Maternal Mindfulness During Pregnancy and Early Child Social-Emotional Development and Behavior Problems: The Potential Mediating Role of Maternal Mental Health

During the perinatal period, women are exposed to major changes, holding possible adverse effects on psychological well-being and child development. An effective way of coping with these challenges and adjustments could be mindfulness. The current study examined associations of mindfulness facets during pregnancy with toddler’s social-emotional development and behavior problems, as well as potentially mediating effects of maternal mental health. A total of 167 women completed questionnaires during pregnancy, the postpartum period, and 2 and 3 years after childbirth, assessing dispositional mindfulness, pregnancy distress, symptoms of anxiety and depression, child social-emotional development, and child behavior problems. There was a positive association between maternal non-reacting mindfulness skills and child social-emotional development at 2 years of age, regardless of mothers’ perinatal mental health, whereas other mindfulness facets were unrelated. Pregnancy distress mediated the association between maternal non-judging skills during pregnancy and child externalizing problem behavior, but no other mediating effects were found. Non-reacting skills in mothers during pregnancy may have favorable implications for child social-emotional development. Due to the relatively high number of tests that were conducted, the mediating effect of pregnancy distress must be interpreted with caution.

learn how to self-regulate their emotions through observational learning, modelling, and social referencing (Morris et al., 2007). Healthy social-emotional development lays a foundation for psychological functioning in later life (Jones et al., 2015;Woodward et al., 2017). Maternal mental health problems (e.g., depressive symptoms) may interfere with this process, as these can lead to less sensitive caregiving behaviors (Bernard et al., 2018;Booth et al., 2018) and modeling of maladaptive coping strategies (Rutherford et al., 2015). Children of mothers that suffer from psychopathological symptoms may therefore be at risk for poor social-emotional development, as well as internalizing and externalizing behavior problems, which are closely related (Huber et al., 2019).
In certain stages of life, women are more vulnerable for developing mental health problems (Tadeka, 2010). In the perinatal phase for example, women are exposed to major physiological, psychological, and social changes, holding possible adverse effects on psychological well-being (Parfitt 1 3 & Ayers, 2014). Prevalence rates of depression are estimated to be 13% prenatally and range from 5 to 25% for the postpartum period (Woody et al., 2017). For anxiety problems, proportions range from 16 to 25% among pregnant women and 17 to 21% among new mothers (Fairbrother et al., 2016;Parfitt & Ayers, 2014). Numerous studies have found unfavorable developmental outcomes for children of mothers faced with postnatal mental health problems. For example, maternal depressive symptoms in the postnatal phase significantly predicted social-emotional problems in children at 6 months (Mason et al., 2011) and 2 years of age (Junge et al., 2017), as well as problem behavior in 2-year-olds (Wesselhoeft et al., 2021). Moreover, maternal postnatal anxiety was associated with adverse social-emotional development of toddlers (Polte et al., 2019;Rees et al., 2019). Furthermore, an increase in anxiety symptoms during the perinatal phase predicted children's social-emotional problems at 12 months of age (Porter et al., 2019).
Even during pregnancy, adequate maternal mental health is important for child developmental outcomes. Previous research established a relation between prenatal stress and children's social-emotional development, behavioral problems (e.g., Hentges et al., 2019;Kvalevaag et al., 2015;Madigan et al., 2018;Wesselhoeft et al., 2021), and neurodevelopmental disorders (Graignic-Philippe et al., 2014;Manzari et al., 2019). The majority of these studies assessed either general stress (Manzari et al., 2019;Wesselhoeft et al., 2021), or anxiety and depressive symptoms as indicators of prenatal stress (Hentges et al., 2019;Kvalevaag et al., 2015;Madigan et al., 2018), whereas stress directly related to pregnancy and upcoming parenthood was disregarded. However, pregnancy-specific stress (e.g., feeling anxious or worry about the baby's health or upcoming delivery) appears to be a distinct construct (Alderdice et al., 2012), yielding greater effects sizes for associations with infant birth weight and gestational age compared to other prenatal stress measures (e.g., trait anxiety; Bussières et al., 2015) and thus, perhaps, functions as a more accurate predictor of children's social-emotional and behavior outcomes.
