Parentification Vulnerability, Reactivity, Resilience, and Thriving: A Mixed Methods Systematic Literature Review

Parentification occurs when youth are forced to assume developmentally inappropriate parent- or adult-like roles and responsibilities. This review thoroughly examines current empirical research on parentification, its outcomes, and related mechanisms to outline patterns of findings and significant literature gaps. This review is timely in the large context of the COVID-19 pandemic, when pandemic-induced responsibilities and demands on youth, and the shifting family role may exacerbate parentification and its consequences. We used the 2020 updated Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) framework to identify 95 studies (13 qualitative, 81 quantitative, 1 mixed methods) meeting eligibility criteria. Representation from six continents highlights parentification as a global phenomenon. Using thematic analysis, we identified five themes from qualitative studies and five from quantitative studies. These were further integrated into four common themes: (1) some parentified youth experienced positive outcomes (e.g., positive coping), albeit constructs varied; (2) to mitigate additional trauma, youth employed various protective strategies; (3) common negative outcomes experienced by youth included internalizing behaviors, externalizing problems, and compromised physical health; and (4) youths’ characteristics (e.g., rejection sensitivity, attachment style), perceived benefits, and supports influenced parentification outcomes. Future methodological and substantive directions are discussed.


Introduction
Parentification-also known as adultification, spousification, child carers, or role reversal-occurs when youth are forced to assume developmentally inappropriate parentor adult-like roles and responsibilities. Definitions highlight that parentification is distinct from supervised or monitored higher-order household responsibilities used by parents to promote positive youth development via leadership skills and character-building. Instead, parentified children and adolescents are expected to become pseudo-parents and pseudo-adults long before they are cognitively and physiologically equipped for these roles. Common roles children assume include household earner, self-carer, family-navigator, language and cultural broker, self-educator, counselor, confidant, caregiver, and emotional supporter (for parents and siblings) [1][2][3].
The prevalence of parentification in the US is unknown. In 2006, Siskowski and colleagues [6] estimated 1.3-1.4 million parentified 8-18-year-olds (2.9% of the population) in the US, an underestimate according to Hooper and colleagues [3]. More recently, it is reported that 2-8% of youth under age 18 are young carers in high-income countries [7]. A recent study of Polish adolescents reported parentification prevalence estimates exceeding 30% during COVID-19 [8].
Now more than ever, understanding the impacts of parentification is a meaningful undertaking when considering the COVID-19 pandemic. COVID-19 resulted in the sudden loss of multiple resources (e.g., childcare, schooling, employment) due to mandated closures, isolation, quarantine, distancing protocols, and the abrupt loss of family members and friends from illness and death. Since March 2020, when COVID-19 began to rapidly spread within the US, nearly 250,000 children lost a caregiver and households with children were more likely to experience financial hardships including the loss of jobs and health insurance [9]. Education was disrupted for over 1.6 billion children with a projected $17 trillion in life-long earning losses, globally [10]. In short, the COVID-19 pandemic exacerbated contributing factors (e.g., caregiver death; loss of job, income, and health insurance; disrupted education) to parentification.
This systematic review aimed to identify the predictive factors contributing to both positive and negative outcomes of parentification and is timely considering the circumstances forced upon the world by the COVID-19 pandemic. Before delving into the empirical findings, we provide a brief background of the dimensions, sources, and consequences of parentification.

Dimensions of Parentification
Classic models of parentification differentiate types of parentification based on the function it serves, typically as either instrumental or emotional parentification. Instrumental parentification involves youth assuming the responsibilities to maintain the household (e.g., meals, chores, finances). Emotional parentification requires youth to tend to the emotional needs of family members. This can include becoming a parents' confidant (e.g., spousification), elevating siblings' self-esteem, and even promoting harmony among the members. Some parentified youth may fulfill both the instrumental and emotional needs of the family.
Alternatively, researchers have also studied parentification by focusing on the various roles that can be assumed; specifically, parent-focused, sibling-focused, and spouse-focused parentification. The role-based approach emphasizes the role a child takes on, such as becoming a parent to care for their own parents (parent-focused) or siblings (sibling-focused), or even a spouse to their parents (spouse-focused). The function-based and role-based approaches to parentification are not mutually exclusive concepts, as parent-focused parentification could provide either/or possibly both emotional or instrumental functionality.

Sources of Parentification
Parentification typically results from the intentional or unintentional abdication of parenting responsibilities, child neglect, or child maltreatment by family-of-origin members, especially primary caregivers. The contributing factors of parentification tend to co-occur and there are countless reasons why parents cannot fulfill their role or why children are forced to assume these roles and responsibilities.
Common sources of youth role reversal and role overload include parental illness (e.g., HIV; opioid addiction), parental loss (e.g., death, divorce, incarceration), parental mental illness and physical disability, crises (e.g., displacement via eviction, war, unemployment), dysfunctional family dynamics (e.g., domestic partner violence), and migration (e.g., refugee, immigration). Additionally, parents who were themselves parentified may expect their children to do the same, creating a culture that is passed on for generations [11].
Although the level and degree of impact in the US are unknown, changing US demographics suggest increasing numbers of children are vulnerable to being parentified. In 2019, 26% of youth lived with only one biological parent and 4.0% lived with neither [12]. Parents who work long hours to meet their financial needs may come home with a diminished capacity to attend to household responsibilities and child needs due to fatigue. Living with single parents and parents who experience financial hardships may increase the risk of parentification of these children. With the increasing prevalence of obesity, diabetes, HIV/AIDS, and other chronic health conditions among adults and children, youth in the family without chronic conditions are at greater risk of becoming parentified, tending to the needs of their parents, siblings, or both.

Consequences of Parentification
As articulated in Minuchin's family system theory [13], a hierarchy of power exists among the family subsystems, and clear hierarchical boundaries between parents and children are considered to be critical to children's positive development. The dissolution in or the alteration of the family structural boundaries, as may occur when children are parentified and essentially become the parents, breadwinners, and other roles with power in the family, has important implications for children's behavioral and moral development. Thus, parentification may be linked to adverse behavioral or life outcomes. Parentification is considered particularly harmful when youth are forced to take on tasks beyond their developmental abilities and when they do not receive adequate support [11,14]. Consistent with this proposition, previous research showed that parentified children experience suboptimal outcomes in adulthood, including higher incidence of depression, anxiety, drug use and addiction, under-and un-employment, poor physical health, and lower educational attainment [15][16][17].
In contrast, it has also been argued in early theoretical work that parentification is not necessarily pathological [18]. When youth assume parental roles of moderate intensity in a time-limited manner and their contributions are appreciated, the parentification experience may instead be adaptive [11]. Parentification could beneficially influence youth development because it provides youth with opportunities to master socialization and coping skills, and be self-reliant, which contribute to healthy identity formation and improved self-esteem. In support of this, early parentification among children of parents with HIV/AIDS were found to have better adaptive coping skills 6 years later [19]. Similarly, parentification was linked to less risky sex among high-risk adolescent girls [20] and with resilience in general [17], although the association may be attributable to other factors such as socioeconomic background.

Purpose of Current Review
This review provides a thorough examination of current empirical research on parentification, its positive and negative developmental consequences, the heterogeneity of parentification impacts, and related mechanisms. Through the examination of the empirical research, we then outline significant literature gaps and provide future research directions.

Materials and Methods
This review was guided by the 2020 updated Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) framework [21,22].

