Eine sichere Bindung scheint die negativen Auswirkungen von pränatalem Stress auf die Temperamentsentwicklung des Kindes abzufedern
Abstract
Abstract.Objective: Secure attachment style is a known protective factor regarding psychopathological development. The infant’s attachment style, which is developed during the first two years of life, is therefore considered a moderating factor on the association between prenatal maternal distress and child temperament development which has repeatedly been reported in previous studies. Method: In this longitudinal study on a new sample of 51 mother-child-dyads, reported maternal distress and maternal empathy were assessed during pregnancy. Infant temperament and motor development were assessed at 12 months, while additionally controlling for the infant’s attachment style as a postnatal factor. Results: Infants with secure attachment style whose mothers had experienced higher prenatal distress showed slightly better gross motor development at the age of 12 months. No association could be found between prenatal maternal distress and infant temperament. Conclusions: The results support the view that secure attachment style in children is a protective factor and softens the effects of prenatal maternal distress on difficult temperament development.
Zusammenfassung.Fragestellung: Studien zu pränatalem Stress zeigten wiederholt und sehr konsistent, dass höherer pränataler Stress mit einem schwierigeren kindlichen Temperament einhergeht. Ein sicherer Bindungsstil gilt entwicklungspsychopathologisch als protektiver Faktor. In der vorliegenden Arbeit wurde angenommen, dass sich ein sicherer Bindungsstil eines Kindes auch als ausgleichender Faktor hinsichtlich pränataler Stresserfahrung verstehen lässt und sich bei sicher gebunden Kindern – im Unterschied zu unsicher gebundenen Kindern – kein Zusammenhang von pränatalem Stress und schwierigem Temperament finden lässt. Methodik: In einer Längsschnittstudie mit 51 Mutter-Kind-Dyaden wurden mütterlicher Stress und mütterliche Empathie während der Schwangerschaft erfasst. Kindliches Temperament sowie motorische Entwicklung wurden im Alter von zwölf Monaten erfasst. Zeitgleich wurde der Bindungsstil des Kindes als postnataler Faktor erhoben. Ergebnisse: Kinder mit sicherer Bindung, deren Mütter mehr pränatalen Stress berichtet hatten, zeigten im Alter von zwölf Monaten eine etwas bessere grobmotorische Entwicklung. Wie erwartet zeigte sich bei den Kindern mit sicherem Bindungsstil kein Zusammenhang zwischen pränatalem mütterlichem Stress und kindlichem Temperament. Schlussfolgerung: Die Ergebnisse unterstützen die Annahme, dass ein sicherer Bindungsstil einen Schutzfaktor darstellt, der anscheinend auch die Auswirkungen von pränatalem Stress auf die Temperamentsentwicklung des Kindes abfedern kann.
Introduction
Maternal distress during pregnancy may influence the infant’s development and well-being as has been demonstrated in a number of studies. Higher reported prenatal stress was associated with a more difficult infant temperament (Austin, Hadzi-Pavlovic, Leader, Saint, & Parker, 2005; Davis et al., 2007; Gutteling et al., 2005; Haselbeck et al., 2013; Haselbeck et al., 2017; Huizink, de Medina, Mulder, Visser, & Buitelaar, 2002; Pauli-Pott, Mertesacker, & Beckmann, 2003), higher motor activity or hyperactivity (O’Connor, Heron, Golding, Beveridge, & Glover, 2002; Van den Bergh & Marcoen, 2004), and more anxiety in infants (O’Connor et al., 2002; Van den Bergh & Marcoen, 2004) as well as with lower cognitive development (Zhu et al., 2014). Mixed results with lower and also higher scores on motor development scales have been reported (Brouwers, van Baar, & Pop, 2001; DiPietro, Novak, Costigan, Atella, & Reusing, 2006; Haselbeck et al., 2013; Haselbeck et al., 2017; Huizink, Robles de Medina, Mulder, Visser, & Buitelaar, 2003; Laplante et al., 2004).
Maternal distress during pregnancy is associated with a higher risk for early regulatory problems in infants. In a study on a high risk sample of pregnant mothers, Sidor and colleagues have shown that higher psycho-social distress is followed by more excessive infant crying. Aditionally, more dysfunctional mother-child-interactions, such as the perception of the infant being difficult, and bonding problems were reported (Sidor, Thiel-Bonney, Kunz, Eickhorst, and Cierpka, 2012). Stressful maternal life events during pregnancy have also been shown to be associated with increased depressive symptoms and major depression diagnosis at the age of 17–18 years in the offspring (Kingsbury, Weeks, MacKinnon, Evans, Mahedy, Dykxhoorn, & Colman, 2016).
According to the concept of acumulating risk factors and protective factors for child development, secure infant attachment is considered to be a protective factor (Sroufe, Carlson, Levy, & Egeland, 1999). A secure attachment style can be developed if the child experiences steady, adequate and sensitive care by their caregiver during the sensitive period of attachment development (Petermann, Petermann, & Damm, 2008; Planalp & Braungart-Rieker, 2013). Parental sensitivity means identifying the child’s emotional and behavioural expressions correctly, interpreting associated needs and reacting promptly and adequately. Securely attached children develop an internal working model about themselves and significant persons around them, which allows them to develop functional emotion regulation strategies (Belsky, 2002; Malik, Wells, & Wittkowski, 2014). Insecure and disorganized attachment has been shown to be a precursor of early emotional an behavioural disturbances (Guttmann-Steinmetz & Crowell, 2006; Pauli-Pott, Haverkock, Pott, & Beckmann, 2007; Smeekens, Riksen-Walraven, & van Bakel, 2007). Attachment theory has been conceptualised as emotion regulation theory, with a correlation between insecure attachment and hyperactivated or deactivated emotion regulation (Malik et al., 2014).
Early deficits in social-emotional abilities in infants in turn are associated with higher risk for the development of emotional and behavioural disturbances, e. g. oppositional behaviour (Ulrich, Petermann, Petermann, & Gust, 2016) or disruptive mood dysregulation disorder (Grau, Plener, Hohmann, Fegert, Brähler, & Straub, 2018). Maladaptive strategies in emotion regulation processes can be seen as a risk factor for later psychopathology, concerning internalizing as well as externalizing problems (Fern, Nitkowski, Petermann, & Petermann, 2018). In general, deficits in emotional regulation and subsequent behavioural dysregulation can be seen as a core part of various psychiatric disorders (Crowell, Price, Puzia, Yaptangco, & Cheng, 2017; Holtman & Petermann, 2018; Schmid, Schmeck, & Petermann, 2008).
