Parenting Styles and Internalizing Symptoms in Adolescence: A Systematic Literature Review

A growing body of recent research has identified associations between various parenting practices and styles and internalizing problems among adolescents. However, the reported findings are inconsistent and the studies in question have been conducted from different theoretical backgrounds. The aim of this systematic review was to synthesize the literature on the association of parental socialization styles with depression, anxiety, and suicidal ideation. To this end, we conducted a systematic search of the PsycInfo, Scopus, Pubmed, and Web of Science databases, covering literature published from 2010 to 2019. The search was restricted to peer-reviewed studies in English or Spanish. The results show that parental warmth, behavioural control, and autonomy granting are inversely related to internalizing symptoms in adolescents. Conversely, psychological control and harsh control by parents are positively associated with adolescent anxiety, depression, and suicidal ideation. Although the associated effect sizes are only small or moderate, the results suggest that these variables should be taken into account when designing programmes aimed at promoting parenting styles conducive to the wellbeing of adolescents.


Introduction
There is a growing interest in the study of child rearing, largely due to the development of theoretical models such as attachment theory [1][2][3] and social learning theory [4], which emphasize the influence that parents have on children's development. Research on child rearing has particularly focused on the role played by general patterns of parental behaviour, referred to as parenting styles. Darling and Steinberg [5] consider parenting style to be a constellation of attitudes, goals, and patterns of child rearing that shape the emotional climate of the parent-child relationship and which remain constant across different life situations.
Building on the work of Baumrind [6] and her description of authoritative, authoritarian, and permissive parenting, Maccoby and Martin [7] proposed a model of parenting styles based on two broad orthogonal dimensions, demandingness and responsiveness, thus giving rise to a typology involving four potential parenting styles: authoritative or democratic (high demandingness and high responsiveness), authoritarian (high demandingness and low responsiveness), indulgent (low demandingness and high responsiveness), and uninvolved or neglectful (low demandingness and low responsiveness). According to Maccoby and Martin [7], the first of these dimensions refers to the limits and rules which parents set in order to regulate their child's behaviour, and to the demand that the child complies with them. The second dimension concerns the extent to which parents show affection, are involved with and accepting of their child's behaviour and feelings, and are sensitive to his or her needs. It must be noted that other theoretical approaches to parenting styles have highlighted the abovementioned orthogonal dimensions, also called acceptance/involvement and strictness/imposition [8,9].
While parenting styles refer to general patterns of parental behaviour, parental practices constitute more specific forms of parent-child interaction in specific situations [5]. For example, behavioural control refers to the ways in which parents monitor or supervise their child's behaviour to ensure that it is appropriate [10,11]. The limits and rules which parents set in this respect can have a positive impact on the child's psychosocial adjustment. However, harsh control, defined as physical and/or verbal punishment by parents, may undermine the child's adjustment [12]. As for psychological control, this refers to parents' attempts to manipulate their child's thoughts and feelings by inducing a sense of guilt or humiliation, or through emotional blackmail or overprotectiveness [13,14]. Another theme that has been studied within the framework of child rearing practices is autonomy granting, defined by McLeod et al. [15] as "parental encouragement of children's opinions and choices, acknowledgement of children's independent perspectives on issues, and solicitation of children's input on decisions and solutions of problems" (p. 162). Finally, parental warmth is described as "the sense of positive regard expressed by the parent toward the adolescent, pleasant interactions shared between parent and adolescent, or parental involvement in the adolescent's activities" [16] (p. 12).
Various authors have stated that, as parenting styles, parental practices can also be classified by the acceptance/involvement and strictness/imposition dimensions [17][18][19][20]. Therefore, behavioural control is characterized by a high acceptance/involvement and strictness/imposition, which fits with authoritative style. However, harsh control and psychological control are characterized by a low acceptance and a high strictness/imposition, related to authoritarian style [21]. Likewise, autonomy granting is defined by a high acceptance/involvement and low strictness/imposition, which are associated with the indulgent style [22]. Finally, parental warmth is related to a high acceptance/involvement, but it remains unclear whether this parental practice is characteristic of authoritative or indulgent style [19,22].
