Introduction

Research has shown that there is a significant decrease in positive mood during adolescence (Weinstein et al. 2007), as well as a dramatic increase in depressive feelings around the time of puberty (Graber and Sontag 2009). Dutch figures indicate that 17 % of the 12- to 17-year-old adolescents experience symptoms of depression or anxiety (Centraal Bureau voor de Statistiek [CBS; Statistics Netherlands] 2003). Symptoms of depression and anxiety during adolescence seem to predict psychological malfunctioning later in life. For example, Kim-Cohen et al. (2003) found that 52.3 % of adults with a depressive disorder (major depressive disorder or dysthymia) at age 26 and 54.5 % of adults with an anxiety disorder at age 26 had received their diagnosis before the age of 15. These findings indicate that the development of internalizing problems in adolescents is a major problem (Abela and Hankin 2008; Meijer et al. 2006).

Many topics have been studied in relation to internalizing problems among adolescents, including the longstanding assumption that parenting practices play a major role in the development and maintenance of psychological problems in children and adolescents (McLeod et al. 2007b). Traditionally, different parenting dimensions have been defined, including responsiveness (also called warmth, support, or connectedness), control (in which the distinction can be made between behavioral control and psychological control), and autonomy support or autonomy granting (e.g., Barber 1996; Darling and Steinberg 1993; McLeod et al. 2007a, b). Studies on the associations between these traditional parenting dimensions and adolescent psychological outcomes have shown that responsiveness and autonomy granting are related to lower levels of internalizing problems (McLeod et al. 2007a, b; Reitz et al. 2006). In contrast, higher levels of parental control seem to be related to higher levels of internalizing problems (McLeod et al. 2007a, b; Reitz et al. 2006).

Many years ago, a new approach to parenting, called “mindful parenting” (Kabat-Zinn and Kabat-Zinn 1997), has been proposed. This refers to parenting that conforms to the principles of mindfulness. Mindfulness is often explained as “paying attention in a particular way: on purpose, in the present moment, and nonjudgmentally” (Kabat-Zinn 1994, p. 4). Various studies have suggested that mindfulness-based interventions, such as mindfulness-based stress reduction (MBSR) or mindfulness-based cognitive therapy (MBCT), are related to a decrease in internalizing symptoms and stress in adolescents and adults (for reviews, see Baer 2003; Fjorback et al. 2011; Keng et al. 2011).

Mindful parenting refers to parenting in which the practices and concepts of mindfulness are integrated into parents’ thoughts, feelings, and behaviors (Coatsworth et al. 2010). Mindful parenting is reflected in the way parents bring an attitude of compassion, acceptance, and kindness into the interactions with their children and in the way parents are fully present during these parent–child interactions (Bögels and Restifo 2013; Coatsworth et al. 2010; Kabat-Zinn and Kabat-Zinn 1997). Based on the concept of psychological mindfulness, mindfulness based interventions, and parenting studies, Duncan et al. (2009a, b) have proposed five different dimensions of mindful parenting: (1) listening with full attention to the child; (2) non-judgmental acceptance of the self and the child; (3) emotional awareness of the self and the child; (4) self-regulation in the parenting relationship, and (5) compassion for the self and the child (for an extensive illustration of these dimensions, see Duncan et al. 2009a). As mindful parenting is based on mindfulness as the underlying theoretical approach, differences between mindful parenting and the traditional parenting dimensions might exist. For example, listening with full attention not only refers to simply hearing words that are said, but also indicates being sensitive to the content of the conversation as well as to the child’s tone of voice, body language, facial expressions and so on. Using these cues should lead to a better detection of the child’s needs or intended meaning of the message (Duncan et al. 2009a). Other dimensions of mindful parenting may show somewhat more overlap with the traditional parenting dimensions. Responsiveness, for example, seems to be related to both emotional awareness of the child and to compassion for the child. Empirical data on the association between these mindful parenting dimensions and the traditional parenting dimensions, however, is not yet available.