The acquisition of knowledge and competences about emotions and their regulation is highly dependant on the qualilty of caregiver-child-interaction (Schipper, Kullik, Samson, Koglin, & Petermann, 2013). Maternal insensitivity can be seen as a failure of maternal emotion regulation. Children of mentally ill parents are at a higher risk of emotional regulation problems and development of psychiatric disorders themselves (Suveg, Shaffer, Morelen, & Thomassin, 2011; Wiegand-Grefe & Petermann, 2016). E. g. borderline personality disorder is conceptualized as a disorder of emotional dysregulation and the dialectical behavior therapy concept, which focuses on the acquirement of competences in functional emotion regulation, has shown to be an effective treatment strategy for adolescents and adults (Maffezzoni & Steinhausen, 2017). Children of mothers with borderline personality disorder have shown higher scores on the temperament dimension of harm avoidance and a higher prevalence of emotional and behavioural problems than children of mothers without psychiatric conditions (Barnow, Spitzer, Grabe, Kessler, & Freyberger, 2006). For emotion regulation, a complex interaction between genetic disposition and environmental experiences, including learning from models, can be assumed (Bridgett, Burt, Edwards, & Deater-Deckard, 2015).
Infant temperament construct is thought to reflect biologically based differences in duration, frequency, and intensity of emotional expressions. Rothbart (2007) empirically found three underlying dimensions of infant temperament: extraversion, negative affect (including soothability) and the effortful control (inhibition) of the first two dimenstions. Therefore emotion regulation can be seen as part of infant temperament and it can be assumed that temperament, like, emotion regulation, is determined by genetics and epigenetics through environmental experiences (Feldman, 2008). The temperament construct by Rothbart is based on a neurobiological theory and higher temperamental effortful control has been shown to be associated with higher volume of the orbitofrontal cortex (Whittle, Yücel, Fornito, Barrett, Wood, Lubman, Simmons, Tantelis, & Allen, 2008). While low effortful control in turn has been shown to be associated with more externalizing problems and substance use later in life, orbitofrontal cortex dysfunction has also been shown to be associated with ADHD and substance abuse. Thus, individual differences in temperament and individual scores at extreme ends of temperament dimensions seem to be associated with risk for behavioural and mental health problems on a neurobiological basis (Whittle et al., 2008). It has been shown that behavioural inhibition in early childhood is a temperament factor that can be associated with adolescent anxiety disorder (Fox & Pine, 2012), especially in the clinical diagnosis of social anxiety (Buzzel, Troller-Renfree, Barker, Bowman, Chronis-Tuscano, Henderson, Kagan, Pine, & Fox, 2017; Rapee, 2014). A recent study on children supports a diathesis-stress model, with early temperament and biological markers as risk factors for increased psychological symptoms after the occurrence of an environmental stress factor (Meyer, Danielson, Danzig, Bhatia, Black, Bromet, Carlson, Hajcak, Kotov & Klein (2017). Biobehavioural studies on individual differences in temperament – starting with the interaction between biological givens and environmental provisions in intrauterine environment – are a step toward a better understanding of risk and resilience factors for later psychopathology (Feldman, 2008). A comprehensive review of key constructs in developmental psychopathology, temperament, neurobiology, and epigenetics is given in “Child Temperament: New Thinking About the Boundary Between Traits and Illness” (Rettew, 2013).
Psychobiological animal studies in rat pups of prenatally stressed mothers have demonstrated that pups who postnatally received high levels of licking and grooming by the nursing rat, showed milder behavioural responses to threat and an increased exploratory behaviour compared to pups that did not receive high levels of licking and grooming after birth (Liu et al., 1997). Maternal care after birth was able to influence transcriptional regulation and the offspring’s behaviour in adulthood by producing epigenetic modifications (Weaver et al., 2004). For a detailed review on a series of epigenetic studies see Roth and Sweatt (Roth & Sweatt, 2011). The effect of prenatal distress could, at least partly, be reversed by postnatal rearing style and lasted into the third and fourth generation while generational transmission was associated with changes in gene expression by epigenetic processes (Liu et al., 1997; Weaver et al., 2004).
A recent human study on “the epigenetic clock at birth” has shown that maternal prenatal depression may contribute to disease trajectories in later life, by assessing the epigenetic gestational age as a biomarker and also assessing the child’s behavior at the age of four years via the Child Behavior Checklist (Suarez, Lahti, Czamara, Lahti-Pulkkinen, Knight, Girchenko, Hämäläinen, Kajantie, Lipsanen, Laivuori, Villa, Reynolds, Smith, Binder, & Räikkönen, 2018). A group of researchers at Charité, Berlin, proposes a model of fetal origins of health and disease risk, which focuses on the interactions of prenatal stress, infant temperament, and the quality of postnatal mother-child-relationship regarding the susceptibility to psychopathology (Buss, Entriger, Moog, Toepfer, Fair, Simhan, Heim, & Wadhwa, 2017).
Despite of the number of studies and promising results, the majority of studies have been characterized by different limitations. To date, to our knowledge, only one prospective study on prenatal maternal distress and infant outcome has included the aspect of infant secure attachment as postnatal protective factor: prenatal cortisol exposure has been shown to predict lower infant cognitive development only for children with insecure attachment, while there was no association for children with secure attachment style (Bergmann, Sarkar, Glover, & O’Connor, 2010).
The aim of this study was to investigate how prenatal maternal stress, indexed by the mothers’ subjective experience is associated with infant temperament and infant motor development while controlling for the child’s postnatally developed attachment style. Additionally, the role of maternal empathy as a part of the mothers’ emotion regulation ability was investigated. Based on previous research, it was expected that mothers with higher prenatal stress, compared to those who experienced less stress during their pregnancy, have infants with a more difficult temperament, better gross motor and a lower fine motor development, only if their children had developed an insecure attachment style. In contrast, no association between prenatal maternal distress and infant temperament or motor development was expected for children with secure attachment style. For the study design see Figure 1.
