankfurt, Germany. At recruitment and 3 months postpartum (pp) mothers and fathers completed an interview including sociodemographic and pregnancy data. Depressive and anxiety symptoms were screened using the Edinburgh Postnatal Depression Scale and the State-Trait Anxiety Inventory. At 3-month pp, fathers also completed the Postpartum Bonding Questionnaire for the assessment of bonding difficulties. A total of 63 couples, from whom data were available for both time points, were included in the final study group.

Results

Depressive and anxiety symptoms before birth are the best predictors for the quality of paternal bonding pp (Total score R² .402 p = .001; Impaired bonding R² .299 p = .019; Rejection and Anger R² .353 p = .005; Anxiety about care R² .457 p = .000). Maternal depression and sociodemographic variables were not significantly associated.

Limitations

High selected small study group.

Conclusions

Paternal depressive and anxiety symptoms during pregnancy are highly predictive for the quality of bonding as well as for the presence of depressive and anxiety symptoms 3 month pp. It is necessary to identify these symptoms as soon as possible in order to prevent later negative impacts on parental mental health and on child developmental outcomes.

KEYWORDS:

• Paternal Bonding
• Depressive symptoms
• Anxiety
• Pregnancy
• Postpartum

1. Introduction

Pregnancy and time after birth are physically and psychologically very intense and challenging periods. The redefinition of the identity, the transformation of the couple’s relationship, the relationship to the family of origin as well as the reorganisation of the daily routines are some of the aspects, that need to be deeply readjusted in this phase. Nevertheless, the development of ‘a relationship with the newborn is the central and most important psychological process of the puerperium’ (Brockington et al., 2006, p. 243). Very soon after delivery, mothers and fathers experience instinctively feelings of growing affection, tenderness and a sense to care and protect the baby. Most parents develop this positive affective state, this emotional bond to the newborn, naturally. In some cases, however, the expected reactions of affection, love and affiliation towards the new-born fail and parents feel distant, indifferent or detached towards the infant. This manifestation of bonding failure, a bonding disorder, also includes increased irritability, aggressive impulses or rejection of the infant. This disorder can lead to a long-term impairment of the parent--child relationship comprising child maltreatment (Brockington et al., 2006; Hornstein & Trautmann-Villalba, 2007; Hornstein et al., 2009; Kumar, 1997) as well as to disturbances regarding the child’s development (de Cock et al., 2016, 2017; Enns et al., 2002; Mason et al., 2011). Longitudinal studies have shown, that the parental bond to the child emerges during pregnancy and is strongly related to the postnatal bond (de Cock et al., 2016; Dubber et al., 2015; Maas et al., 2016; Rossen et al., 2016). This was confirmed by studies that indicate a high stability of bonding across pregnancy, the first few weeks and some months postpartum (de Cock et al., 2017; Moehler et al., 2006).

Even though the development of an emotional bond to the child involves both parents and disorders of this bond could affect mothers and fathers, most of the previous studies only referred to mothers. Adverse maternal mental health aspects, e.g. depressive symptoms (Dubber et al., 2015; Lehnig et al., 2019; Muzik et al., 2017; Tolja et al., 2020), anxiety (Dubber et al., 2015; Tietz et al., 2014) maternal childhood maltreatment (Lehnig et al., 2019; Muzik et al., 2013) or adult attachment style (Nonnenmacher et al., 2016) have been shown to affect the quality of bonding. Other determinants of mother-child-bonding have been described, for example social and partner support as a positive predictor (Kinsey et al., 2014; Wells & Jeon, 2023), unplanned pregnancies and previous pregnancy loss as a negative predictors (Camarneiro & de Miranda Justo, 2017; Laxton-Kane & Slade, 2002). Additionally, there are hints that adverse childhood experience of the mothers might impact the bonding to the child (Cooke et al., 2019; Frohberg et al., 2022). Literature on the development of paternal bonding, the factors that influence this development and the effect paternal bonding disorders have on children’s development is scarce.