One way of coping with the emotional demands, unexpected challenges, and social adjustments of the transition into parenthood could be mindfulness, which is defined as a mental state of moment-to-moment awareness and acceptance of feelings, thoughts, and bodily sensations in a non-judgmental and non-reactive manner (Bishop et al., 2004;Kabat-Zinn, 2015). Mindfulness involves self-regulation of attention towards the here-and-now as it is, rather than ruminating and worrying (Bishop et al., 2004;Shapiro et al., 2008), for instance about the upcoming delivery or future parenting qualities. An important facet of mindfulness involves the non-judgmental attention to and acceptance of thoughts and feelings, which allows for adopting a wider decentered perspective on things, free from own preoccupied interpretations and cognitions (Bishop et al., 2004). This may be particularly beneficial for women in the perinatal phase as self-criticism is known to increase from pregnancy to postpartum (Brassel et al., 2020). Mindfulness is believed to increase awareness that negative and hopeless thoughts are displays of the own mind, rather than a reflection of reality, an insight that will enable more efficient disengagement from such disruptive thoughts (Bishop et al., 2004). Furthermore, the ability to remain calm, notice and observe, and let go of automatic responses towards stimuli that would normally elicit direct reaction (i.e., the mindfulness facet of non-reacting) presumably leads to a greater understanding of emotional states (Bishop et al., 2004) and subsequently adequate coping (Bishop et al., 2004;Desrosiers et al., 2013;Rutherford et al., 2015). Hence, maternal mindfulness might have a protective character for mental health outcomes across the perinatal period, as well as on child social-emotional and behavioral development.
Previous research supports the notion that maternal mindfulness during pregnancy has positive effects on psychological well-being of mothers and their children's development. Mothers' mindfulness skills during pregnancy have been associated with less prenatal and postnatal emotional distress (Braeken et al., 2017) and, in one of our earlier studies using a larger subsample, fewer depressive symptoms at 22 and 32 weeks of gestation and lower chance of child low birth weight (Nyklíček et al., 2018). Moreover, mindfulness interventions have been demonstrated to reduce stress, anxiety, depression, and negative affect during pregnancy (Beddoe et al., 2009;Dunn et al., 2012;Vieten & Astin, 2008), continuing into the postnatal phase (Dunn et al., 2012). Concerning child development, mothers' prenatal mindfulness capacities were negatively related to infant self-regulation problems (Van den Heuvel et al., 2015) and negative social-emotional behavior (Braeken et al., 2017), and this association was mediated by maternal anxiety assessed during pregnancy (Van den Heuvel et al., 2015). Finally, children of parents that reported more dispositional mindfulness in early childhood displayed fewer internalizing and externalizing behavior problems (Parent et al., 2016), as well as fewer symptoms of anxiety and depression (Bird et al., 2021). Examining whether dispositional mindfulness during pregnancy has similar associations with child behavior is important and would contribute to our knowledge on the pathways leading to healthy behavior in early childhood.
The current study examined the association of maternal mindfulness during pregnancy with early child social-emotional development and behavior problems, as well as potentially mediating effects of pregnancy distress and postnatal symptoms of anxiety and depression. First, we hypothesized that higher level of mindfulness skills during pregnancy would be associated with better social-emotional development of children at 2 years of age and less internalizing and externalizing problem behavior at 3 years of age. Second, we expected that mothers' pregnancy distress and postnatal symptoms of anxiety and depression would mediate the relation between maternal mindfulness during pregnancy and children's social-emotional development and problem behavior.