Methods
Sample
In this prospective longitudinal study, 57 mothers and their infants were recruited via advertising in local newspapers and by advertising in birth preparation classes between 2014 and 2015. At the age of 12 months, 51 mothers and their infants could be investigated. After initial exclusion of six interested pregnant women due to our exclusion criteria, there were no dropouts during pregnancy and until 12 months after birth.
For inclusion, participants were required to have sufficient German language skills. Women with high risk-pregnancy were excluded from participation except for higher risk due to age (> 35 years). The exclusion criteria concerning diseases and medication were metabolic diseases, brain organic disorders and psychiatric disorders, substance abuse or substance-related addiction as well as medication with proven influence on cortisol. In order to study the association of prenatal distress and infant outcome variables, and to control for other factors that are known from the literature to influence infant development or temperament, a healthy sample of pregnant women was chosen. In a healthy sample of mothers, as defined by the exclusion criteria, we expected to receive later born children with secure and also insecure attachment style, because the expected distribution pattern in the general public is about 60 % securely and 35 % insecurely attached children, and 5 % with disorganized attachment style, respectively. Screening was conducted via an initial telephone conversation and a subsequent personal appointment, using a semi-structured protocol that, regarding psychiatric conditions, asked for current and/or past psychiatric problems, diagnosis and/or treatment (e. g. depression, anxiety).
All participants gave their written informed consent according to the declaration of Helsinki and the study was approved by the local ethic committee (AZ: A 101/14). For participation the mothers received a compensation of 100.– € at 12 months. Seven mothers had started participation in the study at week 12–14 of gestation (T1), 24 at week 22–24 of gestation (T2), and 20 at week 32–34 of gestation (T3). Mean age was 32.8 years (SD = 4.8; MIN = 23; MAX = 42), six reported a chronic disease (chronic back pain, atopic dermatitis, bronchial asthma, hypothyroidism, Hashimoto thyroiditis, allergies), and two of them regularly used medication (L-Thyroxin). Regarding the 51 new-born babies, 26 of them were girls and 25 were boys, none of them were delivered preterm, all APGAR values ranged from 9 to 10. A total of 36 of the babies were delivered naturally, 14 via a caesarean section and one the mother made no specification about the kind of delivery. The clinical and demographic characteristics of the sample are given in Table 1 and Table 2.
Study procedure
On first personal contact during pregnancy women were handed questionnaires and were thoroughly instructed. At T1 (week 12–14 of gestation), T2 (week 22–24 of gestation), and T3 (week 32–34 of gestation) the pregnant women completed questionnaires (duration approximately one hour) at home. At 12 months after birth, the mothers again completed questionnaires at home. At that time the following investigations were conducted in the laboratory: a standardized developmental assessment battery and video recordings for later evaluation of the infant’s attachment style.
Measures
Based on the stress model by Lazarus and Folkman (Lazarus & Folkman, 1984), stress was assessed multi-dimensionally in order to consider stress provoking events as well as stress resulting reactions on a cognitive, emotional, and physiological level.
Reported maternal distress
For assessment of different aspects of maternal distress, five questionnaires (German versions) were used. Potential stress provoking factors in everyday situations (“daily hassles”) were assessed with the Daily Stress Inventory (Woodside, Winter, & Fisman), for which internal consistency values between Cronbach’s α = 0.94 and α = 0.96 are reported (Traue, Hrabal & Kosarz, 2000). Cronbach’s Alpha was α = 0.95 for this study sample. For subjective stress experience the Perceived Stress Scale (PSS) by Cohen, Kamarck and Mermelstein was used, where internal consistency is reported to be between α = 0.84 and α = 0.86 (Cohen, Kamarck, & Mermelstein, 1983). Cronbach’s Alpha was α = 0.87 in this study sample. Pregnancy-specific stress was measured with the Prenatal Distress Questionnaire (PDQ) by Yali und Lobel for which the reported internal consistency value is α = 0.81 (Yali & Lobel 1999) and α = 0.79 in this study sample. Stress resulting anxiety was investigated with the trait-scale in the State-Trait-Anxiety-Inventory (STAI) by Laux and colleagues for which retest-reliability is reported to be between r = 0.77 and r = 0.91 (Laux, Glanzmann, Schaffner & Spielberger, 1981). Cronbach’s Alpha of the trait-scale was α = 0.87 in this study sample. Stress resulting depression was assessed with the Major Depression Inventory (MDI) by Bech, Rasmussen, Olsen, Noerholm, & Abidgaard for which sensitivity values between r = 0.86 and r = 0.92 were given (Bech, Rasmussen, Olsen, Noerholm & Abildgaard, 2001). Cronbach’s Alpha was α = 0.73 in this study sample.
Maternal empathy
Maternal empathy was assessed prenatally using a German version of the Interpersonal Reactivity Index (IRI) questionnaire (Davis, 1983). The IRI considers empathy to be a multidimensional construct and consists of four subscales: perspective taking (PT) – spontaneous adoption of the point of view of others, empathic concern (EC) – experiencing feelings of compassion for unfortunate others, personal distress (PD) – experiences of distress in response to extreme distress in others and fantasy (FS) – imagination of oneself in fictional situations. Internal consistency is reported to be between Cronbach’s α = 0.73 and α = 0.83 (De Corte et al., 2007). In this study sample, Cronbach’s Alpha was α = 0.78 on the PT subscale, α = 0.83 on the EC subscale, α = 0.70 on the PD subscale and α = 0.62 on the FS subscale.
Maternal report on infant temperament
Infant temperament was assessed postnatally using the Infant Behavior Questionnaire (IBQ) by Pauli-Pott, Ries-Hahn, Kupfer, and Beckmann (1999). In order to minimize maternal observer bias, the items of IBQ consist of behavioural descriptions rather than more abstract characteristics. For the 12 month measurement “activity level”, “soothability” and “fear of novelty” scales were chosen, since on these scales significant correlations with maternal distress had been shown in a previous study of another sample with assessments at the ages of five (Haselbeck et al., 2013) and of 16 months (Haselbeck et al., 2017). Internal consistency is reported to be between Cronbach’s α = 0.72 and α = 0.79.