Eligibility Criteria-Inclusion and Exclusion
To be eligible for inclusion, studies had to meet the following criteria: (1) empirical study including quantitative and/or qualitative primary or secondary data analysis; (2) peer-reviewed publication; (3) include at least one developmental or outcome variable (e.g., mental or physical wellbeing, stress coping, leadership skills, substance use, sexual risk-taking, educational attainment, workforce engagement); (4) written or translated into English; and (5) full text available. Studies could include a wide range of designs and types of data including retrospective, prospective, qualitative, quantitative, experimental, quasi-experimental, and non-experimental studies. Studies were excluded if they were (1) theoretical or review articles with no primary or secondary data; (2) not peer-reviewed (book chapters, reports, theses/dissertations, conference proceedings); and (3) exclusively descriptive of parentification types, prevalence, or incidence with no outcomes reported. If a study did not clearly meet inclusion criteria, two or three authors discussed the study until consensus was reached.
For synthesis, studies were grouped in two ways. First, they were stratified by quantitative versus qualitative and mixed. Then, strata were grouped by study focus: outcomes only (positive, negative, both); outcomes and mechanisms (mediators/moderators).

Information Sources
To ensure a comprehensive literature search, three databases were searched on 21 August 2021, for relevant articles: PsycInfo, Academic Search Complete, and Web of Science. American Psychological Association PsycInfo is a database of abstracts and articles related to psychological, social, and behavioral science. Academic Search Complete (EBSCO Publishing) is a scholarly database spanning numerous disciplines and includes both openaccess and non-open-access peer-reviewed and grey literature (e.g., books, reports). Web of Science database indexes scholarly products (e.g., articles from over 12,000 journals, 148,000 conference presentations) from physical and social sciences, humanities, and arts.

Search Strategy
For each database, articles available from the inception date until July 2021 were included. MESH and Boolean search terms were used to generate relevant articles. Parentification is known by other related terms, including spousification, adultification, and role reversal. Therefore, these and their related terms (denoted by *) were searched. Further, the search was contingent (using AND) on outcome-related terms (e.g., outcome, resilience, thriving, effect) to answer the research question. Article search results (record data including title, authors, publication date, journal name, and abstract) for resultant articles were exported to Excel and imported into SPSS to remove duplicates. Table 1 summarized search terms by database and the number of resultant articles by the database. Table 1. Search Algorithms and Articles Generated by Database.

Search Terms (Same in All Three Data Bases) Articles Count Database
(spousif* or parentif* or adultif* or "role reversal") AND (outcome* or resilien* or thriv* or stress or react* or benefit* or impact* or effect* or affect* or positiv* or negativ*)

Selection and Data Collection Processes
This dataset was divided across three authors (JKD, FRC, and MKN) for the initial review of titles and abstracts for eligibility. Ambiguous articles were cross-checked by a different reviewer. For a few unclear articles, the three authors (JKD, FRC, and MKN) discussed the article until a consensus was reached about inclusion or exclusion status. Then, the resulting list of articles was divided in half for a full article review and data abstraction (Round 1). Two of the authors (MKN, AMC) independently abstracted descriptive data from articles (participants, study design, data type, parentification measures used, main findings, category of study) and flagged ambiguous articles for review by the first two authors and noted articles that did not meet eligibility criteria. Next, half of the assigned articles were cross-validated by a different reviewer for verification of abstracted data and relevance. Then, the list of articles for each reviewer was stratified by quantitative or qualitative type and randomly divided into Tiers 1, 2, and 3.
The team met to discuss additional data fields to abstract including specific outcomes, positive or negative relationship direction, significance, and future directions. Then, two reviewers read assigned Tier 1 articles, updated existing fields, and abstracted data for new fields (Round 2). These results were reviewed by the first two authors and the full team met to discuss the process of data abstraction. Then, the reviewers (MKN and AMC) continued abstracting data for Tier 2 and Tier 3 articles and flagged any challenging articles for JKD and FRC to review. In Round 3, cited studies discovered through article review that met the inclusion criteria were also included and data were abstracted. Every article was reviewed independently by at least two reviewers.

Data Items
Data items extracted for eligible articles are listed and defined in Table 2. Most data items are standard and are discussed below. Both positive and negative outcomes were included. This scoping review intentionally optimized study inclusion by broadly defining relevant outcomes and not placing temporal restrictions on outcomes (e.g., early childhood, adolescence, adulthood). The exposure variable of interest was parentification and no temporal restrictions were placed on when this experience occurred.

Study Risk of Bias Assessment
Every article was reviewed independently by at least two reviewers. Studies were cross-verified for relevance by at least one additional independent coder. If questions persisted, a third independent coder reviewed the article, and a fourth reviewer was engaged only in ambiguous cases for which their content expertise was needed. As needed, discussions were conducted until a consensus for inclusion was reached.
The Mixed Methods Appraisal Tool (MMAT) [23] facilitated study risk of bias, rigor, and quality assessment. Authors independently assessed each article on five components relevant to the article type (e.g., qualitative, quantitative descriptive) with "yes", "no", and "can't tell" designations. The tool was adapted to include partial credit for some criteria, including screening items that required a specified research question to receive a "yes" designation. If articles included a purpose, aim, or objective statement, they were further reviewed. Further, subcriteria were created for double-barreled or complex criteria. For example, five subcriteria criteria-inclusion, exclusion, response rate mentioned, and sampling strategies detailed-were used to assess whether a study represented the target population for quantitative studies. Partial credit was given, and overall quality ratings were adjusted to no, low, moderate, or high. At least 10% of articles were cross-validated by a second independent reviewer. Interrater questions about meeting criteria and sub-criteria were noted. Discrepancies between reviewers (~5%) were discussed as a team and resolved.

Synthesis Methods
For study characteristics and parentification and outcome measures results, patterns of findings were quantified by count. For substantive findings, we utilized the six steps of thematic analysis outlined by Braun and Clarke [24]. The first three steps-data familiarity, initial coding, theme generation-were undertaken by independent reviewers separately for qualitative studies and quantitative studies. The second three steps-theme review, theme definition and naming, and report development-involved coders across both study types. Additional details for the six steps are as follows. Data familiarity involved reading each study. Independent reviewers initially coded studies using inductive open-coding (identifying major and subcodes and definitions). Themes were generated based on shared or associated codes. Themes were reviewed by the coders and then by the larger team. Once themes were agreed upon, definitions and names were drafted and them finalized. Last, these were written up for the manuscript. Objectivity was promoted via multiple approaches. First, each study was read, coded, and themed by independent reviewers. Summaries of themes were written by reviewers leading each article. Second, themes were generated for qualitative articles and quantitative articles separately followed by integration analyses conducted for the two study types. Last, these themes were reviewed by other team members before team discussions that centered on description meaning and wording, theme consistency and inconsistency by study types, integration of themes across studies, and consensus of final themes to include in the report.

Certainty Assessment
For quantitative studies, MMAT includes criteria related to the use of confounders. For mixed methods studies, one criterion assesses whether study components adhere to quality standards of quantitative and qualitative traditions. For quantitative studies, the criterion was assessed using four sub-criteria: any confounders included, relevant demographic characteristics used, other (non-demographic) confounders (e.g., pre-exposure outcome levels; the age of parentification onset), and use of sophisticated statistics to model potential confounders.

Results
The result section is organized into several parts. First, the study selection and characteristics are described (Table 3). Second, measurement patterns regarding parentification ( Table 4) and outcomes (Table 5) are shown. Third, study findings are presented as themes by design and analytic approaches (quantitative followed by qualitative and mixed). Last, theme patterns are synthesized across the study design.

