Access to this full-text is provided by Wiley.
Content available from Journal of Adolescence
This content is subject to copyright. Terms and conditions apply.
Received: 21 February 2024
|
Accepted: 5 May 2024
DOI: 10.1002/jad.12347
RESEARCH ARTICLE
A cross‐sectional investigation into the role of intersectionality
as a moderator of the relation between youth adversity and
adolescent depression/anxiety symptoms in the community
Laura Havers
1
|Kamaldeep Bhui
2,3,4
|Ruichong Shuai
1
|Peter Fonagy
5,6
|
Mina Fazel
7
|Craig Morgan
8,9
|Daisy Fancourt
10
|Paul McCrone
11
|
Melanie Smuk
12
|Georgina M. Hosang
1
|Sania Shakoor
1
1
Centre for Psychiatry and Mental Health, Wolfson Institute of Population Health, Queen Mary, University of London, London, UK
2
Department of Psychiatry, Nuffield Department of Primary Care Health Sciences, and Wadham College, University of Oxford, Oxford, UK
3
Oxford Health and East London NHS Foundation Trusts, Oxford, London, UK
4
World Psychiatric Association Collaborating Centre, Oxford, UK
5
Anna Freud National Centre for Children and Families, London, UK
6
Research Department of Clinical, Educational and Health Psychology, University College London, London, UK
7
Department of Psychiatry, University of Oxford, Oxford, UK
8
Health Service and Population Research, Institute of Psychology, Psychiatry & Neuroscience, King's College London, London, UK
9
ESRC Centre for Society and Mental Health, King's College London, London, UK
10
Department of Behavioural Science and Health, University College London, London, UK
11
Institute for Lifecourse Development, University of Greenwich, London, UK
12
Centre for Genomics and Child Health, Blizard Institute, Queen Mary, University of London, London, UK
Correspondence
Laura Havers and Sania Shakoor, Centre for
Psychiatry & Mental Health, Wolfson Institute of
Population Health, Queen Mary, University of
London, Yvonne Carter Bldg, 58 Turner St,
London E1 2AB, UK.
Email: l.havers@qmul.ac.uk and
sania.shakoor@qmul.ac.uk
Funding information
Cross Council UK Research and Innovation,
Grant/Award Number: MR/W002183/1; National
Institute for Health Research Applied Research
Collaboration Oxford and Thames Valley at
Oxford Health NHS Foundation Trust; Oxford
Health NIHR Biomedical Research Centre; The
National Lottery Community Fund (HeadStart)
Abstract
Background: Adolescents exposed to adversity show higher levels of depression and
anxiety, with the strongest links seen in socially/societally disadvantaged individuals
(e.g., females, low socioeconomic status [SES]), as well as neurodivergent individuals.
The intersection of these characteristics may be important for the differential
distribution of adversity and mental health problems, though limited findings pertain
to the extent to which intersectional effects moderate this association.
Methods: Combined depression/anxiety symptoms were measured using the
emotional problems subscale of the Strengths and Difficulties Questionnaire in
13–14‐year‐olds in Cornwall, United Kingdom in 2017‐2019. In a cross‐sectional
design (N= 11,707), multiple group structural equation modeling was used to
estimate the effects of youth adversity on depression/anxiety symptoms across eight
intersectionality profiles (based on gender [female/male], SES [lower/higher], and
traits of hyperactivity/inattention [high/low]). Moderation effects of these character-
istics and their intersections were estimated.
Journal of Adolescence. 2024;96:1304–1315.1304
|
wileyonlinelibrary.com/journal/jad
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the
original work is properly cited.
© 2024 The Authors. Journal of Adolescence published by Wiley Periodicals LLC on behalf of Foundation for Professionals in Services to Adolescents.
Laura Havers and Sania Shakoor are joint corresponding authors.
Georgina M. Hosang and Sania Shakoor are joint last authors.
Results: Youth adversity was associated with higher levels of depression/anxiety
(compared to an absence of youth adversity), across intersectional profiles. This effect
was moderated by gender (stronger in males; β= 0.22 [0.11, 0.36]), and SES (stronger
in higher SES; β= 0.26 [0.14,0.40]); with indications of moderation attributable to the
intersection between gender and hyperactivity/inattention (β= 0.21 [−0.02,0.44]).