In pioneering work Edhborg et al. (2005) found that, compared to mothers, fathers showed more rejection and anger at one week after birth. However, 2 months after birth, fathers scored significantly higher on the PBQ and on all its subscales. In addition, an impaired bond at one-week postpartum and depressive symptoms in the Edinburgh Postnatal Depression Scale (EPDS) negatively influenced the bonding’s quality of the father 2 month after birth. The association between depression and bonding quality was also confirmed by Kerstis et al. (2016). Paternal impaired bonding at 6 months pp was associated with paternal EPDS scores at 6 weeks and 6 months postpartum. Likewise, paternal impaired bonding was related to maternal depressive symptoms at 6 weeks and 6 months. Since several studies identified maternal depressive symptoms in pregnancy and postpartum as a risk factor for paternal depressive symptoms (for meta-analysis and systematic reviews. Ansari et al., 2021; Thiel et al., 2020; Wang et al., 2021), maternal depression appears to be of particular interest, as a factor that could mediate the influence of paternal depression on the bonding quality.

Finally, prenatal parental bonding levels are positively related to successive bonding to the child after birth. Maternal and paternal bonding feelings were also correlated over time (prenatal to 24 month) (de Cock et al., 2017). To our knowledge, the influence of paternal depressive and anxiety symptoms before birth on the quality of the postpartum paternal bond to his baby, has not yet been studied.

The aim of the following pilot study was to examine the quality of paternal bonding after birth and to determine the influence of father’s depressive and anxiety symptoms prior to and after birth on this bond beyond sociodemographic factors.

Fathers and men in general are not easily recruited into studies, especially if a more extensive assessment of sociobiographic data as well as mental health history and mental disorder symptoms are included. Additionally, there is still scarce knowledge about paternal bonding in pregnancy and early postpartum period and the potential influencing factors. Therefore, we conducted this pilot study to investigate the feasibility of recruiting fathers for this kind of studies and to test first hypotheses on risk factors of impaired bonding. With our results we then will be able to plan an extensively phenotyped and well-powered longitudinal pregnancy and birth cohort study.

2. Methods

2.1 Participants and procedure

This study is part of the Vater-Bindung-Studie (Father-Bonding-Study), a pilot study conducted in Frankfurt, Germany between 2017 and 2019. Expectant parents were recruited, predominately at information sessions in the Department of Obstetrics, Bürgerhospital in Frankfurt or in midwifery practices. The information sessions in the Bürgerhospital were held twice a month, with about 50 expecting couples joining each session. In about 24 months, we therefore approached about 2400 couples in total with information about the study, which means that about 3–4% of the couples agreed to participate. Inclusion criteria were: the women being pregnant ≥ 20 gestational weeks, both parents being over the age of 18, sufficient German knowledge and both parents being willing to participate. Only participants that were able to give written informed consent were included in the study that adhered to the Declaration of Helsinki (64th version, 2013). Exclusion criteria were not being able to give informed consent, not being fluent in German, only one partner being willing to participate. The ethics committee of the University of Frankfurt as well as the Hessische Landesärztekammer approved the study (votum no 135/17). At enrolment (baseline), pregnant women and their partners completed an interview including sociodemographic and pregnancy characteristics, alcohol, tobacco and illicit drug consumption as well as physical and mental health (e.g. ICD-diagnosis, suicide attempts). Families were visited at home 3 month after birth for a second interview. At baseline and at follow up, mothers and fathers also completed some self-report questionnaires. A total of 63 couples, from whom data were available for both time points, were included in the study group of this analysis of the fathers (for more information see Kittel-Schneider et al., 2022). We screened previous mental disorder diagnosis in the participants using the M.I.N.I. (Mini International Neuropsychiatric, Interview, German Version 5.0.0.). Only 10% of the fathers and 11% of the mothers of the whole sample had an own history of migration or came from families that migrated to Germany. Sample characteristics are presented in Table 1.

Table 1. Participants’ sociodemographic information.