These theoretical models have formed the basis for numerous studies examining the role played by different parenting practices in relation to the psychological adjustment of children and adolescents. Therefore, for instance, various meta-analyses have found a negative correlation between parental warmth and internalizing symptoms such as anxiety [15,16,[23][24][25] and depression in young people [26,27]. As regards to harsh control and psychological control, both have been associated with more internalizing symptoms in children and adolescents [28], whereas behavioural control has been linked to fewer symptoms of this kind [12,29,30]. With respect to autonomy granting, several meta-analyses have found a negative correlation between this practice and children's internalizing symptoms [12,15,16]. In addition, some studies have reported an association between suicidal ideation and psychological control [31], as well as an inverse relationship between this kind of control and both maternal warmth [32][33][34] and autonomy granting [33], although no meta-analyses have so far examined these relationships.
With respect to the different types of parenting styles, various meta-analyses have concluded that authoritarian and neglectful parenting are associated with the presence of internalizing symptoms in children [12,23,26], probably due to the lack of acceptance/involvement which characterizes these parenting styles. It should be noted, however, that not all analyses have found a statistically significant correlation in this respect (e.g., [24]). As regards authoritative parenting, this has been associated with fewer internalizing symptoms and, in general, better psychological adjustment in children and adolescents [12,26,35]. Studies which have examined the relationship between parenting styles and suicide risk in young people have found that authoritarian parenting is associated with an increased risk of suicide [36][37][38], whereas authoritative parenting appears to be related to a lower risk [36,37].
Although different theoretical models consider that children's upbringing has an important impact on their subsequent psychological adjustment, several studies have found only a moderate effect size for this relationship [15,27,39]. This could be due to an overly simplistic conception of the relationship between child rearing and psychopathology, since there are multiple causes of the latter [12]. Consequently, these studies highlight the need to consider other variables which may moderate the relationship between parenting styles and internalizing symptoms, for example, genetic factors [15,27,39], parental gender [16,23,24,39], anxiety or depression in the father or mother [40], and the child's age and gender [15,23,27,39]. Furthermore, and as Darling and Steinberg [5] point out, parenting style itself is more likely to be a moderator than a predictor of children's psychosocial outcomes. It should also be borne in mind that the abovementioned studies have been conducted in western countries with participants of a medium socioeconomic status, factors which may also moderate the relationship between child rearing and internalizing symptoms [23].
Studies that have examined the relationship between parenting styles and internalizing symptoms in children and adolescents also suffer from a number of methodological limitations, such as the almost exclusive reliance on questionnaires, some of which lack convergent validity, and this highlights the importance of employing other assessment techniques [15,23,25]. In addition, the fact that the majority of studies are based on non-experimental cross-sectional designs makes it difficult to establish causal hypotheses, and hence there is a need for experimental and/or longitudinal studies in this field [25].
A further issue to consider is the lack of consensus in conceptualizing the different parenting style dimensions [12,25]. More specifically, although there is agreement with respect to the definition of the acceptance/involvement dimension, this is not the case for strictness/imposition [29]. A more precise definition of these dimensions is therefore necessary so that future studies may be based on a consistent set of criteria and constructs [23].
In light of the above, the goal of this systematic review is to synthesize the literature on the association of parental socialization styles and practices with anxiety, depression, and suicidal ideation in youth. In doing so, we aim to fill the gap in the literature regarding the relationship between different parenting style dimensions and practices and suicidal ideation in adolescents, while also examining the methods and characteristics of studies carried out over the past decade on the relationship between parenting styles and practices and depression, anxiety, and suicidal ideation.

Method
This systematic review was conducted in accordance with PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement guidelines [41].

Search Strategy
Systematic searches of the PsycInfo, Scopus, Pubmed, and Web of Science databases were conducted on 15 March 2019, covering literature published from 2010 to 2019. The search terms included 'parenting styles', 'parenting practices', 'parental socialization', 'anxiety', 'anxious', 'depressive', 'depression', and 'suicid*'. The asterisk (*) acts as a truncation symbol to detect various results composed of a single string of text: for example, 'suicid*' detects terms such as 'suicide' and 'suicidal'. Full search strategies are presented in Tables 1-4.