Based on the findings that parenting and mindfulness are related to internalizing problems in adolescents and young adults, researchers have postulated that mindful parenting might also be related to internalizing problems in adolescents (Duncan et al. 2009a, b). Although this has not been empirically tested yet, it is possible that mindful parenting improves both the quality of parenting and the parent–child relationship, which in turn might positively affect adolescents’ psychological functioning (i.e., reduce internalizing problems). Based on these hypotheses, multiple mindful parenting programs or interventions have been developed and studied among various clinical and non-clinical child- and adolescent samples (e.g., Altmaier and Maloney 2007; Bögels and Restifo 2013; Bögels et al. 2013; Coatsworth et al. 2010; Duncan et al. 2009b; Van de Weijer-Bergsma et al. 2012; Van der Oord et al. 2012; Singh et al. 2006, 2007). Most of these programs are based on MBSR and MBCT and often include themes such as self-regulation, forgiveness, kindness, and compassion, and practices such as meditation, yoga, attention-training to focus on everyday activities, and so on (for a detailed description of one of such mindful parenting programs, see Bögels and Restifo 2013).

Studies on the effectiveness of mindful parenting intervention programs provide preliminary support for the assumption that mindful parenting might be related to adolescents’ psychological functioning (Harnett and Dawe 2012). After completing a mindful parenting training program, parents reported they are more aware of how their moods affect how they react and that they think more often before they react in family situations (Coatsworth et al. 2010; Duncan et al. 2009b). In addition, mothers showed an improvement in their anger management and discipline consistency and expressed more positive and fewer negative emotions in their interactions with their adolescent children after the mindful parenting intervention compared to before the intervention (Coatsworth et al. 2010). Furthermore, parents reported less parenting stress and higher well-being after the intervention (Coatsworth et al. 2010; Duncan et al. 2009b; Singh et al. 2007; Van der Oord et al. 2012). As earlier research shows that parental well-being is related to adolescents’ well-being (Shek 2000), there might be an indirect effect of mindful parenting on adolescents’ well-being via parental well-being. However, the direct effects of mindful parenting on adolescents’ psychological functioning are less clear. Singh et al. (2006, 2007) found decreases in aggressive behaviors, compliance, and self-injury in children with a developmental disability after a mindful parenting program, but these studies consisted of small samples (i.e., 3 vs. 4 mother–child dyads). Van de Weijer-Bergsma et al. (2012) also found positive effects among adolescents with ADHD, but as these adolescents also received a mindfulness training themselves, it is impossible to determine whether these effects are the result of the mindful parenting program or not. However, in a recent waiting-list study of Bögels et al. (2013), promising effects were found. This mindful parenting course among parents in mental health care resulted in fewer internalizing and externalizing psychopathology symptoms for the parents themselves as well as for the targeted child. Other studies found mixed results (Van der Oord et al. 2012) or no direct effects (Altmaier and Maloney 2007).

Contrary to existing studies on mindful parenting, the current study measures mindful parenting in the normal population, apart from any intervention. Next, the direct association between these mindful parenting dimensions and adolescents’ internalizing problems, including symptoms of depression and anxiety, will be examined while controlling for the traditional parenting dimensions, to test whether mindful parenting offers a unique contribution to the variance in adolescent internalizing problems above and beyond the traditional parenting dimensions. As previous studies have shown that the psychopathology of the parent is also a predictor of internalizing problems in children or adolescents (e.g., Mesman and Koot 2000) we will also control for symptoms of depression and anxiety of the parent. It is expected that the different mindful parenting dimensions will all be negatively associated with adolescents’ depressive- and anxiety symptoms. Moreover, it is expected that the associations between mindful parenting and adolescent outcomes will remain significant when the traditional parenting dimensions are taken into account because the literature about mindful parenting suggests that mindful parenting is a somewhat different approach to parenting.