Participants
The current study was part of the Holistic Approach to Pregnancy and the first Postpartum Year (HAPPY) project, a large longitudinal prospective cohort study following women from early pregnancy into the first postpartum year (for the complete study design, see Truijens et al., 2014). It concerned a secondary analysis of this study, which was originally designed to examine physiological and psychological well-being and pregnancy-specific symptoms over time. Dutch-speaking white women or third-generation women of other ethnic groups that were in their first trimester of pregnancy were eligible for participation. Exclusion criteria included multiple pregnancies (or higher order pregnancies), endocrine disorder, use of thyroid medication, severe psychiatric disease (schizophrenia, borderline personality disorder, or bipolar disorder), HIV, drug or alcohol addiction problems, or any other disease resulting in treatment with drugs that are potentially adverse for the fetus and need careful follow-up during pregnancy. Eligible women (n = 3160) were informed about the study by their midwife practice at their first antenatal control visit between 6 and 10 weeks of gestation. Of these women, 2269 (72%) consented to participate and completed questionnaires. Only participants who were included in the study between March and December 2013 received the mindfulness questionnaire at 22 weeks of gestation. In total, 984 women (43%) received the mindfulness questionnaire that 912 women (93%) completed. After participation, a sample of 828 (36%) women (selected for purposes that are not relevant for the current study, i.e., [1] elevated levels of depressive symptoms, [2] suboptimal thyroid function or thyroid dysfunction, and [3] healthy controls) was invited for a follow-up study (the HAPPY follow study), of which 485 (59%) agreed to participate. This resulted in a subgroup of 167 women who completed the mindfulness questionnaire as well as the measures on social-emotional development and behavior, and who were therefore eligible for inclusion in the current study (see Fig. 1 for a flow diagram of enrollment). The average age of participants was 30.96 years (SD = 3.59, range 21-40 years). The majority had a higher education or university degree (75%) and was married or co-habiting (99%). Seventy-eight women (47%) were pregnant with their first child. At the start of participation, on average, women were 22.72 weeks pregnant (SD = 1.13, range 21-29.29 weeks). At followup, the mean age of children was 24.38 months (SD = 0.57, range 23-26 months) and 38.22 months (SD = 1.08, range 36-41 months) when assessing their social-emotional development and behavior problems, respectively. Eighty-five children (51%) were boys.

Procedure
Participants received questionnaires via postal mail or email during the second and third trimester of pregnancy to assess dispositional mindfulness and pregnancy distress, respectively, and 12 months postpartum, to assess symptoms of anxiety and depression. These data were collected between January 2013 and September 2014. When the child was 2 years of age, women received a questionnaire via email again, in order to assess the child's social-emotional development. At 3 years of child age, research-psychologists visited women at home. Participating women completed a questionnaire on internalizing and externalizing problem behavior of their child. These data were collected between December 2016 and June 2018.

Dispositional Mindfulness
Mindfulness during pregnancy was assessed using the selfrated Three-Facet Mindfulness Questionnaire-Short Form (TFMQ-SF; Truijens et al., 2016). This questionnaire is derived from the Five Facet Mindfulness Questionnaire-Short Form (FFMQ-SF; Baer et al., 2006) which is a frequently used questionnaire that assesses different facets of mindfulness. The TFMQ-SF holds the three facets that have been found to significantly predict psychological problems in different samples (Baer et al., 2006;De Bruin et al., 2012), omitting the two scales that have been more controversial regarding being part of the concept of mindfulness in nonmeditating samples (Baer et al., 2006;Grossman, 2008). The Dutch version was validated among pregnant women and showed good psychometric properties (Truijens et al., 2016). The questionnaire consists of twelve items divided over three subscales, measuring different facets of mindfulness: (1) Acting with awareness, (2) Non-judging, and (3) Nonreacting. For each item, participants are asked to indicate to what extent this applied to them during their pregnancy on a 5-point Likert scale ranging from 1 = never or very rarely true to 5 = very often or always true. Subscale scores range from 4 to 20. Eight items were reverse scored prior to the data analyses, so that higher scores reflected higher levels of dispositional mindfulness. Examples of items are "I find myself doing things without paying attention" (subscale Acting with awareness; negatively worded item), "I make judgments about whether my thoughts are good or bad" (subscale Non-judging; negatively worded item), and "I watch my feelings without getting carried away by them" (subscale Non-reacting). Cronbach's α coefficients for the current study were 0.86, 0.80, and 0.73 for subscales acting with Acting with awareness, Non-judging, and Non-reacting, respectively. In addition, McDonald omega coefficient indicated a sufficient reliability with ω = 0.91, 0.86, and 0.80 for subscales Acting with awareness, Non-judging, and Nonreacting, respectively.