Study Selection
Two researchers (A.G. and J.A.) conducted the first screening of articles based on titles and abstracts. Inter-rater reliability was satisfactory, with a kappa coefficient of 0.93. Any disagreement was resolved by consensus. Eligible articles were then identified by screening full texts using previously agreed exclusion and inclusion criteria (see below). In this case, the kappa coefficient was 0.81, and any disagreements were once again resolved through consensus.

Eligibility Criteria
The inclusion criteria were (1) studies that assess anxiety, depression, and/or suicidal ideation in adolescents aged between 12 and 18 years, and which examine parenting styles or practices; (2) articles published in English or Spanish; and (3) studies that analyse the relationship between parenting styles or practices and anxiety, depression, and/or suicidal ideation in adolescents.
Accordingly, the exclusion criteria were (1) not having access to the full text (no full text); (2) publications in languages other than English or Spanish (language); (3) theoretical reviews, meta-analyses, systematic reviews, or studies in which no data were collected (article type); (4) studies whose participants were not adolescents aged between 12 and 18 years (sample); and (5) studies that did not examine the relationship between parenting styles or practices and anxiety, depression, and/or suicidal ideation in adolescents (content). As more than one of these exclusion criteria may apply to the same study, records were coded following the order in which the criteria are presented here (see Figure 1).

Eligibility Criteria
The inclusion criteria were (1) studies that assess anxiety, depression, and/or suicidal ideation in adolescents aged between 12 and 18 years, and which examine parenting styles or practices; (2) articles published in English or Spanish; and (3) studies that analyse the relationship between parenting styles or practices and anxiety, depression, and/or suicidal ideation in adolescents.
Accordingly, the exclusion criteria were (1) not having access to the full text (no full text); (2) publications in languages other than English or Spanish (language); (3) theoretical reviews, metaanalyses, systematic reviews, or studies in which no data were collected (article type); (4) studies whose participants were not adolescents aged between 12 and 18 years (sample); and (5) studies that did not examine the relationship between parenting styles or practices and anxiety, depression, and/or suicidal ideation in adolescents (content). As more than one of these exclusion criteria may apply to the same study, records were coded following the order in which the criteria are presented here (see Figure 1).

Data Extraction
All studies were coded using a detailed coding manual and the following data were extracted (where available) from each: country of study, study type/design, sample size/description, age of participants, gender, parenting assessment, depression assessment, anxiety assessment, suicidal ideation assessment, and main results.
In order to evaluate parenting practices, we considered the following practices: parental warmth, which refers to measures of parental care, acceptance, involvement, emotional support,

Data Extraction
All studies were coded using a detailed coding manual and the following data were extracted (where available) from each: country of study, study type/design, sample size/description, age of participants, gender, parenting assessment, depression assessment, anxiety assessment, suicidal ideation assessment, and main results.
In order to evaluate parenting practices, we considered the following practices: parental warmth, which refers to measures of parental care, acceptance, involvement, emotional support, concern, and understanding; psychological control, referring to parental practices that manipulate children's psychological experiences through imposition or conditional love; behavioural control, which refers to parental monitoring or supervision; harsh control, that is, the use of corporal punishment or severe punishment practices; and autonomy granting, referring to practices that explicitly foster children's autonomy. We also considered other practices, such as inconsistent discipline, negative parenting, or over-reactivity, which appear less frequently in the reviewed literature. Finally, we coded the four parenting styles (authoritative, authoritarian, indulgent, and neglectful).
Regarding the outcome variables, it should be noted that suicidal ideation was considered in a broad sense. Therefore, our conceptualization included the lifetime prevalence of suicide attempts, suicidal thoughts, self-harm, or self-destructive thoughts and behaviours.

Literature Search Results
The search strategy yielded 848 results, as shown in Figure 1. After removing duplicates, a first screening based on title and abstracts resulted in 167 articles. Finally, after a full-text read-through, 59 studies [31,36, were selected for inclusion (see Table 5).