Method

Participants

Adolescents

Of the 901 adolescents in the sample, 53.2 % were boys and 46.8 % were girls. On average, participants were 13.4 years old (SD = .60; range 12–15). Most of the adolescents (82.0 %) lived with their biological mother and father, 7.9 % lived with their biological mother, and 3.7 % lived with their biological mother and stepfather. The remaining 6.4 % of the adolescents lived alternately with their mother or father or lived with their adoptive parents or in a boarding school. Almost all children were of Dutch descent (96.8 %). The education of the adolescents was distributed as follows: preparatory middle-level applied education, 6.2 %; a combination of preparatory middle-level applied education and higher general secondary education, 2.9 %; higher general secondary education, 35.2 %; a combination of higher general secondary education and pre-university education, 11.7 %; and pre-university education, 37.0 %. The remaining 7.0 % of the adolescents received a combination of all previous education levels. Compared to the average distribution in the Netherlands, adolescents from higher- or pre-university education were overrepresented in this sample, as in 2010–2011 approximately 34 % of Dutch adolescents received higher- or pre-universal education (CBS 2011), compared to 83.7 % in the current sample.

Parents

The 901 parents who participated were on average 45.2 years old (SD = 4.0; range 32–69). Most of the questionnaires (94.2 %) were completed by the biological mother, 4.6 % by the biological father, and 1.2 % by another caretaker. Of the 901 participating families, 86.8 % were intact (i.e., the parent was married or lived with a (new) partner), 11.6 % of the families were single-parent families, and 1.6 % of the parents reported something else. Ninety-five percent of the parents were of Dutch nationality, while the remaining parents were from different European and non-European countries. Parents’ educational attainment was distributed as follows: lower education, 14.7 %; intermediate education, 37.4 %; and higher education, 44.5 %. The remaining 3.4 % reported something different or left this question blank. There is an overrepresentation of parents with higher education in this sample as the average percentage of people with higher education in the Netherlands is about 28 % (CBS 2011).

Procedure

The current study was part of a study on the effectiveness of a school-based universal depression prevention program for young adolescents (Tak et al. 2012), in which 1,390 adolescents from nine secondary schools were asked to participate. The questionnaires that were completed in this study were part of a larger 45-min test battery at wave 1 (pre-intervention) that the adolescents completed at school during class. After the adolescents completed the questionnaires, the (biological) mother was also asked to complete a questionnaire as asking both parents to complete the questionnaire seemed too demanding and mothers in the Netherlands, on average, spend more time with their adolescents than fathers do (Dubas and Gerris 2002). If it was not possible for the (biological) mother to complete the questionnaire, it could also be filled out by the father or another caretaker. The parents were contacted through a letter that included an Internet link, which connected them to an online version of the questionnaire. After 2–3 weeks, the parents who had not yet responded were sent a reminder letter that included a paper version of the questionnaire. If parents still had not completed the questionnaire 2 weeks after the reminder, they were contacted by phone to ask them to complete the questionnaire. Ten gift certificates of 25 Euros were raffled among all parents who completed the questionnaire.

Of the initial nine participating schools, eight schools gave permission to approach the parents of the adolescents with the request to complete the parent questionnaire, which equated to the parents of 1,251 of 1,390 adolescents. Of these 1,251 adolescents, 19 were excluded from the sample because their parents did not give permission for participation, leaving 1,232 adolescents whose parents were contacted to complete the questionnaire. Of these 1,232 parents, 916 parents responded (74.35 %). For several reasons (e.g., missing personal information of the parent or the adolescent), the questionnaires of 901 parents and adolescents could be successfully matched, which is 73.1 % of the parents who were initially contacted. This sample of 901 parent–adolescent dyads is the final sample that was used in the analyses.

No significant differences in age or gender were found between the adolescents of the final sample (i.e., data of 901 adolescents–parent dyads) and the adolescents that were not included in the final sample (e.g., because they or their parents did not complete the questionnaires) [age: t(1,346) = 1.26, p = .208; gender: χ 2(1) = 0.37, p = .545]. However, it was found that adolescents outside the final sample reported higher levels of depressive symptoms (M = 7.99) than the adolescents within the final sample [M = 7.13; t(1,342) = 2.64, p = .008], as well as higher levels of symptoms of anxiety (M = 7.30), compared to the adolescents that were included in the final sample [M = 6.53; t (1,339) = 2.40, p = .017].