Pregnancy Distress
Pregnancy distress was measured using the self-rated 11-item Negative affect subscale of the Tilburg Pregnancy Distress Scale (TPDS; Pop et al., 2011). Women indicated to what extent they worry about their pregnancy, the upcoming delivery, and the postpartum period on a 4-point Likert scale ranging from 0 = very much to 3 = not at all. Examples of items are "I worry about the health of my baby" and "I worry about my job after the baby is born." Scores ranged from 0 to 33 with higher score reflecting more pregnancy distress. The TPDS demonstrated good validity and reliability, especially for the subscale Negative affect (Boekhorst et al., 2020;Pop et al., 2011). In the current study, reliability was good with a Cronbach's α coefficient = 0.80 and McDonald ω coefficient = 0.83 for the TPDS subscale Negative affect.

Postnatal Symptoms of Anxiety and Depression
Symptoms of anxiety and depression were measured with the self-rated Edinburgh Depression Scales (EDS-10; Cox et al., 1987). Although primarily used for the assessment of depression, the EDS-10 demonstrated to identify anxiety as well (Brouwers et al., 2001;Matthey, 2008;Pop et al., 1992). Therefore, the current study examined anxiety symptoms (Items 3, 4, and 5; e.g., "I have been anxious or worried for no good reason") separately from depressive symptoms (Items 1, 2, 6, 7, 8, 9, and 10; e.g., "I have been so unhappy that I have been crying"). Previous research confirmed a two-factor model of the EDS-10 in a sample of women 6 weeks postpartum (Matthey, 2008). All items are rated on a 4-point Likert scale ranging from 0 to 3, yielding subscale scores of minimum 0 and maximum 9 for Anxiety symptoms and 21 for Depressive symptoms. Higher scores reflected more symptoms of anxiety and depression. In the current study, Cronbach's α coefficients were 0.73 and 0.87, and McDonald ω coefficients were 0.82 and 0.93 for subscales Anxiety and Depressive symptoms, respectively.

Child Social-Emotional Development
Social-emotional development of the child was assessed using a parent-rated questionnaire of the Bayley Scales of Infant and Toddler Development -Third Edition -NL (Bayley-III-NL; Van Baar et al., 2014). The social-emotional questionnaire of the Bayley-III-NL was found valid and reliable in sample of 1845 mothers (Van Baar et al., 2014). Dependent on the child's age, the questionnaire consists of 21-35 items with 6 answer categories (0 = don't know, 1 = never, 2 = sometimes, 3 = half of the time, 4 = often, 5 = always). Examples of items are "Looks at interesting things like faces or toys" and "Seems happy when he or she sees a favorite face." Total scores on all items are transformed into standardized index scores that were used in the current study. Index scores ranged from 40 to 160 with higher scores reflecting better social-emotional competencies. In the current study, reliability was excellent with Cronbach's α coefficient = 0.90 and McDonald ω coefficient = 0.95.

Child Behavior Problems
Internalizing and externalizing behavior problems of the child were measured with the parent-rated Child Behavior Checklist for ages 1.5-5 (CBCL; Achenbach & Rescorla, 2000). The CBCL was found to be a valid and reliable standardized questionnaire for the assessment of children's behavioral and emotional problems (Koot et al., 1997). The version for infants and toddlers consists of 99 items (e.g., "Afraid to try new things" and "Gets in many fights"). For each item, mothers were asked to indicate to what extent the statement applies to their child on a 3-point Likert scale ranging from 0 = not true to 2 = very true or often true. In the current study, standardized t-scores for subscales Internalizing (i.e., the sum of scores on anxious/depressive, withdrawn/depressive, and somatic problems) and Externalizing behavior (i.e., the sum of scores on rule-breaking and aggressive behavior) were used, with 99.73% of scores ranging between 20 and 80 and higher scores reflecting more problematic behavior.