Study Selection
A total of 1111 articles were retrieved across three databases (227 from PsycInfo; 313 from Academic Search Complete; 571 from Web of Science). Of these, 445 were duplicates and 374 were removed for other reasons (e.g., review articles, no outcomes included, no mention of parentification or related terms or constructs). The manual screening was performed on 292 abstracts and 149 were excluded for not meeting inclusion or meeting exclusion criteria. Of the remaining 143 articles, 135 could be retrieved for eligibility assessment, 40 of which were excluded. The 40 articles were excluded for not meeting the inclusion criteria of our review; specifically, these articles were review articles (n = 13), did not include exposure (i.e., at least some participants reporting parentification experiences) or outcome variables (n = 21), reported irrelevant outcomes (n = 4), or summaries in English but the English full-text article could not be located (n = 2). This review focuses on 95 articles successfully screened and meeting all eligibility criteria. The PRISMA flow diagram outlining inclusion and exclusion processes is presented in Figure 1. Table 3 shows study characteristics for one mixed-methods [25], 13 qualitative [26][27][28][29][30][31][32][33][34][35][36][37][38], and 81 quantitative [3,8,17,19,20, studies included in this review (n = 95; ordered alphabetically within study type). Four papers were published before 2000, 20 between 2000 and 2010, 57 between 2011 and 2019, and 14 after 2020. A total of 55 studies (58%) were conducted in the United States, 19 in Europe (20%), eight in the Middle East (8%), seven in Asia (7%), five in Canada (5%), and one in Africa (1%). All studies used non-random sampling methods to recruit participants.

Study Characteristics
Studies varied greatly in sample size (quantitative study range: 41-1796; qualitative study range: 2-34; mixed: 128). Among the quantitative (including one mixed-methods) studies, 17 had fewer than 100 participants, 24 had between 100 and 199, 14 had between 200 and 299, and 24 had 300 or more participants. Among the qualitative studies, six had Twenty-six studies included parent-child dyads or triads, 19 of which were motherchild dyads. This disproportionate focus on mother-child dyads reveals a gap within the literature exploring fathers' and sons' perspectives on parentification. Six studies (6%) included populations of children with HIV/AIDS-infected mothers. Five studies (5%) included populations of typically developing (TD) siblings who reported having a sibling with autism spectrum disorder (ASD). Five studies (5%) included immigrant and refugee populations. Two studies included clinical populations or cases.         : Participants are characterized as adolescents in this review despite that they were referred to as children in the original publication. The reason is that the youth were aged 13 to 19 years; Cimsir, E. and Akdogan, R., 2021 [55]: It contains 5 studies; only Study 5 characteristics are reported here because the first 4 studies are all about development of the emotional incest scale and its validation.

Risk of Bias and Certainty Assessment
As noted above, MMAT criteria ratings were used to assess study risk of bias, rigor, and quality. For the 13 qualitative studies and one mixed methods study (which reported primarily qualitative data), approximately 86% explicitly described appropriate research questions or objectives, data collection methods, findings, and interpretation, whereas the remaining 14% provided some description or narrative to substantively meet the criteria. Nearly 86% had medium to high coherence between data, analysis, and interpretation for at least one study aim or objective. Given these high MMAT criteria ratings, it was concluded that the risk of bias was low, and the study quality was high for the qualitative/mixed studies included in this review. Across the 81 quantitative articles, about 67% used appropriate measures for intervention/exposure and outcomes, nearly 73% specified at least one inclusion or exclusion criteria related to the target population, and 65% met at least one of the criteria for analyses that accounted for confounders. The authors relied on substantive comments rather than calculated scores, as encouraged by tool developers. We conclude that our review reflects a minimum risk of bias as it comprehensively incorporates the study literature.

Parentification Measures
Original measure names and descriptions as well as study-specific modifications and psychometric properties are presented in Table 4 for 11 unique measures used by two or more studies (10 measures were only used by one study; not shown). Measures are organized by frequency of use (alphabetically for tied measures) and studies within a measure are alphabetically ordered by first author's last name. Studies using more than one measure appear more than once in Table 4. Reliabilities were generally acceptable (α: 0.70 to 0.95) with a few exceptions. Of the 78 total entries across 21 measures, 53 (67.9%) made modifications to the original measures. Common modifications to original measures included using select subscales, reducing the number of items, or adopting peer-reviewed translations. Languages in which measures have been administered include Dutch, English, French, German, Hebrew, Polish, and Russian.
The most frequently used measure (n = 18) was the Parentification Questionnaire (PQ). It was first developed by Sessions and first presented in its 42-item format with psychometric properties in a thesis [115]. The original measure includes three subscales assessing instrumental and emotional parentification and perceived unfairness. A few studies used only one or two of the original subscales. Reliabilities ranged from 0.61 to 0.92 with four studies having Cronbach alphas below 0.70 and typically included younger participants (adolescents or young adults).
The second most commonly used measure (n = 14), the Parentification Inventory (PI), was developed by Hooper [116]. The 22-item scale includes three subscales: parentfocused parentification (PFP), sibling-focused parentification (SFP), and perceived benefits of parentification (PBP). Alphas ranged from 0.58 to 0.89 with the SFP characterized by lower reliability. The third most often used measure (n = 10) was the Parentification Scale (PS), which was developed by Mika and colleagues [117]. This 30-item scale includes four subscales: (1) child is parent to parent; (2) child is spouse to parent; (3) child parents siblings; and (4) child acts in all three ways. Alphas ranged from 0.57 to 0.95.
A less commonly used measure (n = 7), the Filial Responsibility Scale, Adult Version (FRS), developed by Jurkovic, Thirkield, and Morrell [118], includes 60 items to assess six subscales: Past and Current Instrumental Caregiving, Past and Current Emotional Caregiving, and Past and Current Unfairness. Alphas ranged from 0.74 to 0.94. The Parentification Questionnaire for Youth (PQ-Y), developed by Godsall and Jurkovic [119] and modified from the adult version [120], was used by five studies. It includes 20 items that assess emotional and instrumental parentification without explicit subscales. Alphas ranged from 0.64 to 0.80. Six measures detailed in Table 4 were used by three (Family Structure Survey and Inadequate Boundaries Questionnaire) or two studies (Child Caretaking Scale, Childhood Questionnaire, Parent-Child Boundaries Scale III, and Relationship with Parents Scale). Several parentification measurement patterns were observed. First, studies typically focused on role-based (e.g., sibling-focused) or function-based (e.g., emotional) parentification, and it is not common that studies assess all these dimensions of parentification. Second, there is a dearth of measures designed to capture the parentification of youths either assessing their current experiences of parentification or retrospective parentification accounts of earlier experiences. Studies that included youth or youth measures reported lower reliabilities, suggesting effort should be made to develop a more reliable measure for youth. Third, most measures assessed adult retrospective accounts of parentification during childhood. Only one of seven studies using the Filial Responsibility Scale, Adult Version used current subscales, whereas the others focused on past subscales.          College students; grew up in an intact family with one mother and one father