Measures

Depressive Symptoms

Adolescents

To measure adolescents’ depressive symptoms, a Dutch version of the children’s depression inventory (CDI) was used, which is a self-report questionnaire for children and adolescents (Kovacs 1992; Dutch version Timbremont and Braet 2002). The CDI contains 27 items; each item consists of three different statements (e.g., “I’m sad sometimes”, “I’m often sad”, “I’m sad all the time”). Respondents had to choose which statement matches their feelings best during the past 2 weeks. Items were scored as 0 (absence of symptom), 1 (mild symptom), or 2 (clear symptom), meaning that higher scores on the CDI indicated more depressive symptoms. Both the original measure and the Dutch version have good psychometric qualities (De Cuyper et al. 2004; Saylor et al. 1984). Cronbach’s alpha in the current sample was .83.

Parents

Parental depressive symptoms were measured with the Beck’s Depression Inventory-II (BDI-II; Beck et al. 1996), which is a widely used self-report questionnaire measuring depressive symptoms in adults. The BDI-II consists of 21 items with four different statements each. The items correspond to the symptoms of depression as listed in the DSM-IV. Participants had to choose the statement that best described their feelings during the past 2 weeks (i.e., “I’m not sad”, “I’m sad most of the time”, “I’m sad all the time”, “I’m so sad that I can’t stand it”). Each item was scored from 0 (absence of symptom) to 3 (most severe symptom), meaning that higher scores on the BDI indicated more depressive symptoms. In this study, a Dutch version of the BDI-II was used, which has been shown to have good psychometric properties (van der Does 2002). Cronbach’s alpha in the current sample was .88.

Anxiety Symptoms

Adolescents

To measure the level of general anxiety in adolescents, the Revised Children’s Manifest Anxiety Scale (RCMAS) was used (Reynolds and Richmond 1978). This widely used self-report scale consists of 28 items measuring anxiety (e.g., “I have trouble making up my mind” or “I am afraid of a lot of things”). Positively formulated items were reversed coded so that all items are rated on a dichotomous yes (1)/no (0) scale, with higher scores on the RCMAS indicating higher levels of anxiety. The RCMAS has been shown to be reliable (Reynolds and Richmond 1978). Cronbach’s alpha in our sample was .88.

Parents

To measure anxiety in the parents, a Dutch version of the State-Trait Anxiety Inventory (STAI-DY) was used (van der Ploeg et al. 1980), which is a self-report questionnaire suitable for adults. The STAI consists of two different subscales: one measuring state anxiety (20 items) and the other subscale measuring trait anxiety (20 items). As only the general levels of parents’ anxiety were relevant in the current study, only the 20 items of the trait scale were used, on which parents responded how they feel in general (e.g., “I feel nervous”, “I find myself worrying about something”). Positively formulated items were reversed coded, so that all items were rated on a 4-point Likert scale, ranging from 1 (almost never) to 4 (almost always), meaning that higher scores indicate higher levels of trait anxiety. The Dutch version has shown good internal reliability and validity (van der Ploeg et al. 1980). Cronbach’s alpha in the current sample was .93.

Traditional Parenting Dimensions

The following scales involve subscales from several parenting questionnaires that have been composed, translated, and adapted by Soenens and colleagues to be suitable for a parenting report (Soenens et al. 2006, 2007). For reasons of parsimony, we refer to Soenens et al. (2006, 2007) for the original scales and information on the validity and reliability (which has been shown to be good). All items of these scales were rated on a 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Negatively formulated items were reversed coded so that a higher score on a subscale reflected higher levels of that parenting dimension. The first scale, responsiveness, was measured using seven items (e.g., “I give my son/daughter a lot of care and attention”). Cronbach’s alpha in the current sample was .79. The second scale, behavioral control, was measured using 16 items tapping parents’ expectations for their child’s behavior (e.g., “I expect my son/daughter to behave in a certain way”) and parental monitoring of the child’s behavior (e.g., “I make an effort to know who my son’s/daughter’s friends are and where he/she spends his/her time”). Cronbach’s alpha was .79. The third scale, psychological control, consisted of eight items (e.g., “I’m trying to change how my son/daughter feels or thinks about things”). Cronbach’s alpha was .67. The fourth and last scale used to measure general parenting, autonomy support, consisted of seven items (e.g., “I allow my son/daughter to choose for him-/herself what to do, whenever possible”). Cronbach’s alpha was .60.