Data Analyses
The Anxiety and Depression subscales of the EDS-10, as well as the subscale Negative affect of the TPDS, were positively skewed and therefore logarithmically transformed. Other outcome variables were normally distributed with values for skewness and kurtosis within acceptable limits. The TPDS, age, and education values were missing for one participant (0.6%). Scores on the EDS-10 and CBCL were missing for 24 (14%) and 4 (2%) participants, respectively. Little's missing completely at random (MCAR) test was not significant (χ 2 (27) = 31.16, p = 0.265) indicating that data was missing completely at random. Power analyses in the software program G*Power (Faul et al., 2007) for linear multiple regression (effect size = 0.15, alpha = 0.05, power = 0.80) revealed a minimum sample size of n = 114. Therefore, cases with missing data were listwise deleted. Independent samples t-tests and chi-square tests were conducted in order to assess differences in characteristics between the sample of the current study and participants of the original cohort study that completed a mindfulness questionnaire but did not participate in the follow-up. Significant differences with medium effect sizes of Cohen's d > 0.30 or Cramer's V > 0.20 were considered clinically relevant (Cohen, 1988).
Next, Pearson's r correlation coefficients were calculated between the three facets of mindfulness (Acting with awareness, Non-judging, and Non-reacting), pregnancy distress, postnatal anxiety symptoms, postnatal depressive symptoms, child social-emotional development, and child internalizing and externalizing behavior problems. For mediation analyses, we used the SPSS macro developed by Preacher & Hayes (2008) for assessing indirect effects, with 5000 bootstrap samples and 95% confidence intervals. We included the three facets of mindfulness during pregnancy as predictors, pregnancy distress and postnatal anxiety and depressive symptoms as mediator variables, and child socialemotional development, or internalizing or externalizing behavior problems as the dependent variable. Maternal age, maternal education level, and sex of the child were included in the analyses as covariates, since previous studies associated these variables with maternal and child well-being (e.g., Biaggi et al., 2016). All statistical analyses were conducted in Statistical Package for the Social Sciences (IBM SPSS version 27.0).