Age in years (mean, range, sd)	36.5 (28–46, 4.35)
Working hours per week (mean h, range, sd)	40,48 (0–55, 9.70)
Education No school degree 9 years schooling 10 years schooling 12/13 years schooling Job specialization University degree Missing	N (%) 0 0 1 (1.6) 2 (3.2) 5 (7.9) 55 (87.3) 0
History of mental illness No Yes (Major depression) Missing	N (%) 56 (88.9) 2 (3.2) 5 (7.9)
Gender of infants Male Female Missing	N (%) 28 (44.4) 31 (49.2) 4 (6.4)
Married N (%)	43 (68.3)
Child is Firstborn N (%)	53 (84.1)

Total sample N = 63 fathers.

2.2 Measures

2.2.1. Edinburg Postnatal Depression Scale (EPDS)

Depressive symptoms were assessed with the German version of the Edinburg postnatal depression scale (Bergant et al., 1998). This well validated 10-item screening instrument has been used in numerous studies for the detection of maternal and paternal depressive symptoms in pregnancy and after birth (Cronbach’s α .81). Each self-rated item scores from 0 (no symptom) to 3 (severe symptom) given a summed maximum score of 30. Some items are reverse scored. A higher score indicates more depressive symptoms. Even though it is possible for screening purposes to set different cut-offs to indicate clinically relevant depressive symptomatology, in this analysis we only used the EPDS score as a continuous variable.

2.2.2. State-Trait Anxiety Inventory (STAI)

Anxiety was assessed with the German version of the State-Trait Anxiety Inventory (Laux et al., 1970). For this study we only assessed trait-anxiety. The STAI-T comprises 20 items with answer options from 0 (almost never/not at all) to 4 (almost always/very much), the sum score ranges between 20 and 80 (Cronbach’s α .88 to .94). A higher score indicates greater trait anxiety.

2.2.3. Postpartum Bonding Questionnaire (PBQ)

The PBQ was originally developed to assess the emotional bond from a mother to her child (Brockington et al., 2001, 2006; German Version: Reck et al., 2006) meanwhile the PBQ has been also frequently used for fathers. The PBQ is a self-report screening instrument and consists of 25 positive and negative statements rated on a scale of 0 to 5. Positive statements are scored as marked, negative statements reversed. The sum of all values constitutes a general score and varies from 0 to 125 (Cronbach’s α .85) In addition, four subscales could be built: Impaired bonding (12 items, range 0–60, Cronbach’s α .78), rejection and anger (7 items, range 0–35, Cronbach’s α .68), anxiety about care (4 items, range 0–20, Cronbach’s α .34) and risk for abuse (2 items, range 0–10, Cronbach’s α .20). For all scales, a higher score indicates more bonding difficulties. Brockington et al. (2001) proposes cut-off points for each scale to identify problematic bonding. For this study, we only used the continuous scores. Only two fathers scored > 1 in the scale risk for abuse, thus this scale was not included in the statistical analysis.

2.3. Statistical analysis

For data management and statistical analysis, SPSS (Version 26, IBM, Armonk, NY, U.S.A) was used. Means and standard deviations for the PBQ, EPDS and STAI scores were calculated and normal distribution was tested. Since not all variables were normally distributed, non-parametric analysis were performed. Spearman-Rho coefficients were used to examine the association between the independent variables EPDS and STAI.

The impact of paternal depression and anxiety at pregnancy and 3 months after birth was examined using a stepwise hierarchical regression analysis (forward) with continuous EPDS and STAI-T scores as predictors and PBQ scores as outcome variables. In a first step, sociodemographic control variables were entered (paternal age, gender of infant, first child, education, working hours per week). To consider a potential effect of maternal depressive symptoms, the EPDS scores at pregnancy and 3 months postpartum were entered in a second (forced) step and in the third step paternal pre- and postpartum EPDS and STAI-T scores followed. The distribution of residuals has been verified. The level of significance was for all tests two-sided and set at p = 0.05.