Mindful Parenting

To measure mindful parenting, the Interpersonal Mindfulness in Parenting scale (IM-P) was used, which is a parent self-report questionnaire developed by Duncan (2007). The IM-P is, to our knowledge, the first and only questionnaire that directly measures mindful parenting and not merely mindfulness among adults or parents. The original IM-P has been developed to measure the five dimensions of mindful parenting as described earlier by Duncan et al. (2009a). In the present study, however, a Dutch version of the IM-P was used (de Bruin et al. 2012), of which the translation has been back translated to English by de Bruin and colleagues and approved by Duncan. As both the original version of the IM-P as well as the Dutch translation are still under construction, de Bruin et al. (2012) conducted a validation study on the Dutch version of the IM-P, which included amongst other samples a sample with the mothers of the current study. The results supported a factor structure with six mindful parenting dimensions, which is slightly different than proposed as some of the items referring to the parent and the items referring to the child loaded on different factors in these Dutch samples (for a more detailed description and a comparison with the original IM-P factor structure, see de Bruin et al. 2012). The Dutch version of the IM-P showed satisfactory validity and reliability, except for the subscale emotional awareness of the self (de Bruin et al. 2012). However, to stay in line with the results of de Bruin and colleagues that are based on multiple samples, we decided to use all six dimensions in the current study.

All 29 items of the IM-P were rated on a 5-point Likert scale, ranging from 1 (never true) to 5 (always true). Negatively formulated items were reversed coded so that a higher score on a subscale reflected higher levels of that particular mindful parenting dimension. The first subscale is listening with full attention, which consisted of five items (e.g., “I pay close attention to my child when we are spending time together”). Cronbach’s alpha in the current sample was .83. The second subscale, compassion for the child, consisted of six items (e.g., “I am kind to my child when he/she is upset”). Cronbach’s alpha was .79. The third subscale, non-judgmental acceptance of parental functioning, also consisted of six items (e.g., “When things I try to do as a parent do not work out, I can accept them and move on”). Cronbach’s alpha was .72. The fourth subscale is emotional non-reactivity in parenting, which consisted of five items (e.g., “I often react too quickly to what my child says or does” (reversed coded)). Cronbach’s alpha was .74. The fifth subscale, emotional awareness of the child, consisted of three items (e.g., “It is easy for me to tell when my child is worried about something”). Cronbach’s alpha was .76. The sixth and final subscale was emotional awareness of self, which consisted of four items (e.g., “When I’m upset with my child, I notice how I am feeling before I take action”). Cronbach’s alpha was .54, which is rather low but in accordance with earlier findings of de Bruin et al. (2012).

Statistical Analyses

To examine whether mindful parenting is associated with adolescents’ symptoms of depression and anxiety when controlled for the traditional parenting dimensions, we conducted hierarchical regression analyses. Since the adolescents are ‘nested’ within 8 different schools, we first performed a MANOVA analysis to test for significant differences between schools with regard to adolescents’ symptoms of depression and anxiety. Based on these findings we decided to include the factor school (dummy coded) in the first step of our regression analyses, in order to control for these school-related differences. We also controlled for adolescents’ age and gender, and parental symptoms of depression and anxiety in step 1. Next, we included the four different traditional parenting dimensions in the analyses (step 2). The last step consisted of the six different mindful parenting dimensions (step 3).

Results

Descriptives

The descriptive statistics of the study variables are summarized in Table 1. Although the mean levels of depressive symptoms and anxiety symptoms in the current sample of adolescents are comparable to adolescents in other non-clinical samples (Timbremont and Braet 2001; Muris et al. 2002), the relatively low scores indicate that, on average, these adolescents experience only few symptoms of anxiety and depression [e.g., a CDI cut-off score of 19 or higher is used as an indication for clinical depression (Timbremont and Braet 2001), while the mean sum score in the current sample is 7.12. In addition, the possible sum scores of the RCMAS range from 0 to 28, while the mean sum score in the current sample is 6.53]. These descriptive statistics also show relatively high scores on parents’ self-reported levels of the different mindful parenting dimensions. In addition, t tests for independent samples showed that girls reported significantly higher levels of depressive symptoms than boys [girls M = 7.89, SD = 5.50; boys M = 6.43, SD = 4.92; t(897) = 4.20, p < .001], and higher levels of anxiety symptoms [girls M = 8.09, SD = 5.88; boys M = 5.14, SD = 4.65; t(893) = 8.37, p < .001]. Adolescents’ age was not significantly related to their levels of depressive symptoms (r = .001, n.s.) or anxiety symptoms (r = −.027, n.s.).