Results
Independent samples t-test and chi-square tests were conducted to examine any differences in characteristics between the subsample of the current study (n = 167) and the participants of the original cohort study that completed the questionnaire on mindfulness but did not participate in the follow-up (N = 745). Women in the current study were slightly older (M = 30.96, SD = 3.59) than women in the original sample (M = 29.93, SD = 3.57, t(910) = − 3.37, p < 0.001, Cohen's d = 0.29: small effect size) and were highly educated more often (χ 2 (4) = 11.85, p = 0.019, V = 0.12, small effect size). No significant differences were found for gestational age at the start of participation and marital status. In addition, the current subsample of women did not significantly differ from the original sample on mindfulness facets acting with awareness and non-judging, as well as pregnancy distress, and postnatal symptoms of anxiety and depression. For the mindfulness facet non-reacting, a difference was found (t(910) − 2.75, p = 0.006, Cohen's d = 0.26: small effect size) with slightly higher scores for participants in the current subsample (M = 12.37, SD = 3.81) than in the original sample (M = 11.39, SD = 4.22). These differences with small effect sizes show that there were no clinically relevant differences between the current sample and the original sample. Table 1 presents the sample characteristics and bivariate correlations between all variables. Concerning different maternal mindfulness facets, acting with awareness was positively associated with non-judging (r = 0.37, SE = 0.09, p < 0.001, CI [0.28,0.49]), but both were unrelated to nonreacting. A significant association was found between nonreacting and child social-emotional development (r = 0.20, SE = 0.09, p = 0.009, CI [0.05,0.34]). Acting with awareness and non-judging were unrelated to child social-emotional development. No associations were found between facets of mindfulness and child internalizing and externalizing behavior problems. Further, acting with awareness was significantly negatively associated with pregnancy distress ( All maternal mental health variables were unrelated to child social-emotional development. Next, mediation analyses were conducted. Regression coefficients for a-and b-paths and indirect effects are presented in Table 2. Results revealed only one significant mediating effect: pregnancy distress significantly mediated the association between maternal non-judging skills during pregnancy and child externalizing problem behavior (ab = − 0.08, SE = 0.06, CI [− 0.17, − 0.00], p = 0.036; see Fig. 2). No mediating effects were found for postnatal symptoms of anxiety and depression for either child social-emotional  development or internalizing and externalizing problem behavior. Also, analyses indicated no significant mediation effect of pregnancy distress and postnatal symptoms of anxiety and depression for the other two facets of mindfulness (acting with awareness and non-reacting), and all three child outcomes (social-emotional development and internalizing and externalizing problem behavior). However, two facets of mindfulness (i.e., acting with awareness and non-judging) were significantly associated with maternal mental health variables (a paths), while the third facet (non-reacting) showed a significant total effect on social-emotional development (c = 0.45, t(132) = 2.04, p = 0.044, CI [0.01,0.89]). In addition, pregnancy distress was significantly associated with child externalizing behavior problems and-in one model-with child social-emotional development (b paths; see Fig. 3). Concerning the covariates in this study, analyses that included acting with awareness and non-reacting as predictors revealed a significant effect of maternal age on pregnancy distress (B = − 0.01, t(129-132) = − 2.69 to − 2.34, p = 0.018-0.021) and-in two total effect models-child internalizing problem behavior (B = − 0.47, t(129) = − 2.03 to − 2.01, p = 0.045-0.046), with older mothers reporting less distress during pregnancy and fewer internalizing problems in their children at 3 years of age.