Outcome Measures
Measure names, descriptions, study-specific modifications, and psychometric properties are presented in Table 5 for the constructs studied more than twice. Close to 30% of the studies are not included in Table 5 because they included constructs studied no more than twice. Broadly, the constructs excluded from Table 5 can be grouped into positive life outcomes, family relationships, and clinically relevant outcomes. Constructs in Table 5 are ordered based on the frequency of being studied. Within each construct, measures are ordered from the highest to the lowest frequency. Note that studies may appear multiple times when they investigated more than one construct or used more than one measure for one construct.
The most frequently studied construct was depression (n = 21), followed by internalizing and/or externalizing problems (n = 12). These constructs were all measured with well-validated instruments, and depression had good reliability. However, over half of the studies on internalizing and/or externalizing problems failed to report reliability for their specific studies.
Other negative outcomes include psychological/emotional distress which was often analyzed as a latent variable (n = 8), substance use (n = 10), antisocial behavior (n = 3), and risky sex (n = 3). The last three constructs were behavior-based measures and often measured by indicating the presence or frequency of the behavior [132,133]. Except for three studies that used the same instrument for substance use [3,65,67], no two studies used the same measures for antisocial behavior, risky sex, or substance use. Positive outcomes include self-esteem (n = 5), satisfaction with life (n = 3), and efficacy (n = 3), and they were measured similarly across studies with good reliability.
Several patterns were observed for outcome measures. First, outcome constructs were routinely assessed with well-validated instruments and reported good reliability. Second, only very few studies examined positive outcomes. This focus on adverse outcomes neglects the potential positive outcomes of parentification. Third, when constructs were only studied a couple of times, there is not ample evidence to conclude whether and how these constructs are related to parentification. Fourth, studies that focused on the same constructs only a few times were often conducted by the same researcher groups, which may suggest the limited research scope and lack of attempts to integrate individual research into the broader literature of parentification.          [152]: The name CAGE derives from the four items: "Cut back on drinking", "Annoyance at criticisms about drinking", "Feeling Guilty about drinking" and "Using alcohol as an Eye opener". All items require "yes" or "no" answers Dragan and Hardt, 2016 [56] Subjects all registered with a market research company The item "Guilt" was excluded from the analysis due to the excessive difference between the samples of Poles and Germans; A binary variable "problematic alcohol use" was set to "1" whenever at least one of the three items was endorsed None stated Paternal role reversal ns

Substantive Themes-Quantitative Studies
Most quantitative studies were cross-sectional and only six [19,70,80,92,95,104] were longitudinal studies. As reviewed above, the parentification construct was measured in various ways, including overall level, functionality (instrumental and emotional), roles (parent-focused, sibling-focused), and perceptions (perceived unfairness and perceived benefit). Thus, the link between parentification and outcomes was often contingent upon the dimensions of the parentification being examined. The most studied negative mental health outcome was depressive symptoms (25% of all quantitative studies). Across the 21 studies on depressive symptoms, emergent patterns of findings included: (1) a positive association for emotional parentification (five positive findings vs. one null finding) and parent-focused parentification (four positive vs. two null); (2) null rather than positive associations for instrumental parentification (three null vs. one positive) and sibling-focused parentification (five null vs. one positive); (3) when parentification was an overall construct or studied as role reversal without differentiating any further dimensions, the findings were mixed in its connection with depressive symptoms (four positive vs. two null vs. two negative); (4) perceived benefit of parentification, when studied, was consistently negatively associated with depressive symptoms, suggesting its protective role.
Among the studies on internalizing problems (n = 11; measured either by forming a latent variable or summing across anxiety and depressive symptoms) and studies on psychological/emotional distress (n = 8), only two studies differentiated instrumental from emotional parentification [17,51], two studies exclusively focused on emotional parentification [86,88], and two other studies focused on role-based parentification [95,100]. More emotional parentification was consistently linked with internalizing problems or distress, whereas instrumental parentification consistently showed a null association with internalizing problems or distress. Furthermore, role-based parentification was not associated significantly with internalizing problems or distress.
The remaining 13 of the 19 studies on internalizing problems or distress examined the parentification construct as a whole. Researchers interested in internalizing problems or psychological distress likely investigated broader categories of psychopathology (i.e., a spectrum of problems) instead of focusing on more specific subscales of symptoms (e.g., anxiety alone). This tendency may have also spilled over to their treatment of parentification, where the focus is on the overall construct rather than dimensions. Parentification as an overall construct was positively associated with internalizing problems across studies except for one [19], yet role reversal as an overall construct as studied in two papers was not associated with internalizing problems or distress [72,73]. Regarding temporal order, one longitudinal study found that adolescent but not childhood boundary dissolution predicted internalizing problems in adolescence [92].

Theme 2.
Externalizing Problems (Broad Spectrum, Antisocial Behavior) are Associated with Emotional Parentification, and with Overall Parentification to a Lesser Degree; and Neither Substance Use nor Risky Sex was Linked with Parentification Consistently.
The next most studied outcome was externalizing problems (n = 12), including the broad spectrum of externalizing problems (n = 8) and aggression/antisocial behavior (n = 4). Only one study examined emotional versus instrumental parentification [70] and only one examined maternal emotional parentification [86], both of which found that more emotional parentification was associated with more externalizing problems. Role-based parentification was studied only in one study and was not linked with antisocial behavior [53]. Parentification as an overall construct was positively linked with externalizing problems in four studies [77,82,95,104], yet an almost equal number of studies (n = 3) found null results [76,113,161]. The difference in findings cannot be explained by power/sample size because studies that found a null association had comparable sample sizes as the ones revealing a positive association. Together, it suggests that the positive association between overall parentification and externalizing problems was weak at best.
Most of the 11 studies involving substance use as an outcome revealed null findings. Across the 18 associations tested in these 11 studies, only five associations were statistically significant and three were in the unexpected negative direction where more parentification was associated with fewer substance use problems [19,72,96]. None of these studies examined instrumental or emotional parentification as they either focused on overall or role-based parentification.
Like findings in substance use, a null finding between parentification and risky sex was more common. Among the six associations examined in three studies on risky sex, the only two significant associations were in opposite directions where more parentification was associated with lower sexual intention [20], and the more parental role was linked with more concurrent risky sex measured as a latent variable [95]. Compared to the negative mental health outcomes, very limited studies focused on positive outcomes, and the largest limitation is the small number of studies focused on the same construct to allow a clear conclusion to be drawn. This was true even for the most studied positive outcomes: self-esteem (n = 5), self-efficacy (n = 3), and satisfaction with life (n = 3). Parentification was not associated with self-esteem in general as only three of the 11 associations tested in the five studies were significant: Self-esteem was positively associated with perceived benefit [45] and parentification [106], and negatively linked with perceived role unfairness [80]. No consistent findings emerged for efficacy or satisfaction with life across six studies.
Other positive life outcomes such as posttraumatic growth [17,66], happiness [97], school achievement [8], social competence [102], prosocial behavior [42], and empathy [103] were studied no more than twice, but findings tentatively suggested that parentification may be associated with beneficial outcomes. For example, more role-fulfilment from parentification was linked with a higher daily happiness rating [97], more instrumental parentification with better school achievement [8], more parental role with more social competence [102], more parentification with better cognitive empathy [103], and better coping skills 6 years later [19]. However, studies also suggested a lack of association between parentification and happiness [97], emotional parentification and school achievement [8], and parentification and affective cognitive empathy [103]. Future studies on the same constructs are needed to allow any consistent results patterns to emerge. Seventeen studies examined mechanisms through which parentification affected the outcomes. The only mediators that were tested more than once were the differentiation of self (defined as "capacity for emotional self-regulation and the ability to regulate the relational impulses of separateness and togetherness" [67,68]), rejection sensitivity [62,63], and attachment styles [41,75,93]. All three constructs were significant mediators connecting parentification with various outcomes. Differentiation of self mediated the effect of perceived unfairness on general mental health outcome [68] and alcohol use [67]. Rejection sensitivity mediated the boundary dissolution effect on false-self behavior [62], and the effect of parentification on same-sex friend intimacy [63]. Lastly, the emotional parentification effect on constructive communication was mediated via anxiety but not avoidance adult attachment [75]; paternal parentification effect on adult relationship satisfaction and insecurity was mediated via avoidance/anxiety attachment [41]; and enmeshed attachment mediated the relationship between role reversal and paranoid and schizotypal personality disorder traits [93]. Interestingly, the studies examining the above mediators are largely the work of one research group, except for attachment. The remaining eight studies examined other mediators and outcomes, including quality of life, perceived stress, and communication about alcohol; given the once-off nature of the studies, it is difficult to ascertain patterns.
Seven studies examined parentification as a mediator for other associations, typically associations between other family environments and health/behavior outcomes [39,48,[58][59][60]64,88]. Four studies found parentification as a significant mediator connecting family risks with negative outcomes [39,48,59,88], even though the exact family risk factors ranged from domestic violence to parental alcohol misuse, and the outcome also ranged from internalizing problems to aging. Heterogeneity of parentification effects on outcomes is organized into two primary categories based on features of moderators: demographic features, mostly sex, and race, versus non-demographic characteristics (e.g., social support). Nine studies examined nondemographic moderators for the parentification effect, and exacerbating and buffering moderators were identified. Maternal depression appeared to exacerbate the negative effect of parentification on adjustment [89,111]. Some protective factors emerged that either buffer the negative effect of parentification or facilitate the effect of parentification on positive outcomes, including disclosure of worries, social support, and religious service attendance. For example, more sibling-and parent-focused parentification was associated with negative interaction with siblings and less wellbeing, but only at low disclosure of worries about autistic sibling to parents [78]; language brokering was linked to less selfefficacy, but only at low parental/social support [85]; parentification was linked with distress, but only at low levels of social support [101]; more perceived unfairness was linked to more alcohol use for those not attending religious service [67].
In summary, parentification, particularly emotional parentification was linked with more depressive symptoms and internalizing problems. Findings on externalizing problems were weaker and revealed null associations more often than not for substance use and risky sex. Positive outcomes of parentification were much less investigated, but research suggested a potential positive link between parentification and desirable life outcomes (e.g., coping skills). Differentiation of self, rejection sensitivity, and attachment styles were promising mechanisms that elucidate the link between parentification and negative health outcomes. Parentification also mediated the associations between various family risk factors and adverse outcomes. Lastly, while exploring the heterogenous effect of parentification, researchers explored not only exacerbating factors (e.g., maternal depression) for the negative impact of parentification, but also buffering factors (e.g., social support) that exerted more of a protecting role to buffer the negative impact of parentification and/or to facilitate the positive outcomes of parentification.