Table 1 Descriptive statistics of the study variables
Full size table

To take the nested structure of our data into account, we tested whether adolescents’ levels of depressive- and anxiety symptoms were related to their school. The result of a MANOVA with school as independent variable and adolescents’ symptoms of depression and anxiety as dependent variables showed to be significant [F(14,1,772) = 2.37, p < .003]. Based on this finding we decided to include school as a control variable in our hierarchical regression analyses.

Correlations Among Model Variables

Table 2 depicts the Pearson correlations among all the model variables. As expected, adolescents’ depressive symptoms and anxiety were highly correlated with each other (r = .77), as was the case for parental depressive symptoms and anxiety (r = .73). Parental symptoms were only weakly correlated with adolescents’ symptoms (r = .10 to r = .12). Furthermore, the traditional- and mindful parenting dimensions were not or only weakly associated with adolescents’ symptoms of depression and anxiety (ranging from r = .01 to r = −.18). Parental symptoms of depression and anxiety, however, were overall moderately associated with the different traditional- and mindful parenting dimensions (ranging from r = −.06 to r = −.41), indicating that higher levels of parental internalizing problems were, in general, associated with lower levels of mindful parenting as well as lower levels of responsiveness, behavioral control and autonomy support, and higher levels of psychological control.

Table 2 Pearson correlations between the study variables
Full size table

Looking at the correlations between the different mindful parenting dimensions, the results show that they are, overall, moderately correlated with each other (ranging from r = .27 to r = .46). The same conclusion is applicable to the correlations among the traditional parenting variables (ranging from r = .08 to r = −.43). The correlations between the mindful- and traditional parenting dimensions are more diverse. Some dimensions are not correlated with each other (e.g., behavioral control is unrelated to emotional non-reactivity in parental functioning and to non-judgmental acceptance of parental functioning), but others are higher than the intra-correlations as mentioned before (e.g., r = .58 between responsiveness and compassion for the child) (see Table 2).

Hierarchical Regression Analyses

The two hierarchical regression analyses (Table 3) show that adding the traditional parenting dimensions (step 2) in the prediction of adolescents’ symptoms of depression or anxiety, after controlling for adolescents’ age, gender, school, and parental symptoms of depression and anxiety (step 1), resulted in a non-significant R 2 Change. These results indicate that the traditional parenting dimensions did not explain any additional variance of adolescents’ symptoms of depression and anxiety when relevant control variables were accounted for.

Table 3 Results hierarchical regression analyses
Full size table

Most important for our research question was the inclusion of the mindful parenting dimensions in the last step of these regression analyses (step 3). The results showed a significant R 2 Change, indicating that the different mindful parenting dimensions together did explain additional variance of adolescents’ symptoms of depression and anxiety, when controlling for the traditional parenting dimensions and other control variables. Looking at the mindful parenting dimensions separately, results showed that only non-judgmental acceptance of parental functioning was significantly associated with adolescents’ symptoms of depression (β = −.14, p = .002) and anxiety (β = −.14, p = .002). These results indicate that higher levels of non-judgmental acceptance of parental functioning as reported by the parents were related to lower levels of depressive and anxiety symptoms based on adolescents’ self-report.

Discussion

This study examined the associations between six mindful parenting dimensions and adolescents’ internalizing problems, controlling for traditional parenting dimensions (i.e., responsiveness, psychological and behavioral control, and autonomy support). Our findings revealed that overall mindful parenting was associated with adolescents’ symptoms of depression and anxiety while controlling for the traditional parenting dimensions. However, of all mindful parenting dimensions (i.e., listening with full attention, compassion for the child, non-judgmental acceptance of parental functioning, emotional non-reactivity in parental functioning, emotional awareness of the child, and emotional awareness of self), only non-judgmental acceptance of parental functioning was associated with lower levels of adolescents’ internalizing problems.