In a post-hoc power analysis of the linear multiple regression analysis on the PBQ scores with maternal and paternal depressive symptoms and paternal anxiety at pregnancy and postpartum, we could achieve a marginally sufficient power with 0.78 (G*Power 3.1.9.4).

3. Results

3.1 Bonding, depression and anxiety

Descriptive statistics for PBQ, EPDS and STAI-T are presented in Table 2. With regard to paternal depressive and trait anxiety symptoms, in general, fathers rated both as very low during pregnancy as well as after birth (Table 2). As expected, symptoms of depression and trait anxiety were strongly correlated. In all PBQ scales, the self-rated scores 3 month after birth were also very low. PBQ scales correlated stronger with EPDS and STAI-T scores after birth than at pregnancy. Maternal and paternal scores were strongly correlated for all PBQ values. Surprisingly, no association with EPDS values could be found with either the mothers or fathers at pregnancy or after birth as reported before (Kittel-Schneider et al., 2022). Only paternal EPDS at 3-month pp correlated with impaired bonding (Table 3).

Table 2. EPDS, STAI-T at pregnancy and 3 month pp and PBQ at 3 month pp – Descriptive statistics and Correlations (only Fathers).

				Variables
	M	SD	Min.-Max.	1.	2.	3.	4.	5.	6.	7.	8.
1. EPDS pregnancy	3.21	4.17	0–23	-
2. EPDS 3 month pp	2.95	3.02	0–13	.485***	-
3. STAI-T pregnancy	29.64	8.31	20–54	.656***	.579***	-
4. STAI-T 3 month pp	29.43	9.13	20–58	.457***	.554***	.706***	-
5. PBQ Total score	10.89	8.20	0–36	.233	.441**	.312*	.493***	-
6. PBQ Impaired bonding	6.08	4.83	0–18	.207	.412**	.300*	.452***	.961***	-
7. PBQ Rejection and anger	2.19	2.51	0–10	.216	.275*	.296*	.356**	.819***	.695***	-
8. PBQ Anxiety about care	2.62	1.81	0–9	.328*	459***	.313*	.452***	.671***	.538***	.515***	-

***p < 001; **p < .01; *p < .05.

Table 3. EPDS (Edinburg Postnatal Depression Scale) at pregnancy and 3 month pp (postpartum) and PBQ (Postpartum Bonding Questionnaire) at 3 month pp – Intercorrelations mothers and Fathers and descriptives statistics (only for mothers)¹.

	Mothers
Fathers	1.	2.	3.	4.	5.	6.
1. PBQ Total score	.512***	.491***	.467***	.259*	.078	.157
2. PBQ Impaired bonding	.539***	.531***	.476***	.283*	.047	.202
3. PBQ Rejection and anger	.318*	.303*	.352**	.059	.102	.038
4. PBQ Anxiety about care	.399**	.304*	.364**	.283*	.108	.106
5. EPDS pregnancy	−.031	−.004	−.003	.056	.221	−.008
6. EPDS 3 month pp	.232	.259*	.180	.178	.042	.085
M (S)^1	11.55 (8.25)	6.10 (4.44)	2.83 (2.99)	2.62 (1.93)	4.24 (3.36)	5.45 (4.52)
Min.-Max^1	0–36	0–19	0–11	0–8	0–14	0–20

^aThis information for Father variables has been shown in Table 2.

***p < .001; **p < .01; *p < .05.

3.2. Influence of depressive symptoms and anxiety on paternal bonding

The results of the stepwise regression analysis on the PBQ scores with maternal and paternal depressive symptoms and paternal anxiety at pregnancy and postpartum are shown in Table 4. The variables paternal age, gender of infant, first child, education, working hours per week – included for control in the analysis – were not significant for all the scores, therefore, they were not included in any of the further regression models. Likewise, the forced inclusion of maternal depressiveness (EPDS score before and after birth) was not significant in any of the models. Paternal depressive symptoms before birth were the best predictor for the PBQ total score and for all subscales, while the same symptoms after birth were only significant regarding the subscale anxiety about care. Furthermore, the effect of paternal trait-anxiety during pregnancy was significant for all other but this subscale. For all other scales, this effect was significant and surprisingly the opposite as expected: the higher the anxiety symptoms, the less paternal bonding difficulties after birth. In addition, trait anxiety 3-month postpartum was only significant for the subscale rejection and anger, but here in the contrary direction as with anxiety before birth.