Infant development
For standardized developmental assessment the developmental test battery “Entwicklungstest ET 6–6” (Petermann, Stein, & Macha, 2008) was used. Besides the international gold standard Bayley Scales III (Bayley, 2006), the ET 6–6 is a good comparable alternative with norms. Internal consistency of scales is reported to be between Cronbach’s α = 0.66 and α = 0.77. The standard procedure allows a detailed analysis of the developmental status in children between six months and six years. In this study, the scales for gross motor development (body movements), fine motor development (hand movements) and cognitive development (action strategy) were used in order to determine gross- and fine-motor as well as cognitive development.
Infant attachment style
For standardized assessment of the infant’s attachment style, the strange situation test by Ainsworth (Ainsworth & Wittig, 1969) was conducted. The strange situation test is a standardized paradigm to assess a child’s attachment style at early infancy. It consists of eight 3-minute-episodes during which the child’s interactional behaviour is rated by two independent trained raters, blind to the hypotheses. According to the classic concept of attachment by Bowlby and Ainsworth, attachment style can be secure, insecure-avoidant, insecure-ambivalent (Ainsworth & Wittig, 1969) or insecure-disorganised (Ainsworth, Blehar, Waters & Wall, 1978). In this study, the child’s attachment style was assessed and videotaped. Videotapes were analysed by two independent raters, who each time reread the criteria for secure, insecure-avoidant and insecure-ambivalent behaviours before watching and rating episode 5 and subsequently episode 8 of one child’s video recording. Children finally were divided into two subgroups (secure and insecure attachment) in order to conduct separate statistical analyses for each subgroup, while children with insecure-disorganised attachment style were excluded from further analyses.
Statistical analyses
Statistical analyses were conducted using PASW Statistics 18 for Windows. Hypotheses concerning linear relationships were tested using Pearson product-moment correlation coefficients and linear regression analyses.
The association of prenatal distress with infant temperament and also with infant motor development was investigated with correlation analyses, separately for the subgroups of secure and insecure attachment style. Additionally, the role of maternal empathy as a part of the mothers’ emotion regulation ability was investigated by (partial-)correlation analyses. It was expected that mothers with higher prenatal stress have infants with a more difficult temperament, better gross motor, and a lower fine motor development, only in the subgroup of children with insecure attachment style. In contrast, no association between prenatal maternal distress and infant temperament or motor development was expected for children with secure attachment style.
The predictability of infant motor development and infant temperament by prenatal distress was tested using simple and multiple linear regression analyses. Multiple linear regression analyses were also used in order to investigate if maternal empathy additionally helps predicting infant motor development or temperament. The predictability of infant attachment style by prenatal distress, maternal empathy and infant temperament was tested by logistic regression analysis for binary dependent variables.
Potentially confounding (socio-demographic) variables were statistically controlled by computation of partial correlations. Statistical significance level was set at 5 % (p ≤ 0.05) and Bonferroni corrections for multiple comparisons were conducted where necessary.
To derive a comprehensive index of “overall stress”, a principal component analysis was conducted, based on maternal stress reports, using four out of the five stress questionnaires and weighing each scale equally. The MDI questionnaire, which assesses depressive symptoms as a resulting stress aspect, was excluded from the principal component analysis, as values for 98 % of the participating mothers were 0; only for one woman in trimester two the value was 1, therefore, there was no data variance in order to conduct any further correlational analyses with this questionnaire. The one factor solution revealed one factor with an eigenvalue of 2.6, explaining 65 % of total variance of maternal stress reports. Accordingly, the scale “overall stress” was built, and an index on this scale was computed for every participant by building linear combinations based on the factor values for the main component.
Before conducting the principal component analysis, in a first step, intercorrelations over all three trimesters were screened for each of the four questionnaires separately, which all were significant (r = 0.59 to r = 0.99). In a second step, mean scores over all three trimesters were built for each questionnaire separately. Then, for the mean scores as well as for each trimester separately, intercorrelations between all four questionnaires were screened, which all were significant except for one intercorrelation in trimester one and two intercorrelations in trimester two (mean scores: r = 0.31 to r = 0.76; trimester one: r = 0.53 to r = 0.85; trimester two: r = 0.29 to r = 0.72; trimester three: r = 0.35 to r = 0.73). Finally, based on the mean scores, the principal component analysis was conducted.
Variable sample sizes were due to missing values in trimester one (n = 7), trimester two (n = 31), and trimester three (n = 49), concerning the prenatal maternal stress scores. Since a comprehensive index of overall stress was built based on mean scores for each participant, there was no missing value for the global prenatal maternal stress score (n = 51). There also were no missing values for the maternal empathy scores or the children’s temperament scores and motor development scores.
Results
Infant attachment style
Attachment style was assessed by two trained raters, who independently, and blinded to mothers prenatal distress or socio-economic data, rated the child’s attachment style, based on the video recordings of the strange situation test (Ainsworth et al., 1978). Inter-rater accordance was 100 %. Against our expectations, all infants, except for one child, clearly showed secure attachment style. Therefore, it was not possible to divide the sample into subgroups for further separate statistical correlation analyses. Also, logistic regression analysis for binary dependent variables could not be conducted due to the small number (n = 1) in the subgroup of insecurely attached children. Instead it was decided to exclude the one child of our study sample that had shown insecure attachment style from further statistical analyses. Therefore, statistical results of the correlation and linear regression analyses refer to the subgroup (n = 50) of securely attached children.
Correlation analyses
Prenatal stress reports associated with infant gross motor development
Results of the correlation analyses revealed a positive linear relationship (r = 0.261, p = 0.034) between prenatal maternal reported stress and the gross motor development score on the ET 6–6. No relationship was found between prenatal maternal stress and fine motor development (Table 3). Statistically significant correlations were independent of parents’ socio-economic status and remained after statistical control of postnatally reported maternal distress via partial correlations.
Prenatal maternal stress not associated with infant temperament
Results of the correlation analyses were not significant (Table 3).
Maternal empathy not associated with infant temperament or motor development
Results of the correlation analyses revealed a positive linear relationship between maternal empathy and a more difficult temperament (r = 0.239, p = 0.049) only for the empathy subscale “empathetic concern” and the temperament subscale “soothability”. There was no significant correlation on any other subscale. Due to multiple testing regarding the association of maternal empathy and infant temperament with four subscales on the empathy questionnaire and three subscales on the temperament questionnaire, Bonferroni adjustment still needed to be performed. This lowers alpha to the level of α = 0.017, which renders the result above non-significant.