Study Characteristics
Half of the included studies (N = 29; see Table 5) had been published since 2016. Overall, the studies reviewed had been conducted in 28 different countries, most notably the USA (33.3% of studies), followed by China (11.7%), and Mexico, Portugal, and Spain (each 5%). A cross-sectional design was used in 78% of studies, while the remaining 22% were longitudinal. Sample size ranged from 27 [65] to 71,891 [36], with a median of 445 participants. The proportion of females in the samples ranged between 0% [89] and 92.7% [88], with a median of 52%.
Regarding outcome variables, 66.1% of studies assessed depression, 27.1% anxiety, and 18.6% suicidal ideation. In approximately 15% of studies, the results for boys and girls were analysed separately, while in 34%, the results relating to parenting style were analysed separately for mothers and fathers. In addition to parenting, this review is also concerned with depression, anxiety, and suicidal tendencies in adolescents. Depression was most often assessed with the Clinical Diagnostic Interview (CDI; [107]; 10 studies), the Center for Epidemiologic Studies Depression Scale (CES-D; [108]; 10 studies), the Youth Self Report (YSR; [109]; 5 studies), the Symptom Checklist 90 Revised (SCL-90-R; [110]; 3 studies), the Diagnostic Interview Schedule for Children (DISC; [111]; 2 studies), the Beck Depression Inventory (BDI; [112]; 2 studies), and related measures (7 studies). Anxiety was most often measured with the Revised Child Anxiety and Depression Scale (RCADS; [113]; 2 studies), the Revised Children's Manifest Anxiety Scale (RCMAS; [114]; 2 studies), and related measures (12 studies). Finally, suicidal ideation was most often assessed either with measures created ad hoc (6 studies), or with selected items from standardized measurement instruments designed to assess other constructs (2 studies).

Association of Parenting Practices and Styles with Depression, Anxiety, and Suicidal Ideation in Adolescents
The results of our analysis showed an inverse relationship of moderate effect size between parental warmth and depressive symptoms in adolescents. Correlation coefficients ranged between −0.75, reported by Jelenova et al. [65] in a sample of adolescents with inflammatory bowel disease, and −0.03, with a median of −0.26. Regarding the inverse association between parental warmth and anxiety, the median correlation indicated a small effect size (−0.14), with coefficients ranging between −0.51 (once again, in the study by Jelenova et al. [65]) and a minimum of −0.14. The relationship between parental warmth and suicidal ideation also showed a small effect size (median correlation of −0.17), with coefficients ranging between −0.71 ([50], in a sample of adolescents who had assaulted their parents) and 0.03.
In general, there was a positive relationship with a small or moderate effect size (median 0.24) between parental psychological control and depression in adolescents, with correlation coefficients ranging between 0.48 [42] and 0.05. The effect size for the relationship between psychological control and anxiety was smaller (median 0.16), and in this case, coefficients ranged between 0.47 [78] and 0.03. Finally, the relationship between psychological control and suicidal ideation yielded a median correlation of 0.20, ranging between 0.41 [83] and −0.08.
Behavioural control showed an inverse association of a small effect size with depression in adolescents (median −0.02), with coefficients ranging between −0.53 ([88], in a sample of adolescents hospitalized for a suicide attempt) and −0.03. This parenting practice was only weakly associated with anxiety (median −0.03), with coefficients ranging between −0.19 [85] and 0.07. Finally, the relationship between behavioural control and suicidal ideation was almost null, with the highest coefficient being −0.12, in the study by Florenzano et al. [61].
Harsh control showed a positive association of a moderate effect size (median 0.25) with depression in adolescents, with coefficients ranging between 0.78 [65] and 0.14. The effect size was also moderate (median 0.34) for the relationship between this practice and anxiety, and in this case, coefficients ranged between 0.68 [65] and 0.24. An association of a moderate effect size was also found for the relationship between harsh control and suicidal ideation, although this is based on just two studies; the highest correlation was 0.28, in the study by Nunes and Mota [80].
Finally, autonomy granting showed a negative association of a moderate effect size with depression in adolescents (median −0.28), and a slightly weaker inverse relationship with anxiety (median −0.22) and suicidal ideation (median −0.22). Coefficients ranged between −0.44 [65] and −0.03 for the relationship between autonomy granting and depression, between −0.45 [65] and −0.08 for the association with anxiety, and between −0.35 [83] and −0.02 for the relationship with suicidal ideation.
Regarding parenting styles, indulgent parenting has been found to be associated with less suicidal ideation in adolescents [36], while authoritative parenting has been related to less depression [66,74,86,90]. Conversely, authoritarian [71,74] and negligent parenting styles [71,86] have been associated with more symptoms of depression in adolescents.     With respect to other parenting practices that have been examined in a smaller number of studies, positive associations of a moderate effect size have been reported between adolescent depression and inconsistent discipline [44], negative parenting [46], family dysfunction [49], a desultory parental style [65], and over-reactivity [91]. Conversely, a negative relationship has been observed between adolescent depression and effective parenting [76]. As regards anxiety in young people, a positive relationship has been found with respect to both family dysfunction [49] and overprotection [78]. Finally, one study reported a negative and moderate correlation between suicidal ideation in adolescents and parental imposition [31], while another found a positive association between internalizing behaviour in adolescent girls and parental over-involvement [77].