Discussion
The present study aimed to examine the association of maternal mindfulness during pregnancy with early child social-emotional development and internalizing and externalizing behavior problems, as well as potentially mediating effects of pregnancy distress and postnatal symptoms of anxiety and depression. Results showed a significant total effect for mindfulness facet non-reacting on child socialemotional development, with mothers who demonstrated more non-reacting skills during pregnancy, reporting more social-emotional competence in their children at 2 years of age. This association was not mediated by either pregnancy distress or postnatal symptoms of anxiety and depression. In contrast to non-reacting, mindfulness facets acting with awareness and non-judging were unrelated to child Fig. 2 Indirect, direct, and total effect of non-judging and child externalizing behavior problems. Note. Coefficients are standardized Fig. 3 Paths from mediation analyses. Note. Child social-emotional development was assessed at 2 years of age. Child internalizing and externalizing behavior problems were assessed at 3 years of age. Coefficients are standardized. Slight deviations in coefficients were found for the effect of non-judging on postnatal anxiety and depressive symptoms (i.e., − 0.42 ** and − 0.35 ** respectively for the model with outcome variables child social-emotional development), as well as the effect of pregnancy distress on child externalizing behavior problems (i.e., 0.18 for acting with awareness and 0.23 * for nonreacting as predictor variables). In addition, the association between pregnancy distress and child social-emotional development was only significant in the model with non-judging as a predictor variable social-emotional development. In addition, no associations were found between mindfulness facets and child internalizing and externalizing behavior problems.
In line with the hypotheses, better non-reacting skills in mothers during pregnancy were associated with more socialemotional capacities in children at 2 years of age. These findings were in line with earlier research that associated more maternal mindfulness during pregnancy with less infant selfregulation problems (Van den Heuvel et al., 2015) and negative social-emotional behavior (Braeken et al., 2017). The association between non-reacting and child social-emotional behavior in the current study was, against expectation, not mediated by pregnancy distress or postnatal symptoms of anxiety or depression. Thus, rather than maternal mental health, other mechanisms might play a role. It could be that non-reacting skills extend into parenthood. Non-reacting implies being aware of own emotional states without getting carried away by them, which is presumed to lead to a greater understanding of own emotions (Bishop et al., 2004). In a parenting context, it may enhance insight into the child's emotions as well. The capacity to empathize with the child might benefit parental coregulating competences and subsequently socialization of adequate emotion regulation (Bird et al., 2021). In addition, mothers may apply non-reacting skills in interaction with their children. Non-reactive parenting involves a delay of response towards emotional displays of the child, enabling a parent to select a supportive and appropriate coping strategy, as an alternative to excessive, overreactive, or automatic responses (Duncan et al., 2009;McKee et al., 2018). More dispositional mindfulness in parents has indeed previously been associated with more mindful parenting (Parent et al., 2016), which in turn was related to better emotional self-regulation capacities in children (Evans et al., 2020). Non-reacting could be particularly relevant in toddlerhood as parent-child conversations manifest and calmly elaborating on children's negative feelings may facilitate their emotional understanding (Laible, 2011).
In contrast to non-reacting, mindfulness facets acting with awareness and non-judging were unrelated to child socialemotional development. These findings were not in accordance with earlier studies where more prenatal mindfulness was related to less infant self-regulation problems (Van den Heuvel et al., 2015) and less negative social-emotional behavior (Braeken et al., 2017). In both studies, however, significant associations were demonstrated for only 1 out of 5 subscales of child social-emotional development. Neither study differentiated between facets of mindfulness. Results of the current study suggest that non-reacting has distinct features, at least when measured during pregnancy. In contrast to non-reacting, mindfulness facets acting with awareness and non-judging were associated with maternal mental health variables (as in line with earlier studies, e.g., Braeken et al., 2017;Nyklíček et al., 2018), both during pregnancy and postpartum, indicating that these facets are more closely linked to mood. This finding is in line with a meta-analysis of Carpenter et al. (2019) that demonstrated the strongest correlations between these facets and negative affect (e.g., depression, anxiety), which were largely unrelated to socioemotional development in the present study.
As hypothesized, pregnancy distress mediated the association between non-judging and child externalizing behavior problems. Mothers with more non-judging skills during pregnancy reported less pregnancy distress, and in turn, fewer externalizing behavior problems in their children at 3 years of age. This finding was in line with earlier studies that associated more maternal mindfulness with less prenatal distress (e.g., Braeken et al., 2017), and less prenatal distress with less externalizing problems in children (Wesselhoeft et al., 2021). Adapting a non-judgmental attitude implies compassion towards oneself and one's experience, which is likely to protect against self-criticism, negative or disruptive thoughts, and worrying or ruminating, for instance about the physical changes inherent to pregnancy or upcoming parenthood (Bishop et al., 2004;Hollis-Walker & Colosimo, 2011). Subsequently, less maternal distress during pregnancy can safeguard healthy development of the child. Different mechanisms might be involved. For example, there is evidence that elevated stress during pregnancy is a risk factor for adverse birth outcomes (e.g., low birth weight; Hobel et al., 2008), which in turn have been associated with child behavior problems (Van Os et al., 2001). Alternatively, since stress during pregnancy has been related to postpartum stress (Štěpáníková et al., 2020), parental stress may negatively affect parenting quality and subsequently child behavior (Choi & Becher, 2019). It should be noted however that the number of tests conducted in the current study was relatively high and we cannot rule out that this finding was an incidental one. In addition, when not being included simultaneously in the analysis, nonjudging and maternal distress were both unrelated to externalizing problems, further suggesting that the present mediation effect may have been spurious. Therefore, caution is needed when interpreting this result.
Against expectations, mindfulness facets during pregnancy proved unrelated to internalizing and externalizing child behavior. Although previous research associated more maternal mindfulness with fewer internalizing and externalizing behavior problems (Bird et al., 2021;Parent et al., 2016), it is difficult to compare our findings since those studies did not assess maternal mindfulness during pregnancy and included children that were older than the children in the current study. Perhaps, such relations are not yet visible during the prenatal period. Alternatively, it might be that child behavior problems manifest more clearly at a later developmental stage. Hence, including our young age group might explain the absence of significant correlations in the present study. The fact that also mindfulness facet non-reacting was unrelated to child behavior problems seems to reflect a unique effect of maternal non-reacting skills on child social-emotional development. As proposed earlier, this might be due to mothers' ability to empathize with the child and adequate modeling and socialization of adaptive ways to regulate emotions, both very important aspects of social-emotional development.