Regression analyses
Prediction by overall stress
Simple linear regression analyses were conducted in order to test the predictability of infant motor development and infant temperament by overall prenatal maternal stress. Results of the simple linear regression analyses revealed a significant prediction of gross motor development (R2 = 0.068, p = 0.034) by “overall stress”. Therefore, 7 % of the variance in data can be explained by the overall stress variable. Prediction of fine motor development and infant temperament was not significant in any of the simple linear regression analyses.
Prediction by stress questionnaires in different trimesters
Multiple linear regression analyses using the five stress questionnaires as five separate predictors were conducted for each trimester separately. Results of the multiple linear regression analyses revealed a significant prediction of gross motor development (R2 = 0.369, p = 0.039) by all five stress questionnaires in trimester two. Prediction of gross motor development remains significant, if only two predictors (PSS and PDQ) are used (R2 = 0.304, p = 0.007) and also if only one predictor (PSS) is used (R2 = 0.206, p = 0.012). Of the variance in gross motor development data 21 % can be explained by perceived stress in the second trimester. Pregnancy-specific stress can additionally explain 9 % of the data variance, while adding three more predictors altogether help to explain the additional 7 %. Prediction of fine motor development and also of infant temperament was not significant in any of the multiple linear regression analyses.
Adding the four subscales of the maternal empathy questionnaire to the multiple regression analyses did not ameliorate prediction of infant development or infant development. Prediction of motor development or infant temperament were not significant in any of the multiple linear regression analyses using maternal empathy as additional predictor(s). Thus, adding maternal empathy to the regression analysis renders the significant prediction of gross motor development by maternal distress in trimester two non-significant.
Discussion
The aim of this study was to evaluate how prenatal maternal distress is associated with infant temperament and development. Based on previous studies, we expected a more difficult temperament, lower fine motor developmental status, and better gross motor developmental status in relation with higher prenatal stress, only for insecurely attached children. For securely attached children, no relation was expected between prenatal distress and the child’s temperament or development.
Infant development
In securely attached children, higher prenatal distress was associated with slightly better gross motor development. Infant gross motor development can be predicted by prenatal maternal distress. Perceived stress during the second trimester explains about 20 % of variance in gross motor development data. These findings are in line with the findings of our first study, where infants at the age of five months had also shown better gross motor development scores (Haselbeck et al., 2013). The results are also in line with DiPietro and colleagues (2006) and as DiPietro suggests, a bell-shaped relationship might be assumed between prenatal stess and gross motor development. If a study sample is not representative for the whole population, a bell-shaped relationship between two variables can seem like a positive or negative linear relationship, depending on the sample selected. Looking at the socio-demographic parameters, the extremly good compliance, and low dropout rate of our study sample, it must be asumed that the sample is not representative for the whole population concerning stress variance. Therefore, if the women in this study stample with the highest stress values represent women with medium stress values in the whole population, a positive linear relationship between stress and gross motor development must be expected for this study sample.
Again in line with the findings of our fist study (Haselbeck et al., 2013), no association between prenatal distress and fine motor development could be found at the age of 12 months. It is assumed that prenatal distress might stimulate the children’s gross motor development to further their autonomy at the expense of their later emerging fine motor development; in our previous study, at the age of 16 months (Haselbeck et al., 2017) higher prenatal distress was associated with slightly lower fine motor develompent scores.
Taken together, the correlation observed in securely attached children was only moderate.
Infant temperament
As expected, higher reported prenatal distress was not associated with a more difficult infant temperament at the age of 12 months, in the group of securely attached children. However, if testing for H0 (“No association between two variables”), conventionally alpha needs to be set high (.25) and no Bonferroni adjustments must be done. Test logic tells that, if no Bonferroni adjustment is done and alpha is high and there are still no significant results, it can be assumed that there is no association. Furthermore, different studies have reported higher prenatal distress to be associated with a more difficult temperament (Austin et al., 2005; Davis et al., 2007; Gutteling et al., 2005; Huizink et al., 2002). It seems possible that prenatal distress as a risk factor for a difficult temperament is moderated by postnatal secure attachment style. This would be in line with the concept of accumulating risk and protective factors for child development (Petermann & Petermann, 2011; Petermann et al., 2008). Difficulties in early mother-child-interaction are known to be an antecedent of problems in emotion regulation, insecure attachment, and later psychopathology (Malik et al., 2014).
Limitations
The strengths of this study are the successful realization of a longitudinal study design with a low dropout rate and a good operationalization of theoretical concepts including the strange situation test, which is quite absorbing concerning the time of researchers and participants, respectively. Still, there are also weaknesses that have to be mentioned:
In this study sample, the strange situation test revealed that 50 out of 51 children were securely attached. While for the individual child this is an excellent result, for our study design, we had expected to achieve the typical distribution that can be found in a normal population which would have consisted of about 60 % securely attached children and about 40 % insecurely attached children. For statistical analyses we subsequently could not divide the sample into subgroups with secure and insecure attachment. According to the concept of accumulating risk and protective factors, we had expected to find correlations between prenatal distress and child temperament as well as development only in the subsample with insecure attachment. Since we considered secure attachment to be a protective factor, we expected no correlation between prenatal distress and infant temperament or development. Statistical comparisons between secure and insecure attachment group could therefore not be conducted (e. g. difference of differences between mean scores), our findings are limited to the subgroup of securely attached children and no generalisations for the whole population can be made. In our previous study sample (Haselbeck et al., 2017), the strange situation test had shown that 18 of the 30 children (60 %) were securely attached, while 12 (40 %) were insecurely attached (no children showing disorganized attachment). This approaches the distribution in a normal population, but due to small subgroups further separate statistical analyses for each subgroup could not be conducted in the previous study. The bias observed in attachment style in the study may reveal a quite strong selection bias in the sample. The reason for the uniform sample in this new study with securely attached children only (except for one child) is not entirely clear; however, this time, recruiting method included not only advertisement in the local newspaper but also advertisement in birth preparation classes through personal contact with the researchers. Therefore, it can be speculated that the subpopulation of women who are attending a birth preparation class differs from the whole population. Potentially, women attending birth preparation classes are generally more compliant and more interested in learning about attachment development. Independent from the benefits of participating in a birth preparation class, these women might also generally be more sensitive. Another possibility is that the women who attended a birth preparation class have become more sensitive in parenting and their child subsequently developed a secure attachment style. Nevertheless, the biggest shortcoming of this study is the lack of the subgroup with insecure attachment style. Differences between the subgroups of secure and insecure attachment style concerning the relation between prenatal maternal stress and infant development remain speculative. In the next step, our study must be repeated including children with insecure attachment style.