The significant association between the mindful parenting dimension non-judgmental acceptance of parental functioning and adolescents’ psychological functioning is meaningful because it indicates that parents who report higher levels of non-judgmental acceptance of their own functioning as a parent (i.e., parents who report being less hard on themselves, less self-blaming regarding their parenting) are more likely to have children who report fewer symptoms of depression and anxiety. Non-judgmental acceptance of parental functioning shows strong similarities with the concept of self-compassion, which entails being kind and understanding toward oneself, perceiving one’s experiences as part of the larger human failure, and keeping painful feelings and thoughts in mindful awareness as opposed to over-identifying with them (Neff 2003; Neff et al. 2007). Not surprisingly, the concepts of self-compassion and mindfulness are assumed to be closely related to each other (see Neff 2003 for a more comprehensive description). In addition, higher levels of self-compassion have been found to be associated with lower levels of depression and anxiety in (young) adults as well as adolescents (Neff et al. 2007; Neff and McGehee 2010). It is possible that, by displaying higher levels of non-judgmental acceptance of parental functioning (i.e., higher levels of “self-compassion of the parent”), parents instill higher levels of self-compassion in adolescents, such as through modeling or imitation (e.g., Bandura 1986).

Since our data are cross-sectional, we cannot assume any causality. We cannot rule out that the direction of affect is not from mindful parenting to adolescents’ adjustment, but rather that adolescent adjustment affects mindful parenting. For example, parents report higher levels of non-judgmental acceptance of parental functioning when their children have fewer psychological problems because it might be easier for parents to be non-judgmental toward their own functioning as a parent when their children do not experience internalizing problems. Future studies should explore the causality of this association.

In addition, although adolescents’ internalizing problems were only associated with one of the mindful parenting dimensions, our findings based on the correlations showed that parental internalizing problems were associated with all six dimensions of mindful parenting, indicating that higher levels of parental symptoms of depression and anxiety were associated with lower levels of mindful parenting. Duncan et al. (2009a) suggested an effect of mindful parenting on parental well-being, yet it is also possible that it is easier for parents with lower levels of depressive and anxiety symptoms to be more mindful in their parenting. In accordance with this assumption are the findings of Parent et al. (2010) on general parenting, who showed that higher levels of depressive symptoms among parents were related to more negative parenting behaviors and less positive parenting behaviors. However, they also showed that lower levels of depressive symptoms were related to higher levels of mindfulness among these parents. This is in line with a study by Neece (2013) in which parents received a mindfulness training to increase their own levels of mindfulness. Although this training did not focus directly on parenting practices or child outcomes, it has been found that this training not only reduced parental stress and increased parental well-being, but also reduced ADHD-symptoms in their children (Neece 2013). Thus, it might be that the association between parental psychological functioning and mindful parenting, as found in the current study, is also mediated by higher levels of mindfulness among parents.

Besides the somewhat unexpected results on the associations between most mindful parenting dimensions and adolescents’ internalizing problems, it should be noted that the overall relation between all (traditional and mindful) parenting dimensions and adolescents’ internalizing problems appeared to be very weak, and various expected associations were not found. For example, no association has been found between adolescents’ internalizing problems and the traditional parenting dimensions when control variables were taken into account. Several explanations are possible for the absence of some associations. First of all, several characteristics of the sample may have biased the results, such as an overrepresentation of adolescents and parents with higher education in our sample, relatively low prevalence of depressive symptoms and anxiety symptoms among parents and adolescents [which may be attributed to the high average educational levels of both parents and adolescents (De Graaf et al. 2012)], or to some kind of ‘participation bias’, supported by the finding that adolescents outside the final sample had higher levels of depressive- and anxiety symptoms than the adolescents within the current sample. It is possible that some associations will be found in samples that are more heterogeneous regarding the economic background and participants’ internalizing problems.