Table 4. Influence of depressiveness and anxiety on father’s bonding - Stepwise (foward) regression analysis.

	PBQ Total Score*		PBQ Impaired bonding*		PBQ Rejection and anger*		PBQ Anxiety about care*
Model 1	ß	T (p)	ß	T (p)	ß	T (p)	ß	T (p)
EPDS Mother pregnancy	.148	.896 (.375)	.049	.296 (.769)	.218	1.330 (.190)	.234	1.435 (.158)
EPDS Mother 3 mo. pp	−.050	−.300 (.766)	.072	.433 (.667)	−.155	−.944 (.350)	−.199	−1.223 (.228)
	R^2 .018; F = .408; p .667		R^2 .011; F = .246; p .783		R^2 .041; F = .958; p .391		R^2 .052; F = 1.236; p .300
Model 2
EPDS Mother pregnancy	.019	.133 (.895)	−.070	−.458 (.650)	.082	.560 (.579)	.156	1.158 (254)
EPDS Mother 3 mo. pp	−.092	−.654 (.517)	.043	.286 (.777)	−.193	−1.323 (.193)	−.259	−1.937 (.060)
STAI-T Father pregnancy	−.666	−2.438 (.019)	−.613	−2.073 (.045)	−.613	−2.156 (.037)	−.514	−1.974 (.055)
STAI-T Father 3 mo. pp	.490	1.794 (.080)	.416	1.407 (.167)	.628	2.209 (.033)	.221	.851 (.400)
EPDS Father pregnancy	.502	2.952 (.005)	.480	2.607 (.013)	.400	2.261 (.029)	.427	2.636 (.012)
EPDS Father 3 mo. pp	.331	1.867 (.069)	.268	1396 (.170)	.160	.871 (.389)	.556	3.296 (.002)
	R2 .402; F = 4.597; p .001		R2 .299; F = 2.910; p .019		R2 .353; F = .3.726p .005		R2 .457; F = 5.743; p .000

excluded variables: paternal age, gender of infant, first child, education, working hours per week.

*3 mo. pp.

Discussion

The purpose of this pilot study was to examine the quality of the paternal bond after birth, the influence of maternal and paternal depressive symptoms as well as paternal trait-anxiety during pregnancy and 3 month after the birth controlled for some demographic factors. The aim was also to test the feasibility of recruiting new fathers for a study that required a rather deep phenotypic characterisation of mental health symptoms and bonding to the child to prepare for a larger pregnancy and birth cohort study.

Regarding the paternal scores on all PBQ scales, it is evident that these were overall low, indicating no severely disturbed emotional relationships between fathers and their babies in our sample. While other studies also found low values for paternal bonding disorders, the scores registered in our study are slightly higher. These studies, however, measured the bonding quality at other time points (Bieleninik et al., 2021; Edhborg et al., 2005; Kerstis et al., 2016).

Paternal age and educational level, first paternity, and depressive symptoms of mothers during pregnancy and after birth did not demonstrate an effect on the quality of paternal bond. These results are in line with other recent studies on predictors of postpartum bonding. Bieleninik et al. (2021), for example, found no significant association between paternal postpartum bonding and older age, low educational level or being a first time father.