Discussion
The main goal of this study was to synthesize the literature on the association of parenting practices and styles with anxiety, depression, and suicidal ideation in adolescents. Specifically, we reviewed 59 studies that analyse the relationship between parenting and one or more outcome variables. The evidence base is largest with respect to depression, followed by anxiety, but there is considerably less research on the relationship between parenting and suicidal ideation in young people.
The fact that the reviewed studies were conducted in almost thirty different countries, with half of them being published since 2016, indicates that the relationship between parenting and internalizing problems in young people continues to be a topic of research interest. Despite this interest, however, some of the problems identified in previous meta-analyses have yet to be resolved. For example, and with respect to a shortcoming noted by Wood et al. [25] and Pinquart [12], the present review shows that there is still considerable heterogeneity in the conceptualization of different parenting styles and practices, making it difficult to draw firm conclusions from the results. This heterogeneity is illustrated by the fact that the studies reviewed used around 40 different instruments, with some being standardized, whilst others were developed ad hoc. There was also notable variation in the parenting practices and styles considered, even across studies that used the same measurement instrument. All these aspects hamper the generalization of results.
As regards the operationalization of outcome variables, the 39 studies which examined depression used 12 different instruments, most frequently the Children's Depression Inventory (CDI) and the Center for Epidemiologic Studies Depression Scale (CES-D). Anxiety was also operationalized in a variety of ways, and only two instruments, the Revised Child Anxiety and Depression Scale (RCADS) and the Revised Children's Manifest Anxiety Scale (RCMAS), were used in more than one study. With respect to suicidal ideation, it is noteworthy that instruments designed specifically to measure this construct, such as the Suicidal Ideation Questionnaire (SIQ, [115]) or Suicide Intent Scale (SIS, [116]), were rarely employed. Some studies used two items from the Youth Self-Report (YSR, [117]), namely "I deliberately try to hurt or kill myself" and "I think about killing myself", and one study used a single ad hoc item, "I've thought about ways of killing myself". It should also be noted, in relation to a point made earlier, that we found very few studies analysing the relationship between suicide and parental practices, and hence, in the present review, suicidal ideation was considered in a broad sense (lifetime prevalence of suicide attempts, suicidal thoughts, self-harm, or self-destructive thoughts and behaviours), including both passive and active ideation [118].
Although various meta-analyses have examined the relationship between parenting practices and depression [26,27,35], anxiety [15,23,25,39,40], or both constructs [12,16,24] in children and adolescents, no meta-analysis or systematic review has considered studies published in the period 2017-2019, and none has included the variable suicidal ideation. The latter, however, may be a precursor of suicidal behaviour [119], and it is therefore an important variable in the context of adolescent wellbeing. Although, in the period considered by this review, only a small number of studies have examined the relationship between parenting practices and suicidal ideation in young people, our analysis suggests that suicidal ideation is negatively associated with parental warmth, behavioural control, and autonomy granting, and positively associated with psychological control and harsh control.
It should be noted, however, that the corresponding effect sizes are small. Some research also suggests that indulgent parenting may protect against suicidal thoughts, and a relationship has been reported between suicidal ideation and both parental imposition and over-involvement.