Substantive Themes-Qualitative and Mixed Methods Studies
Across the body of 14 qualitative and mixed methods studies, parentification was not consistently related to negative or positive outcomes. Ten studies reported both positive and negative outcomes, two studies reported only positive outcomes [28,36], and two studies reported only negative outcomes [31,35]. Five themes emerged across the studies.

Theme 1. Personal Growth and Strengthened Sibling Relationships Come from Adversity.
Reported positive outcomes from parentification included personal growth (e.g., emotional intelligence, prioritizing others, social skills, independence) and strengthened sibling relationships. In seven qualitative and mixed methods studies, participants reported personal growth as a positive outcome of their parentified experiences. For some, personal growth meant enhanced emotional intelligence, characterized by increased perspective-taking, enhanced empathy, and a desire to understand and care for others [28,29,34]. Personal growth was also defined as prioritizing others above self-interests and the development of social skills [26]. For others, personal growth was described as developing greater independence [32], agency [26], or grit [33].
A strengthened relationship between and among siblings was a means of relational coping, and was particularly important in contexts where youth were physically or socially isolated (especially through stigma), experienced mothers' mental health concerns, or were abused, neglected, or witnessed domestic violence [26,27,34]. Strengthened relationships with siblings were both an outcome of parentification and helped reinforce and promote personal growth. When building stronger sibling bonds, these youth developed empathy, understanding, acceptance for others, and other social skills (e.g., positive, emotional connections, providing social support); these all contributed to them aspiring to have future careers in human services fields to care for others and striving to break down mental illness stigma [26,34]. Theme 2. Self-Preservation Mechanisms-Self-sacrifice, Distancing, and Balancing Desire for Closeness.
To manage their experiences, parentified children developed numerous coping strategies, many of which were suboptimal by conventional standards but adaptive for their situation [25,30,32,35,37]. Some youth expressed a self-imposed heightened sense of protecting parents from worry and stress [32]. For example, parentified youth did not speak about abuse by employers because of worry about losing employment and no longer being the family breadwinner [35]. This was a form of emotional self-sacrifice to promote instrumental resources for the family, and it also demonstrates how self-sacrifice in one domain (downplaying emotional needs or workplace silence) serves as self-preservation in another domain (minimizing the effect of additional stress on parents or potentially losing household income).
While some youth protected their parents by not sharing worrisome and stressful matters, some youth chose to emotionally distance themselves from mothers to avoid criticism and retraumitization [31,34,37] Mother-child relationship quality influenced how much and why youth shared information: protect mothers (self-sacrifice) or protect themselves (distancing). The participants who preferred more distance from mothers limited contact to avoid conflict despite the tension of wanting closeness.
Some youth found ways of balancing family demands with self-care and individual needs. For instance, some scheduled time to be alone or interact with friends. Chee and colleagues [28] described this as "children's parentification was found to be a process involving intense yet subtle dynamics of cooperation, negotiation, and resistance" (p. 209). Some parentified youth discussed how greater maturity and independence relative to peers prepared them for a transition to adulthood, but these positive attributions were often accompanied by negative attitudes about forfeited childhoods and compromised adult outcomes. Participants described a "lost childhood" marked by insufficient time to engage in play and other activities with same-aged mates [32], and basic attachment-related needs for comfort were not met in childhood [25,31].
Due to premature transitions into adulthood roles, parentified youth commonly experience compromised human capital in the form of incomplete educational attainment due to school dropout or low attendance, limited to no time with peers to develop relationships and friendships, and early transition into parenthood partially the result of early or risky sexual behavior and substance use [29,32,38]. Compromised educational attendance and attainment was a common outcome for parentified youth [27,29,32,33,35]. Forming relationships with others is sometimes difficult due to limited time, issues of trust and fear, for example, parentified girls avoided emotionally intimate relationships and did not accept support from others because they fear being parentified by support persons as a quid pro quo, meaning they will owe that person for any support received [31].

Theme 4. Suboptimal Mental and Physical Health among Parentified Youth.
Whereas some youth perceive and respond to parentified experiences with positive perceptions and coping strategies, other parentified youth experience anger, loneliness, resentment, feelings of being overwhelmed, and substance use [32,33,38]. Either by parent reports or youth self-reports, youth experienced suboptimal mental and physical health outcomes. Mental and physical health concerns include depression, anxiety, worry, toxic stress and trauma, dementia, domestic violence, substance abuse, displacement (war, foster care, orphaned), poverty, and HIV [25,31,34].
The suboptimal physical health outcomes of parentification could be exacerbated by co-occurring risk factors that many parentified youth experience. These co-occurring risk factors include a lack of access to affordable health care, healthy and stable housing, sufficient and quality water and food; harsh or abusive working conditions; domestic violence, and harsh parenting [25,27,35]. Addressing these needed resources and supports may prevent or mitigate some of the negative outcomes of parentification.