Concerning the influence of pre- and postpartum paternal depressiveness and anxiety as well as maternal depressive symptoms before birth on paternal bonding, contrary to our expectation, we found no significant influence of paternal postpartum depressive symptoms 3 month after birth. Depressive symptoms during pregnancy was the variable, which best predicted all dimensions of paternal bonding. These results are particularly surprising, as a number of studies have shown that maternal depression, especially postpartum (even related to symptoms rather than specific depression diagnosis), is a critical factor in the development of bonding disorders (Dubber et al., 2015; Lehnig et al., 2019; Muzik et al., 2017; Tolja et al., 2020). We expected a similar independent effect by fathers. A possible explanation of these results could be the strong correlation between pre- and postpartum depressiveness (Table 3), in which case it could be inferred that paternal depressive symptoms 3 months postpartum are in this case a continuation of those already existing during pregnancy (for a Review Wee et al., 2011). Maternal depression during pregnancy and after birth were not significant either, indicating an independent paternal bonding process, which is not influenced by the maternal emotional state or by her own bonding process.

There is a body of evidence showing increased rates of anxiety in mothers and fathers in pregnancy compared to the postpartum period (Boyce et al., 2007; Dubber et al., 2015; Figueiredo & Conte, 2011; Lee et al., 2007). In a systematic review about anxiety in fathers in the perinatal period, F. Philpott et al. (2019) reported, that particularly at the time of birth, fathers experience more anxiety with a decrease from birth to the later postnatal period. Similarly, Figueiredo and Conte (2011) found that the prevalence of anxiety decreased from the first trimester to the third trimester, and was the lowest at 3 months postpartum. According to this information, an increase in anxiety symptoms in pregnancy would be expected for all fathers as state anxiety. As we assess trait anxiety, a more constant variable, which may not be specifically highly affected by pregnancy (Huizink et al., 2004; Robertson Blackmore et al., 2016), the decrease in anxious symptoms after the birth that other authors have found was not corroborated in our study. On the contrary, the high correlation we found (Table 3) confirms the stability between trait anxiety during pregnancy and 3 months after birth.

An association between mother’s prenatal anxiety and the quality of postpartum bonding have been investigated in several studies with somewhat inconsistent results. There were hints of an association with higher anxiety with impaired bonding (Dubber et al., 2015; Feldman et al., 1997; Motegi et al., 2020; Nath et al., 2019). However, others are showing that those observations might be only mediated by co-occuring depressive symptoms (Tolja et al., 2020). One study even showed improved bonding in increased anxious mothers (Edhborg et al., 2011). In fathers, a role of anxiety for reduced responsiveness and impaired bonding has been show, whereas this was not true in the mothers (Nakić Radoš, 2021). Data, however, are still very scarce in fathers. We found an interesting result for the fathers in our study. As with the depressive symptoms, paternal trait anxiety only in pregnancy has an independent effect on the total PBQ score, on impaired bonding and on rejection and anger (Table 4): higher levels of trait anxiety were associated with lower scores on the PBQ scales, which indicates lower levels of disturbed bonding. In turn, a lower level of trait anxiety is associated with less bonding quality. Since these results are partly the opposite as described for mothers in the literature, they are difficult to interpret. A possible explanation could be that fathers have difficulties perceiving their partners’ pregnancy as real because they do not experiences the physical and physiological changes women undergo (Fenwick et al., 2012; Genesoni & Tallandini, 2009). Therefore, it might be hard for them to recognise the unborn baby as real. We could hypothesise that this fact could negatively affect the bonding ability of the fathers. In addition, fathers with trait anxiety are just as fearful about these new exiting situations as they usually are. Consequently, they might spend more time being preoccupied with the baby, its health, its needs and future life. Fathers might also be more concerned with their own abilities to cope with the baby and with the changes and challenges in their new life as a father. Through this confrontation, the baby becomes real and the development of a positive emotional bond to the unborn child is then already possible. In this way, the protective effect of trait anxiety exerts on the quality of the bond in the first months after birth could be explained. As we also have not observed father-child-interaction in our study, only assessed the parent--child bonding with the PBQ, we therefore cannot speculate about potential differences between perceived bonding quality and objective quality of parent-child-interaction.