Regarding depression and anxiety, overall, our results are consistent with published findings, since previous meta-analyses have likewise reported a negative relationship between parental warmth and internalizing symptoms in children and adolescents [15,16,[23][24][25], as well as a positive association of both psychological control and harsh control with the development of depression and anxiety symptoms [12,29,30]. Regarding behavioural control, and in line with previous meta-analyses, our results showed that this practice was negatively associated with both depression and anxiety, although the average effect size is very small. Also consistent with existing meta-analyses, we found a negative relationship between autonomy granting and internalizing symptoms in adolescents [12,15,16]. We also observed a relationship between adolescent depression and a number of other parenting practices, including inconsistent discipline, negative parenting, family dysfunction, a desultory parental style, over-reactivity, and effective parenting. Additionally, we found an association between anxiety and family dysfunction and overprotection. Yap et al. [16] reported similar results in relation to inconsistent discipline and overprotection. Finally, as regards parenting styles, our results suggest that neglectful and authoritarian parenting are associated with higher levels of depression, whereas an authoritative parenting style is related to less depression. These results are partially consistent with the classic studies conducted in the United States (e.g., [8,9]), although there are studies conducted in other cultural contexts in which the indulgent style is the most optimal (e.g., [120,121]).
In general, and in line with the results of previous meta-analyses (see, for example, [12,15,27]), the effect sizes for these associations were either small or moderate. Therefore, these effect sizes are as expected and are considered adequate, since many other potentially important variables are related to internalizing problems in adolescents. However, this does not mean that parental practices are not relevant in terms of generating or protecting against such problems. Therefore, several primary studies have also pointed out the importance of such results, despite the effect sizes found [8,122]. In this respect, it is worth noting that some of the studies included in this systematic review, which report the largest effect sizes, have involved clinical samples of young people, for example, adolescents with inflammatory bowel disease [65], adolescents who have assaulted their parents [50], or adolescents hospitalized following a suicide attempt [88]. This suggests that it is in these kinds of populations where adequate parenting practices may have a greater protective effect in relation to adolescent wellbeing. This review has certain limitations. First, the considerable variation in the conceptualization of parental socialization and the diverse range of instruments used hampers the comparison of results across studies. A related issue here is that, despite having conducted an exhaustive database search, it is possible that some studies were missed due to their use of different terminology. A further limitation is that although we considered a broad period (2010-2019), the number of studies reviewed was not sufficient to analyse the influence of potential moderator variables, such as sample characteristics, cultural context, participants' gender, parental gender, research design, or the quality of the measurement instruments used. In this regard, it should be noted that many studies have had to be excluded due to two main reasons: that the participants were not adolescents between the ages of 12 and 18 and that the studies did not examine the relationship of parenting practices or styles with anxiety, depression, and/or suicidal ideation in adolescents. Therefore, for example, it was common to find studies in which anxiety or depression had been evaluated in parents and not in adolescents, or studies in which participants were adults. Finally, our systematic review has been limited to studies that have used questionnaires and this fact may be related to the effect sizes found.
As areas of future research, it is worth mentioning the need to systematically analyse the parental styles that may be more effective for the psychosocial adjustment of adolescents depending on the cultural context, since, as previously noted, there is no consensus on this issue. On the other hand, a systematic review could be carried out with various operationalizations of parental socialization.

Conclusions
The studies reviewed suggest that parental warmth, behavioural control, and autonomy granting are inversely associated with internalizing problems, especially depression, in adolescents. Conversely, psychological control and harsh control by parents show a positive relationship with adolescent anxiety, depression, and suicidal ideation. Although the associated effect sizes are small or moderate, the results suggest that these variables should be taken into account when designing programmes aimed at promoting parenting styles conducive to the wellbeing of adolescents.

Conflicts of Interest:
The authors declare no conflicts of interest.