Theme 5.
Modifying Influences: Perception, Acknowledgement, Initial Competencies, and Supports Make the Difference between Floundering, Resilience, and Thriving.
The role of perceived fairness [36], being rewarded with statements of appreciation [28,36], positive relationships with parents [34], and receiving community, social, and service supports [27,34] influenced whether youth experienced floundering, surviving, resilience, or thriving outcomes. Youths' perceptions about their experience of parentification (e.g., fair, positive, unfair, loss) and the degree to which emotional and resource supports are provided (e.g., appreciation, food, housing) influence the trajectories they traverse-resilience and thriving or surviving and succumbing [36,86]. A lack of parental emotional support in childhood may take the form of parents dismissing or underplaying the children's parentification experience, and providing less warmth and support [86], assuming children should and will take on parenting responsibilities, which could dismantle children's trust in parents and other adults to care about them.
In contrast, small displays of appreciation for children's contributions to the family were related to positive outcomes for children (e.g., confidence) [28,36]. Among orphaned youth, some youth were reported to be resilient by going to night school to earn their degree and providing food and housing for siblings with community assistance; and these resilient youth had more social capital and support from community members, siblings, and peers [27]. Further, support from family and friends alleviated the negative impact of parentification, as did grit (personal strength) and sense-making or meaning-making, which depended on cultural context [33]. Initial competence and locus of control levels also help determine if children experience resilience [36].

Integrated Themes-Quantitative, Mixed, and Qualitative Studies
Theme alignment across qualitative, quantitative, and mixed methods studies is presented using an integration matrix (Table 6). This section focuses on four overlapping themes. Two quantitative themes map onto one qualitative theme, resulting in a total of four integrated findings.
First, across qualitative, quantitative, and mixed methods studies, positive outcomes in the form of resilience and positive coping were reported. Influential factors include the type of parentification-instrumental and the strengthening of relationships such as those with siblings. Although specific outcome constructs varied, these factors confer feelings of contributing to the household or family and opportunities to develop skills related to empathy, agency, esteem, and prioritization of others.
Second, youth employed several strategies to protect themselves or their parents from additional trauma. Mechanisms include not sharing or disclosing information with parents that would increase their worry and stress; distancing themselves from parents, even though they experience a tension between desired closeness and avoiding additional suffering. The family environment (e.g., maternal mental health or support) served to either protect or place youth at a greater risk for negative outcomes.
Third, parentified children experienced negative outcomes, including internalizing problems (e.g., depression, anxiety), externalizing problems (e.g., substance use, sexual risk-taking), and physical health (e.g., physical abuse, poor nutrition). Relationships between parentification and physical health and externalizing outcomes were more present in qualitative accounts than quantitative statistical tests.
Last, reports of positive or negative outcomes were influenced by afflicted youths' characteristics (e.g., self-differentiation, rejection sensitivity), attachment style, and perceptions of the benefits or fairness of adult responsibilities. These factors moderate how parentification relates to outcomes-protective/buffering or increasing risk. The buffering effect of other social supports (e.g., community members, teachers, and others) was reported to reduce the outcome severity youth experienced. Perceived benefits and fairness were associated with fewer negative mental health outcomes. These factors influenced the type of trajectory youth traversed.
Although not substantial enough to qualify as themes for qualitative or quantitative studies due to a dearth of longitudinal studies of multiple generations, there was mention of intergenerational impacts of parentification in several studies. For example, the mother's role reversal with her mother predicted the mother-toddler role reversal over and above attachment, suggesting a need for preventive interventions to address not only attachment disorganization, but also the mother's own history of parentification [74]. In another study, among a high-risk community sample of first-time mothers, maternal history of destructive parentification had an indirect effect on child externalizing behavior through warmth and responsiveness [82]. In yet another study, Tedgård and colleagues studied how parents' substance abuse translated into instrumental and emotional parentification of their children and the intergenerational impact on parenting. These parentified youth had an attribution bias toward danger and threat that resulted in being overprotective and overbearing of their own children [38]. • Instrumental parentification and perceived benefit linked to more positive outcomes • Too few studies examined the same positive construct for specific patterns, but across constructs, resilience emerged.
Across qualitative, quantitative, and mixed methods studies, positive outcomes in the form of resilience and positive coping were reported. Influential factors include the type of parentification-instrumental and the strengthening of relationships such as those with siblings. Although specific outcome constructs varied, these factors confer feelings of contributing to the household or family and opportunities to develop skills related to empathy, agency, esteem, and prioritization of others.

Self-Preservation Protective
Mechanisms-Self-sacrifice and Distancing • Youth utilized several strategies to protect themselves or their parents from additional trauma.
Exploration of moderators to explain effect heterogeneity: exacerbating effect of maternal depression and protective roles of disclosure of worries, social support, and religious service attendance (Quant Theme 5).
• Disclosure of worries, social support, and religious service attendance either buffered the negative outcomes of parentification or facilitated the positive outcomes. • Maternal depression exacerbated the negative outcomes of parentification.
Youth employed several strategies to protect themselves or their parents from additional trauma. Mechanisms include not sharing information with parents that would increase their worry and stress; distancing themselves from parents even though they experience a tension between desired closeness and avoiding additional suffering. The family environment (e.g., maternal mental health or support) served to either protect or place youth at greater risk for negative outcomes. Parentified children experienced negative outcomes in the areas of mental health or internalizing problems (e.g., depression, anxiety), externalizing problems (e.g., substance use, sexual risk-taking), and physical health (e.g., physical abuse, poor nutrition). Relationships with physical health and externalizing outcomes were more present in qualitative accounts than in quantitative statistical tests. The report of positive or negative outcomes was influenced by afflicted youths' characteristics (e.g., self-differentiation, rejection sensitivity), attachment style, and perceptions of the benefits or fairness of these adult responsibilities. These factors mediate how family risk factors relate to outcomes-protective/buffering or increasing risk. The buffering effect of other social supports (e.g., community members, teachers, and others) was reported to reduce the outcome severity youth experienced. Perceived benefits and fairness were associated with fewer negative mental health outcomes. These factors influenced the type of trajectory youth traversed (e.g., severity degree and positive/negative nature).

Summary of Findings
A majority of studies relied on adult retrospective accounts, participants tended to be mostly female or college students, and there was representation from six continents, demonstrating this is a global phenomenon. This systematic review revealed that parentification may have both positive and negative impacts on coping, family relationships, mental and physical health, and human and social capital. Altogether, this work implies that parentified youth need support, especially those experiencing early-onset parentification when their cognitive and emotional systems are still immature. The early intervention of social, emotional, and cognitive support could mitigate adverse impacts stemming from sources of parentification, such as parent mental and physical health challenges, exposure to domestic violence, war, and unstable housing.
A few overlapping themes emerged from quantitative and qualitative studies. These include positive outcomes of enhanced coping and resilience; negative outcomes of suboptimal mental health and problem behaviors; protective effects of perceived benefits and social support; and mediating roles of self-preservation mechanisms. The results indicate that parentification is a complex process in which linear explanations may not be sufficient and different mediators and moderators should be considered.
We have found parentified children experience various suboptimal outcomes in adulthood [15][16][17]. One reason is that these children perceived their obligatory adult roles negatively, as unfair and "robbing" them of their childhood, and experienced stress, role overload, and resentment. Another explanation is that parentified youth and their siblings are expected to prematurely assume adult-like responsibilities during developmental stages (e.g., childhood, adolescence) marked by immature brain development, which is not fully optimized until age 25 [162,163]. In addition, brain development is negatively impacted by early and chronic stressors [164], such as parentification. These youth may have limited opportunities to learn, observe, and utilize positive coping strategies typically derived from parental influences. Rather, parentified youth are disadvantaged by a lack of emotional and instrumental support from parents who cannot meet their emotional and psychological needs [25,31].
Resilience among parentified youth is not as well documented, in part due to quantitative study focus bias on negative outcomes. Qualitative studies in this review, however, provide insight into how resilience develops from parentification experiences. The studies suggest that it may be important to recognize that some parentified youth may emerge into adulthood relatively unscathed or robust against adversities they experienced. Unfortunately, this area of the parentification literature is scarce, and we currently know little about the positive outcomes of parentification and the mechanisms underlying resilience.