The results of this study are limited in terms of external validity. Not only has the infants’ attachment style been distributed in another way than in general population, but also mothers had a higher education level than can be expected in the general population. Looking at the socio-demographic parameters of this study sample, 85 % of the mothers had a university entrance diploma and 63 % a university degree, which clearly does not represent the general population. Furthermore, 98 % of the mothers showed no depressive symptoms at all, which also differs from the distribution in the general population. Like many other studies on psychiatric risk factors in mother-child-dyads, this study sample consists of highly educated and highly motivated families with fairly little stress. The study sample is neither representative for the general population nor for a population at risk that takes advantage of child and adolescent psychiatric care. For studying solely the effect of prenatal distress on child outcome variables, in a first step, a group of healthy women was chosen and factors like substance abuse during pregnancy or a history of maternal psychiatric illness were intentionally excluded at the cost of external validity. However, this limitation in variance reduces the exploratory power of the results quite significantly. Neither children nor mothers of our study sample show the whole spectrum of symptoms that can occur in a high risk study sample. If variance is missing in both independent and dependent data of a given sample, statistical analyses like correlation and regression analyses cannot become significant. In our study sample, infant attachment style could not be predicted by prenatal distress, maternal empathy or infant temperament. Almost all of the children were securely attached, of normal birth weight, and had been delivered without complications. Taken together, the results of our study describe a normal healthy development.
In order to improve external validity, in a next step, this study must be repeated recruiting directly from a high risk population. The introduction of a high risk population increases both variance and expected effect size. Participants could be recruited in cooperation with a youth welfare office, family nurses or mother-and-child facilities. However, in a study on a sample from a high risk population, a more difficult recruitment of participants, because of motivational problems, and a higher dropout rate must be expected. Since compliance can be ameliorated by reducing demands on the participants, a future study in a high risk population might want to collect stress data only once during pregnancy and use less stress questionnaires. As suggested by the results of this study, stress data collection could e. g. focus on trimester two or on perceived stress in order to reduce time required for participation. Furthermore, in a high risk population, not only stress variance increases, but also a number of confounding variables will emerge in the study sample (e. g. domestic violence, maternal alcohol, nicotine or other drug abuse), which are known to potentially influence infant development. Thus, in a study sample from a high risk population, the effect of prenatal distress seems to be inseparable from the effects on the infant development of other variables.
Like infant temperament, infant motor development also depends on both genes and environmental factors. Environmental factors that might stimulate early gross motor development are e. g. toys that are available for the child, the participation in a class for infants and their mothers like PEKIP, DELFI or baby swimming. Also, the living circumstances might influence gross motor development due to resulting restrictions: living in a big vs. small flat, living in the countryside vs. big city. Future studies also should assess these postnatal factors if gross motor development is assessed.
Although, in this study, we controlled for many variables, we did not control for genetic factors. Mother and child share genetic polymorphisms of vulnerability. A link between mother and infant temperament must be expected due to a genetic factor. In this kind of behaviour study, the question of a direct link between the frailty of the mother and of the child cannot be ruled out as possible explanation, which clearly is a methodological limitation. For future studies investigating the development of infant temperament and emotion regulation, while also considering prenatal and early postnatal factors, especially the mother’s emotion regulation, it might be helpful to combine these behavioural data with genetic and epigenetic analyses. (Epi-)genetic analyses can be conducted in a non-invasive manner using buccal swabs. In genetic and epigenetic analyses, larger study samples are required, which in turn have the advantage of good statistical power when small effects are expected, which is the case for effects of prenatal maternal distress. For health prevention in general public epigenetic studies could be of great value because it can be assumed that early prevention also changes gene expression (Schmidt, Petermann, & Schipper, 2012).
Clinical implications
Epidemiological studies have shown a positive relationship between reported prenatal maternal distress and higher prevalence of psychiatric disturbances in the general population, e. g. ADHD, autism-spectrum-disorders, depression, and schizophrenia (Beversdorf et al., 2005; Clements, 1992; Selten, van der Graaf, van Duursen, Gispen-de Wied & Kahn, 1999; van Os & Selten, 1998; Watson, Mednick, Huttunen & Wang, 1999). Epigenetic studies in animal models have shown that genes can be “switched off” by mechanisms like methylation of the DNA and that these mechanisms are influenced by environmental factors (Roth & Sweatt, 2011). Epigenetic studies in humans have reported that infants of mothers with high anxiety and depression during pregnancy show a higher methylation of the glucocorticoid receptor gene (Oberlander et al., 2008). For a detailed review on epigenetic studies on prenatal distress and methylation see Stonawski, Frey, Golub, Moll, Heinrich, and Eichler (2018). Epigenetic studies have introduced a new paradigm for understanding psychiatric disturbances and have the potential to show the meaning of early health prevention.
Secure attachment style has been shown to be a protective factor for later development. The results of this study support the view that early caregiver-child interventions, which help to build a secure attachment style during the first two years of life, might help to protect from future psychopathology and recover from previous distress. Therefore, early life interventions for parents seem to be important instruments to improve health in children at risk in the general population. Instruments that can be implemented are e. g. video-feedback – in order to improve the caregiver’s sensitivity to the infant (Hänggi, Schweinberger, & Perrez, 2011) and to support the development of secure attachment – or the help of a family midwife throughout the child’s first year of life (“Frühe Hilfen”). Studies on the efficacy of the German “Frühe Hilfen” have shown small to medium effects for some of the existing programmes on maternal and also on child outcome variables (for a detailed review on existing studies see Taubner, Munder, Unger, & Wolter, 2013). Still, more follow-up-studies on the efficacy of the German “Frühe Hilfen” are much needed in order to ensure evidence-based practice in this field of work. Infants and parents from high risk populations, e. g. if one parent is mentally ill, should be offered intensive counselling and educational support, especially throughout the child’s first two years of life, e. g. in classes for mentally ill parents and their children. Classes for mentally ill parents could be organized by psychiatrists for adults and should be combined with parallel offers of classes for their children. The children at risk should be offered classes at different ages, depending upon their needs as babies, infants, school children or adolescents. These group offers for mentally ill parents and their children could be organized by psychiatrists for adults in cooperation with psychiatrists for children and adolescents. Common efforts of both sides, adult and child psychiatry, could stop a vicious circle and prevent transgenerational transmission of stress-related disorders like trauma by pre-emptively focusing on the needs of children at risk.