Limitations and Future Research
The present study has a number of strengths. The first is the longitudinal design of the study, including assessments during different trimesters of pregnancy and the postpartum phase, as well as a follow-up at 2 and 3 years after childbirth. Even though the current study includes a subsample of the original cohort, the current study did not show any differences (with clinically relevant effect sizes) in demographic and obstetric variables compared to the original cohort. Second, the study examined facets of mindfulness-rather than mindfulness in general-and demonstrated these facets to be differentially associated with child social-emotional development. Third, we studied distress directly related to pregnancy and upcoming parenthood separately from symptoms of anxiety and depression, whereas previous research generally assesses anxiety and depression as representations of distress. Despite these strengths, there are also some limitations. First, all questionnaires were completed by mother as the only respondent, but using multiple informants (e.g., fathers) for the assessment of child behavior preferable, particularly in the case of depressive mothers (Ordway, 2011). Second, while different predictors and outcome variables were included, multiple tests had to be conducted. This can increase the chance of a false-positive finding (Ranganathan et al., 2016) and therefore, cautiousness is required when interpreting the significant mediating effect found in the current study. Third, the current participants were more highly educated compared and fewer participating women had an ethnic minority background compared to the general population in the Netherlands. This may limit generalizability to the general population. Fourth, the a priori power analysis that was conducted in the software program G*Power (Faul et al., 2007) did not take testing for indirect effects into account and the required sample size of n = 114 might not have been sufficient. Finally, the design of the study does not enable conclusions on causality of the associations found, nor can it be determined whether effects of mindfulness were specific for the pregnancy or if similar relations would have been found when mindfulness was measured in another period (i.e., before or after pregnancy).
The current study contributed to the literature by examining mindfulness and mental health of women during pregnancy and postpartum, in relation to early child development and behavior. Non-reacting skills in mothers during pregnancy may be a facet of mindfulness that holds favorable implications for child social-emotional development, regardless of mothers' mental health during pregnancy and the postpartum period. Future research is necessary to examine whether enhancing maternal non-reacting skills would be valuable, either for preventative purposes or in families at risk. Moreover, it would be interesting to examine whether non-reacting skills during pregnancy extend into later parenting, since this may provide a window for enhancement of positive parental behavior in the earliest stage of parenthood. Additionally, further research is needed to study the different facets of mindfulness in relation to early child development, as well as other potential mechanisms such as mindful parenting. Including postpartum measures of dispositional mindfulness or mindful parenting would allow for examining potential mediating and direct effects, as well as the stability of mindfulness throughout the perinatal period.
Acknowledgements The authors thank all midwife practices for their support in recruitment and all women for their participation in the study. In addition, the authors thank all the researchers for their contribution to the data collection.

Author Contribution
All authors contributed to the study conception and design. Material preparation and analysis were performed by Noor de Waal, Myrthe Boekhorst, and Ivan Nyklíček. The first draft of the manuscript was written by Noor de Waal and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding This study was funded by Stichting de Weijerhorst.
Data Availability Data will be made available on reasonable request.

Declarations
Ethics Approval This study was performed in line with the principles of the Declaration of Helsinki. Approval of the HAPPY study was granted by the Ethical Board of Tilburg University, and has been evaluated by the Medical Ethical Committee of the Máxima Medical Centre Veldhoven. The HAPPY follow study was approved by the Medical Ethics Committee of the Máxima Medical Centre (protocol number NL54558.015.15).

Conflict of Interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.