Furthermore, it should be taken into account that a number of other variables, which have not been considered in this study, could mediate the effect of trait anxiety on bonding. For example, other research has shown that an initially clear effect of anxiety on the bonding quality is only mediated by other variables like depression, as stated above, or stress, so that the direct effect was no longer detectable when mediating variables were included in the equation (Bieleninik et al., 2021). In addition, just as surprising is the fact that trait-anxiety measured 3 month after birth has a negative effect on the paternal bonding quality: fathers with high trait-anxiety after birth showed significantly more rejection and anger. Although this result is also difficult to interpret, it could be hypothesised that fathers with trait anxiety, in direct contact and handling with the real baby, become fatigued (Tzeng et al., 2009), insecure (Pinto et al., 2016), stressed (Prino et al., 2016) and more negative regarding their experiences as fathers, so that their trait anxiety increases (L. Philpott et al., 2017; Vreeswijk et al., 2014) along with the negative effects on bonding that are already known by mothers. This seems to be a conceivable explanation, especially considering that rejection and anger was the only subscale in which this effect was significant. However, the levels of trait anxiety reached by the fathers 3 months postpartum were relatively low compared to clinical samples, as they were during pregnancy.

Strengths and limitations

A significant strength of our study is that our findings underline the relevance of paternal mental health for the development of an emotional bond to their babies. While research has focused on the effects of depressive and anxiety symptoms on the quality of bonding in mothers, our study shows the important role of paternal emotional well-being, and not just after the birth. Depressive and anxiety symptoms during pregnancy are highly predictive for the quality of bonding as well as for the presence of these symptoms 3 month after birth.

However, when interpreting our findings some limitations should be taken into account. In this exploratory study, the sample mainly consists of a small group of highly educated, prosperous, German speaking fathers, who were mostly married and expecting their first child. Moreover, we only used self-rating-scales for the assessment of depression and anxiety and, even if the use of self-ratings is a common and valid method for the assessment of symptoms, it can be assumed that the estimation of depressive and anxious symptoms could have been different if we had used clinical interviews and diagnoses. In addition, the paternal symptoms were not really severe, so the results concerning the role of depression and anxiety in predicting the quality of paternal bonding could possibly be different in seriously ill fathers. An important aspect that was not considered in this study is the quality of the couple relationship and coparenting, which can influence the emotional state of the parents and indirectly the quality of the bond.

On the other hand, more anxious and depressive fathers might be less motivated to participate in a study like this one, so possibly we have some selection bias in the study groups. Therefore, the results cannot be generalised. One final important point that can put our results into perspective is that we used the same factor solution for the PBQ that was calculated for women (Brockington et al., 2006) although it is conceivable that the definition of bonding disorders for men has a different focus than that for women. Therefore, even though our results are promising, more research is needed to verify the results obtained and to be able to generalise them to other groups.

Conclusions and clinical implications

Symptoms of depression and anxiety as well as disturbed bonding are well known factors, which negatively affect parental well-being, their parental capabilities and thereby the development of the children. Hence, it is necessary to identify these symptoms as soon as possible in order to prevent later negative impacts on parental mental health and on child developmental outcomes. At the perinatal period, both parents are regularly in contact with the health system. Therefore, there are many opportunities for the early detection of symptoms, for early treatment and for prevention. It seems to be necessary and useful to consider to also screen fathers during pregnancy. The results of our study highlight the importance of being mindful not only of the mother’s mental health, but also that of the father’s in the perinatal period. They also suggest that fathers, like mothers, need to be supported during the transition to parenthood and at the time after birth.

Availability of data and material

The anonymised raw results can be made available on request.

Consent to participate

The participants gave informed written consent to participate in the study.

Consent for publication

The participants gave informed written consent that their anonymised data could be published.

Ethics approval

The Ethics Committee of the University of Frankfurt and the Ethics Committe of the Hesse State Medical Association (Hessische Landesärztekammer) approved the study (votum no 135/17).

Disclosure statement

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

Additional information

Funding

Not applicable. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.