Second, the absence of some of the hypothesized associations may also be explained by the fact that parenting was measured through parents’ reports instead of adolescents’ reports on parenting. Research has shown that during adolescence, there is only weak agreement on parenting practices between adolescents’ reports (how adolescents perceive their parents’ parenting) and parents’ reports (how parents perceive their own parenting; Cohen and Rice 1997). It has been suggested that the way in which children perceive parenting behavior is more strongly related to their own behavior than the actual parenting that takes place or the parenting behavior as perceived by the parent (Cohen and Rice 1997). Indeed, studies in which parenting is reported on by adolescents show stronger associations between parenting and adolescents’ psychological functioning than the present study (Oldehinkel et al. 2006; Reitz et al. 2006).

Furthermore, parents in the current study reported quite high levels of mindful parenting, which may be the result of social desirable answers. It is also possible, however, that these parents (who generally will have less knowledge about mindfulness or mindful parenting as compared to parents participating in a mindful parenting program), have more difficulties to come up with a reliable assessment regarding their levels of mindful parenting. Future research should study the sensitivity and specificity of the IM-P questionnaire to measure mindfulness in the general population.

In addition to these explanations that can clarify the weakness or absence of some of the associations, the current study has also some limitations. It should be kept in mind that the generalization of the results is limited given that the majority of the parents in our sample were mothers (>93 %), which might not be representative for two parent households in which the father spends equally as much or even more time with the adolescent than the mother, and because parents and adolescents with higher or pre-universal education levels were overrepresented in this study. Another limitation of this study concerns the use of a cross-sectional design; thus, no causal conclusions can be drawn based on the results presented in this study. Previous research has suggested the influence of adolescents’ internalizing symptoms on mothers’ expressed emotions (Hale et al. 2011) as well as a bidirectional relationship between general parenting and adolescents’ psychological functioning (e.g., Reitz et al. 2006); these suggestions might very well apply to mindful parenting as well.

The finding that only one dimension of mindful parenting is related to adolescents’ internalizing problems after controlling for traditional parenting may be the result of the explanations and limitations described before. However, it is also possible that overall, contrary to our hypothesis, mindful parenting is not very important for the well-being of adolescents. Although earlier studies on mindful parenting interventions demonstrated some positive effects, it should be stressed that generally the designs of these studies were suboptimal with relatively small sample sizes (e.g., Singh et al. 2007), no randomization of conditions (e.g., Bögels et al. 2013) and these studies mostly focused on the effects on parent- or parenting-related constructs (e.g., lower parental stress or more self-awareness of the parent regarding their parenting), while the direct effects on adolescent psychological functioning were less convincing. Together with the current findings, one can question the effects of mindful parenting on adolescents’ internalizing problems. In addition, considering our finding that only the mindful parenting dimension non-judgmental acceptance of parental functioning (i.e., parental self-compassion) was significantly related to adolescents’ internalizing problems, it could be argued that these parenting programs might focus primarily on increasing parents’ self-compassion with regard to their parenting, and less on other mindful parenting practices such as listening with full attention to the child, and so on. For the clinical implementations of mindful parenting intervention programs, it is important that future research demonstrates the effective components of these (mindful) parenting programs in different clinical and non-clinical samples covering a broad range of socio-economic backgrounds, as well as whether these programs could have a positive direct effect on adolescents’ psychological functioning or not.

Despite several limitations, this study covers a relatively large sample in which mindful parenting is measured in the general population, apart from any intervention. In addition, the current study provides more insight into mindful parenting as a new approach to parenting, specifically with regard to the association with adolescents’ internalizing problems. The findings revealed that in the current sample only the mindful parenting dimension non-judgmental acceptance of parental functioning is related to adolescents’ internalizing problems, when controlled for traditional parenting dimensions. Since this dimension is closely related to the concept of self-compassion, the finding suggests that a non-judgmental and compassionate stance of parents towards their parenting may be one of the core elements of mindful parenting, and maybe also function as one of the ‘active ingredients’ of existing mindful parenting programs. Implications of these findings are that, in future parenting research, it is also relevant to take parental thoughts, feelings, and attitudes with regard to their own role as a parent into account when studying the association between parenting and adolescents’ internalizing problems. Moreover, future research should focus on the theoretical conceptualization of mindful parenting and its possible overlap with the traditional parenting dimensions, as well as on the possible direct causal effects of mindful parenting programs on adolescents’ internalizing problems.