Conflicts of interests: There are no conflicts of interest existing.
Literature
(1978). Patterns of Attachment. A psychological study of the strange situation. New York: Hilsdale.
(1969).
Attachment and the exploratory behavior of one-year-olds in a strange situation . In B. M. Foss (Ed.), Determinants of Infant Behavior. London: Methuen.(2005). Maternal trait anxiety, depression and life event stress in pregnancy: relationships with infant temperament. Early Hum Dev, 81(2), 183–190.
(2006). Individual Characteristics, Familial Experience, and Psychopathology in Children of Mothers With Borderline Personality Disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 45(8), 965–972.
(2006). Bayley Scales of Infant and Toddler Development, Third Edition. San Antonio: Psychological Corporation.
(2001). The sensitivity and specificity of the Major Depression Inventory, using the Present State Examination as the index of diagnostic validity. J Affect Disord, 66(2–3), 159–164.
(2002). Developmental origins of attachment styles. Attach Hum Dev, 4(2), 166–170. doi: 10.1080/14616730210157510.
(2010). Maternal Prenatal Cortisol and Infant Cognitive Development: Moderation by Infant-Mother Attachment. Biol Psychiatry, 67(11), 1026–1032.
(2005). Timing of prenatal stressors and autism. Journal of Autism and Development Disorders, 35, 471–478.
(2015). Intergenerational transmission of self-regulation: A multidisciplinary review and integrative conceptual framework. Psychol Bull, 141(3), 602–654. doi: 10.1037/a0038662.
(2001). Maternal anxiety during pregnancy and subsequent infant development. Infant Behav Dev, 24, 95–106.
(2017). Intergenerational Transmission of Maternal Childhood Maltreatment Exposure: Implications for Fetal Brain Development. Journal of the American Academy of Child & Adolescent Psychiatry. doi: 10.1016/j.jaac.2017.03.001.
(2017). A Neurobehavioral Mechanism Linking Behaviorally Inhibited Temperament and Later Adolescent Social Anxiety. Journal of the American Academy of Child & Adolescent Psychiatry. doi: 10.1016./j.jaac.2017.10.007.
. (2014). Mother-child and father-child interaction with their 24-month-old children during feeding, considering paternal involvement and the child‘s temperament in a community sample. Infant Ment Health J, 35(5), 473–481.
(1992). The incidence of attention deficit-hyperactivity disorder in children whose mothers experienced extreme psychological stress. Georgia Educational Research, 91, 1–14.
(1983). A global measure of perceived stress.J Health Soc Behav(24), 385–396.
(2017). Emotion dysregulation and autonomic responses to film, rumination, and body awareness: Extending psychophysiological research to a naturalistic clinical setting and a chemically dependent female sample. Psychophysiology, 54(5), 713–723. doi: 10.1111/psyp.12838
(2007). Prenatal exposure to maternal depression and cortisol influences infant temperament. J Am Acad Child Adolesc Psychiatry, 46(6), 737–746.
(1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113–126.
(2007). Measuring Empathic tendencies: Reliability and validity of the Dutch version of the Interpersonal Reactivity Index. Psychologica Belgica, 47(4), 235–260.
(2006). Maternal psychological distress during pregnancy in relation to child development at age two. Child Dev, 77(3), 573–587.
(2008). The Intrauterine Environment, Temperament, and Development: Including the Biological Foundations of Individual Differences in the Study of Psychopathology and Wellness. Journal of the American Academy of Child & Adolescent Psychiatry, 47(3), 233–235.
(2018). Zur Bedeutung funktionaler und dyfunktionaler Emotionsregulationsstrategien für die depressive Symptomatik bei Jugendlichen. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherpie, 46, 17–27.
(2012). Temperament and the Emergence of Anxiety Disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 51(2), 125–128.
(2018). Prevalence rate and course of symptoms of disruptive mood dysregulation disorder (DMDD). A population-based study. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie, 46, 29–38.
(2005). The effects of prenatal stress on temperament and problem behavior of 27-month-old toddlers. Eur Child Adolesc Psychiatry, 14(1), 41–51.
(2006). Attachment and externalizing disorders: a developmental psychopathology perspective. J Am Acad Child Adolesc Psychiatry, 45(4), 440–451. doi: 10.1097/01.chi.0000196422.42599.63
(2011). Feinfühligkeitstraing für Eltern. Bern: Huber.
(2013). Fötale Programmierung Der Einfluss von pränatalem mütterlichen Stress auf Entwicklung und Temperament des Säuglings. Kindheit und Entwicklung, 22(4), 1–8.
(2017). Prenatal maternal distress seems to be associated with the infant‘s temperament and motor development: an explorative study. J Neural Transm (Vienna), 124(7), 881–890. doi: 10.1007/s00702-017-1712-0.
(2012). Early family context and development of adolescent ruminative style: moderation by temperament. Cogn Emot, 26(5), 916–926.
(2018). Emotionale Dysregulation. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie, 46(1), 5–6.
(2002). Psychological measures of prenatal stress as predictors of infant temperament. J Am Acad Child Adolesc Psychiatry, 41(9), 1078–1085.
(2003). Stress during pregnancy is associated with developmental outcome in infancy. J Child Psychol Psychiatry, 44(6), 810–818.
(2016). Stressful Life Events During Pregnancy and Offspring Depression: Evidence From a Prospective Cohort Study. Journal of the American Academy of Child & Adolescent Psychiatry. doi: 10.1016/j.jaac.2016. 05.014.
(2004). Stress during pregnancy affects general intellectual and language functioning in human toddlers. Pediatr Res, 56(3), 400–410.
(1981). State-Trait-Angstinventar STAI. Weinheim: Beltz.
(1984). Stress, appraisal, and coping. New York: Springer.
(1997). Maternal care, hippocampal glucocorticoid receptors, and hypothalamic-pituitary-adrenal responses to stress. Science, 277(5332), 1659–1662.
(2017). Ambulante Dialektisch-Behaviorale Therapie für Adoleszente (DBT-A) mit impulsiven und selbstschädigenden Verhaltensweisen. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie, 45(6), 453–462.