The Role of Culture and Context
Across the 95 studies, 42% represented non-US countries: Nine of the 14 (64%) qualitative or mixed studies and 31 of the 81 (38%) quantitative studies. The sources, meanings, and practices of parentification may vary across these contexts. For example, countries with higher prevalence of infectious diseases such as HIV/AIDS, that increase the likelihood of parental death, may result in fewer adult supports for children who are orphaned or lose one parent. Countries experiencing pressures for flight may result in refugee families whereby children are expected to serve as language brokers and workers because it is easier for them to learn the language of new countries. These language brokers may experience both heightened instrumental and emotional parentification that may be further exacerbated depending on the child's age. For example, if youth are particularly young (pre-pubescent; a time when learning multiple languages is easier), they may be asked to translate adult conversations that require an emotional maturity they do not have, making these conversations particularly difficult for youth.
The fact that the parentification inventory has been translated into many languages and has demonstrated reliability and validity in studies across continents-North America, Australia, South America [165], Africa [166], Europe [167], and Asia [55], suggests the phenomenon of parentification is global. Policies and resources provided to parentified children also vary by countries [168]. Although some studies reviewed here did discuss and interpret their findings in light of the cultures where they were embedded, the crossculture comparison was not explicitly tested because studies were generally carried out in one country and precluded such empirical tests of cultural differences. For example, these studies had discussed the readiness of youth taking on care-giving roles in a more collectivistic culture in Korea [54], the implications of relatively higher degree of emotional interdependence in Turkish culture that affects parentification [55], and the impacts of historically collective and communal focus versus the modern individualistic orientation in Jewish Israel culture on parentification and enmeshment [63]. In short, none of the studies in this review explicitly conducted cross-national comparisons of parentification experiences and associations with outcomes, a gap future studies should address.

Dimensions of Parentification and Promising Research Directions
One challenge involved in integrating findings across quantitative studies concerns the dimensions of parentification being investigated. Many studies used standardized measures either just on the functions parentification served (emotional versus instrumental) or just on roles (sibling-, parent-, or spouse-focused). Results were not directly comparable across studies focused on function versus roles of parentification.
Emergent quantitative findings suggested emotional parentification may be more detrimental than instrumental parentification, in that emotional parentification was consistently linked with depressive symptoms and a broader spectrum of internalizing and externalizing problems. In contrast, no clear pattern for the link between role-based parentification measures and outcomes emerged. Emotional parentification may be more overwhelming for youth, who at the age of parentification onset, were not developmentally equipped with the necessary skills to adequately cope with the imposed emotional demand. As seen in qualitative studies, parentified children resort to suboptimal coping strategies to protect themselves and their parents from additional emotional trauma (e.g., not sharing about stressors, self-sacrifice, and distancing).
Instrumental parentification, on the other hand, may be more predictable and prescribed (e.g., working hours, household responsibilities), making it more likely to be within the youth's capabilities, even if it is demanding in time and effort. From an evolutionary psychology perspective, youth taking on instrumental responsibility (e.g., caring for younger siblings, doing house chords) to contribute to family economy reflects our prehistoric past as a species, since children and adolescents took on varies duties in tribal societies that make up most of the human cultures. Being able to make contributions to the family in an instrumental way may be psychologically favorable for self-esteem development; hence, explaining why instrumental parentification has not been found consistently linked to a host of negative outcomes. The mixed findings for role-based parentification may be explained because any of them (e.g., sibling-or parent-focused) could be any combination of emotional or instrumental parentification elements. If the association with mental health indeed resides in emotional parentification and it was not directly measured by a role-based instrument, then it is not surprising to see mixed findings for role-based parentification. Given that emotional parentification was more consistently associated with suboptimal mental health, focusing on emotional versus instrumental parentification may present new research avenues for identifying mechanisms, such as focusing on emotion regulation, positive coping, and social support as mediators for emotional parentification [27,32,34]. In contrast, role-based parentification likely limits further formulation of mechanisms, as it may be difficult to draw on existing psychological theories to hypothesize why and how parent-focused and sibling-focused parentification have differential outcomes, and how they affect different mediators to exert their impact on mental health outcomes. We argue that the studies operationalizing parentification through emotional versus instrumental dimensions will advance the study of parentification in a more fruitful direction.

Perceptions and Perceived Benefits of Parentification
The significance of perceptions emerged from both quantitative and qualitative studies. As revealed in the results, perceived benefits of parentification were generally linked with more positive outcomes and less negative outcomes across studies. We want to highlight an important conceptual issue that has not received enough consideration: should perceived benefits be treated as a dimension of parentification equally as function-based parentification or role-based parentification, or should it be conceptualized as a moderator or mediator of the impact of parentification on outcomes? Currently, only the former conceptualization has been explored, largely because the perceived benefits of parentification have been measured as a subscale in parentification measures. Researchers typically conduct main effect analyses to examine how perceived benefits were linked to outcomes, and no existing quantitative research examined it as a moderator for the effects of parentification. Perceived benefits of parentification being treated as a subscale of parentification may have hindered efforts to explore how perceptions are a moderator or mediator for the parentification effects.
The reverse causation issue was not addressed and rarely discussed in research that found positive outcomes of perceived benefits. Positive outcomes may cause higher ratings of perceived benefits of parentification in quantitative research and more positive perceptions of parentification in qualitative studies. This issue is particularly salient in retrospective research where individuals may interpret their past parentification experiences differently. We firmly believe it is critical to understand individuals' perception of benefits from their experiences, yet this should be separated from measuring the parentification experience itself. By clearly conceptualizing perceived benefits as separate from parentification and identifying ways of dealing with the potential reverse causation issue, we can better advance our understanding of parentification and its heterogenous impacts.

Adverse and Positive Life Outcomes
Although over three-quarters of the quantitative studies focused on the negative outcomes of parentification, surprisingly few consistent findings could be identified. Except for depressive symptoms, broad spectrum internalizing and externalizing problems, and substance use, most constructs were studied only a handful of times and some constructs were almost exclusively studied by the same research groups. To advance an area of knowledge, individual studies should build on the existing literature and situate themselves well into the larger literature. Yet, many "one-off" studies were observed, the findings of which were neither replicated nor carried forward. Future research is needed to address these problems and meaningfully build the parentification literature with more depth and consistent use of outcome measures and constructs rather than casting a wider net by adding more "one-off" outcomes. For example, there were consistent findings for emotional parentification and depressive symptoms. Future research can build upon these replicated findings by identifying various pathways of how emotional parentification exerts its effect through physiological or psychological mediators (e.g., stress cortisol reactivity, resting heart rate variability for emotion regulation capacity, coping strategies). Focusing on the how and why will move the field further forward and provide insights into supports, programs, and interventions needed to mitigate the negative outcomes that these parentified youth experienced.
It is clear from this review that resilience remains understudied. Less than 15% of the quantitative studies included positive or resilient outcomes. Although most qualitative studies were open-ended, none explicitly posed research questions addressing the positive outcomes of parentification. Nonetheless, studies presented promising findings that parentification or dimensions of parentification were linked with beneficial outcomes such as coping skills, social competence, or general well-being. The "one-off" nature of parentification research is even more pronounced when looking at studies that measured positive outcomes. Parentification research can benefit from the positive youth development and positive psychology approaches by including resilience and positive coping into their research designs.