(2014). Emotion regulation as a mediator in the relationship between attachment and depressive symptomatology: A systematic review. J Affect Disord, 172(C), 428–444.
, et al. (2017). Neural Biomarker Prospectively Predicts Increases in Anxiety Symptoms in Children After Hurricane Sandy. Journal of the American Academy of child & Adolescent Psychiatry. doi:10.1016/j.jaac.2017.02.005.
(2008). Prenatal exposure to maternal depression, neonatal methylation of human glucocorticoid receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics, 3, 97–106.
(2002). Maternal antenatal anxiety and children‘s behavioural/emotional problems at 4 years. Report from the Avon Longitudinal Study of Parents and Children. Br J Psychiatry, 180, 502–508.
(2007). Negative emotionality, attachment quality, and behavior problems in early childhood. Infant Ment Health J, 28(1), 39–53. doi: 10.1002/imhj.20121.
(2003). Ein Fragebogen zur Erfassung des “frühkindlichen Temperaments” im Elternurteil. Kinder- und Jugendpsychiatrie und Psychotherapie, 31(2), 99–110.
(1999). Konstruktion eines Fragebogens zur Erfassung des „frühkindlichen Temperaments“ im Elternurteil – Ergebnisse für den Altersbereich 3–4 Monate. Praxis der Kinderpsychologie und Kinderpsychiatrie, 48, 231–246.
(2011). Prävention. Kindheit und Entwicklung, 20(4), 197–200.
(2008). Entwicklungstest ET6. Frankfurt a. M.: Pearson Assessment.
(2008). Entwicklungspsychopathologie der ersten Lebensjahre. Zeitschrift für Psychiatrie, Psychologie und Psychotherapie, 56(4), 243–253.
(2013). Temperamental precursors of infant attachment with mothers and fathers. Infant Behav Dev, 36(4), 796–808. doi: 10.1016/j.infbeh.2013. 09.004.
(2014). Preschool Environment and Temperament as Predictors of Social and Nonsocial Anxiety Disorders in Middle Adolescence. Journal of the American Academy of Child & Adolescent Psychiatry, 53(3), 320–328.
(2013). Child Temperament: New Thinking About the Boundary Between Traits and Illness. New York: Norton.
(2011). Annual Research Review: Epigenetic mechanisms and environmental shapin of the brain during sensitive periods of development. Journal of Child Psychology and Psychiatry, 52(4), 398–408.
(2007). Temperament, Development and Personality. Current Directions in Psychological Science, 16(4), 207–212.
(2013). Emotionsdysregulation im Kindesalter. Psychologische Rundschau, 64, 228–234.
(2008). Persönlichkeitsstörungen im Kindes- und Jugendalter? Kindheit und Entwicklung, 17(3), 190–202.
(2012). Epigenetik – Revolution der Entwicklungspsychopathologie? . Kindheit und Entwicklung, 21(4), 245–253.
(1999). Psychotic illness after prenatal exposure to the 1953 Dutch Flood Disaster. Schizophrenia Research, 35, 243–245.
(2012). Das exzessive Schreien bei 5 Monate alten Säuglingen und prä-, peri- und postnatale Belastungen ihrer Mütter in einer Hoch-Risiko-Stichprobe. Zeitschrift für Kinder- und Jugendpsychiatrie und Psychotherapie, 40(4), 239–250.
(2007). Multiple determinants of externalizing behavior in 5-year-olds: a longitudinal model. J Abnorm Child Psychol, 35(3), 347–361. doi: 10.1007/s10802-006-9095-y.
(1999). Implications of attachment theory for developmental psychopathology. Dev Psychopathol, 11(1), 1–13.
(2018). Affektive Belastungen der Mutter in der Schwangerschaft und assoziierte epigenetische Veränderungen beim Kind: Eine Übersicht. Zeitschrift für Kinder- und Jugendpsychatrie und Psychotherapie, 46(2), 155–167.
(2018). The Epigenetic Clock at Birth: Associations With Maternal Antenatal Depression and Child Psychiatric Problems. Journal of the American Academy of Child & Adolescent Psychiatry. doi: 10.1016/j.jaac.2018.02.011.
(2011). Links between maternal and child psychopathology symptoms: mediation through child emotion regulation and moderation through maternal behavior. Child Psychiatry Hum Dev, 42(5), 507–520. doi: 10.1007/s10578-011-0223-8.
(2013). Wirksamkeitsstudien zu Frühen Hilfen in Deutschland. Zeitschrift für Kindheit und Entwicklung, 22(4), 232–243.
(2000). Alltags Belastungs Fragebogen (ABF): Zur inneren Konsistenz, Validierung und Stressdiagnostik mit dem deutschsprachigen Daily Stress Inventory. Verhaltenstherapie und Verhaltensmedizin (1), 15–38.
(2016). Verhaltensauffälligkeiten bei Kindern mit Entwicklungsdefiziten im Vorschulalter. Zeitschrift für Entwicklungspsychologie und Pädagogische Psychologie, 48(2), 80–89.
(2004). High antenatal maternal anxiety is related to ADHD symptoms, externalizing problems, and anxiety in 8- and 9-year-olds. Child Dev, 75(4), 1085–1097.
(1998). Prenatal exposure to maternal stress and subsequent schizophrenia. The May 1940 invasion of The Netherlands. British Journal of Psychiatry, 172, 324–326.
(1999). Prenatal teratogens and the development of adult mental illness. Developmental Psychopathology, 11, 457–466.
(2004). Epigenetic programmin by maternal behavior. Nature Neuroscience(7), 847–854.
(2016). Kinder psychisch erkrankter Eltern. Kindheit und Entwicklung.
(2008). Neuroanatomical Correlates of Temperament in Early Adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 47 (6), 682–692.
(1991). Salivary cortisol in children: correlations with serum values and effect of psychotropic drug administration. Can J Psychiatry, 36(10), 746–748.
(1999). Coping and distress in pregnancy: An investigation of medically high risk women. Journal of Psychosomatic Obstetrics and Gynecology(20), 39–52.
, (2014). Does prenatal maternal stress impair cognitive development and alter temperament characteristics in toddlers with healthy birth outcomes? Dev Med Child Neurol, 56(3), 283–289.