Adopting a Developmental and Systems Perspective
Only a handful of studies examined parentification from the entire family and at multiple time points. Fathers and sons are undoubtably understudied within the parentification literature. From a systems perspective, examining dyads, triads, or the family unit is important given the ecological and biosocial implications of parentification within a family system. Further, from a behavioral genetics perspective [169], children raised in the same household have shared and non-shared experiences and biological vulnerabilities (genetic predispositions) that are important considerations in understanding the consequences of parentification. The source of parentification may be parents' mental health conditions, such as depression, schizophrenia, and narcissism [31,34], implicating biological and social vulnerabilities for children to experience similar or related conditions. With the increased prevalence of restructured families where stepsiblings, half-siblings, and adopted siblings are more commonplace, the context and implications of parentification may vary. Future studies could explore parentification as a function of different levels of sibling relatedness, which contributes to differentiating gene (biological) and environmental (behavioral) influences, helping to clarify shared genetic vulnerability and environmental influences.
From a socialization perspective, studying parent-child dyads may reveal how parentification impacts and is impacted by relationship quality. As noted above, one theme that emerged from qualitative studies was how parentified children withheld information from parents for reasons of protecting parents from additional worry and self-preservation through maintaining independence and emotional distancing. Further, parentified youth are disadvantaged by a lack of emotional and instrumental supports from parents who cannot meet their emotional and psychological needs [25,31]. The role of parent-child relationship quality in predicting, being predicted by, and moderating parentification outcomes was not directly or explicitly measured in the included studies, and is an area for future research, especially longitudinal designs.
In addition, prospective studies are important given the developmental implications of parentification. Qualitative study findings highlight compromised educational attainment as a negative outcome of parentification, yet most quantitative studies that we reviewed were conducted with college-going samples and cross-sectional, which together raise further concern that youth who did not go to college are underrepresented. Longitudinal studies would capture important characteristics and contexts of parentification, including onset timing, duration of experience, short-term vs. long-term adaptations, sleeper effects, and retrospective reframing. These factors should be considered to better understand how the effects of parentification may vary based on developmental periods.
Adopting a developmental and systems perspective to research parentification requires a large sample size, longitudinal design, and a focus beyond college students, which is lacking in the current research. For example, the majority of quantitative studies had relatively small sample sizes (less than 300 participants), suggesting statistical power may be a concern already, even with a cross-sectional design, for the number of dimensions studied, multiple tests run, and sophisticated analyses including latent variable, mediation, and moderation modeling.

Youth Measure and Non-Linear Modeling and Explanation Considerations
From a measurement perspective, measures specific to youth reports warrant additional refinement, including improved reliability. These measures have been adapted from adult versions and few studies have utilized them. It may be useful to use cognitive interviewing and mixed methods approaches to develop these youth measures to ensure their reliability.
Related to modeling, considering non-linear explanations and descriptions of parentification is recommended. A linear explanation of the relationship between parentification and outcomes does not capture the nuances of the phenomenon. Parentified individuals report both positive and negative perceptions and outcomes. For the majority of qualitative studies (n = 10), both positive and negative outcomes were found (e.g., parentified children reported both personal growth and high school dropout [29]). Callaghan and colleagues refer to this as an "empowering and constraining" duality (p. 662) [26]. Further, as Chee and colleagues highlight [28], curvilinear relationships between functioning and parentification need to be explored, whereby those not at extremes adapt well. A different approach warranting future exploration in quantitative research is to consider the combined effects of various parentification domains. For instance, youth reporting high emotional coupled with high instrumental or low emotional coupled with high instrumental may report more positive outcomes compared to youth experiencing high emotional-low instrumental parentification. There may be a buffering effect of instrumental parentification (associated with leadership skills, esteem from responsibility fulfillment, and achievement) not otherwise captured if we modeled dimensions of parentification separately. The quantitative studies reviewed rely heavily on linear modeling and assumptions, which may need to be relaxed to ensure a more holistic understanding of the parentification phenomenon.

Strengths and Limitations
Our review has several limitations. First, only one mixed-method study was identified with our inclusion criteria. This necessitated its grouping with qualitative studies and limited our ability to contrast this study design with quantitative and qualitative studies. Second, although 81 quantitative studies were identified and included in the review, about 20% of these studies contributed less weight to the synthesized findings because of the heterogeneity in outcome constructs, where outcomes were only investigated once or twice. Third, conclusions rest largely on a majority of studies that used cross-sectional designs and employed non-probability-based sampling. This precluded causal inference and may introduce selection effects. Nonetheless, we identified several meaningful gaps and important directions for future studies.
These limitations should be balanced with the strengths and significance of our review. It is the first systematic literature review to examine parentification studies with outcomes. This study was inclusive of qualitative, mixed, and quantitative studies from across the world and utilized three databases to ensure comprehensiveness. MMAT provided additional analysis of study bias and quality, evidencing that the risk of bias is minimal. The use of mixed methods integration analyses enabled the comparison of themes across different study designs. Five qualitative and four quantitative themes overlap, suggesting consistency and congruity in findings despite outcome measures varying. Policy and programmatic recommendations were gained.

Conclusions
The study of parentification has public health significance and practical implications. Although parentification may confer positive outcomes, its relationship with resilience has been grossly understudied. The predominantly negative consequences of parentification are far-reaching, including the parentified individual, the immediate family, and intergenerationally, speaking to the need to further understand its causes, consequences (both positive and negative), typologies, and intervention. Many parentified youth experience compromised human, social, and financial capital that strongly predict long-term outcomes for these youth and their families, including under-and un-employment and mental and physical health disparities. Identifying and providing needed support early and often is needed.
In the US, demographic changes related to family structure (e.g., increasingly unstable, complex, and "fragile" families), economic demands, and health disparities have and will further increase the number of parentified youth. Factors contributing to this role strain include household poverty, siblings with additional needs, parental illness, parental divorce or separation, and other stressors affecting youth and families. In the US, COVID-19 has disproportionately negatively impacted marginalized racial and socioeconomic strata with Black adults and youth more likely to experience compromised household income, housing stability, and mental and physical health, [170,171] which are sources of parentification pressures for children. These pressures are not isolated to the US, but rather are globally relevant with additional concerns in other countries including wars, humanitarian atrocities, and epidemics that have far-reaching impacts and disproportionately impact the most vulnerable and lowest-resourced families. These circumstances create a contagion of risk factors with youth particularly vulnerable to parentification, whereby children assume adult responsibilities. This study elevates the importance of this topic and provides feasible and actionable substantive and methodological recommendations for future studies. Further, this review highlights (a) the need for social support for these youth, and (b) investment in programs and policies that invest in youths' ability to increase their human (e.g., education) and social (e.g., peer relationships) capital and physical health (e.g., medical, nutrition) while helping to provide for their households (e.g., rent, food).
The implications of parentification expand beyond the parentified individual (e.g., psychological, cognitive, and physical health outcomes) to the family of origin (e.g., sibling outcomes) and future generations through an intergenerational transmission [1]. The fact that the prevalence of parentification in the US is unknown, yet is greater than 30% of youth in Poland and impacts family members-horizontally to siblings and vertically to offspring-further amplifies the public health need for studying this phenomenon, negative sequelae, and opportunities to promote resilience for youth and families.

Data Availability Statement:
No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest:
The authors have no known conflict of interest to disclose.