Conclusions: Youth adversity is associated with heightened depression/anxiety across
intersectional profiles in 13–14‐year‐olds. The stronger effects observed for males, and
for higher SES, may be interpreted in terms of structural privilege. Preliminary
findings suggest that vulnerability and resilience to the effects of youth adversity may
partially depend on specific intersectional effects. Importantly, the current results
invite further investigation in this emerging line of inquiry.
KEYWORDS
ACEs, adversity, anxiety, depression, intersectionality, moderation
1|INTRODUCTION
1.1 |Youth adversity and mental health problems
Youth adversity encompasses stressful and traumatic experiences during childhood or adolescence, occurring both within the
home (e.g., domestic violence, problem drug/alcohol use in the family, often referred to as adverse childhood experiences
[ACEs] [Felitti et al., 1998]), as well as outside the home (e.g., bullying‐victimization) (Kalmakis & Chandler, 2014). This
adversity is strongly linked to mental health difficulties throughout the life course (Bellis et al., 2019). For instance, 25%–40%
of adult depression and anxiety cases are estimated to be attributable to adverse experiences in youth (Bellis et al., 2019).
Moreover, exposure to youth adversity has proximal negative effects on mental health during childhood and adolescence
(Scully et al., 2020). With evidence suggesting that half of all lifetime mental health problems emerge by adolescence (Caspi
et al., 2020; Kessler et al., 2005), understanding the factors that may exacerbate or mitigate the effects of youth adversity on
adolescent mental health is crucial. This knowledge can inform the development of more targeted clinical and community‐
level interventions during this critical developmental stage (Griner & Smith, 2006).
1.2 |Background characteristics
While youth adversity and mental health problems affect people across demographic divides, disadvantaged and/or
minoritized societal and neurodivergent groups experience higher rates of youth adversity and elevated levels of mental
health problems, compared to their counterparts. This includes, for example, individuals diagnosed with neurodivergent
conditions, including autism and attention deficit hyperactivity disorder (Lai et al., 2019) and individuals with heightened
levels of neurodivergent traits such as hyperactivity/inattention (Craig et al., 2020), those from lower socioeconomic status
(SES) backgrounds (Walsh et al., 2019), females (Campbell et al., 2021), and members of underrepresented ethnic groups
(Assini‐Meytin et al., 2022).
The mechanisms linking these background characteristics with youth adversity and mental health problems are likely to
be multifaceted and complex. For instance, in individuals diagnosed with and or self‐identifying as neurodivergent, as well as
those with heightened traits of neurodivergence –an individual's response to as well as from social/societal systems tailored to
neurotypical functioning are likely to contribute to their experiences of adversity and mental ill health (Pantazakos &
Vanaken, 2023). The specific mechanisms underlying these links are likely to differ to those underlying, for instance, ethnic
marginalization and mental health problems (Williams et al., 2003), and observed gender differences in manifestations of
mental health (Blakemore et al., 2010; Udry, 2000). Importantly, however, in the context of the current study –the collective
findings strongly indicate that minoritized, disadvantaged, and neurodivergent individuals, are at heightened risk of
experiencing adversity and mental health problems compared to their (socially/societally advantaged, and or neurotypical)
counterparts.
1.3 |Multiple background characteristics and intersectionality
Further to these observations relating to background characteristics reflective of societal/social minoritization and
neurodivergence in isolation –adolescents with multiple characteristics of disadvantage exhibit lower levels of wellbeing
JOURNAL OF ADOLESCENCE
|
1305
compared to those with less disadvantage (Kern et al., 2020). Notably, in the context of the current study, because the impact
of youth adversity on mental health problems has been found to be stronger in disadvantaged groups compared to their
advantaged counterparts (Scully et al., 2020)–vulnerability to mental health problems in the face of youth adversity may be
intensified in individuals with multiple characteristics of disadvantage (e.g., a neurodivergent female from a low SES
background) (Ghavami et al., 2016).
Importantly, existing empirical findings highlight the emergent effects that may arise when considering the influence of
multiple, interacting characteristics on youth adversity and mental health. For example, youth adversity was found to be
more prevalent among nonimmigrant Hispanic compared to nonimmigrant White young people in the USA, with greater
disparity among high compared to low SES families (Slopen et al., 2016). The findings from such empirical studies
demonstrate the importance of considering markers of disadvantage, like low SES, in the context of other characteristics.
However, only one study to our knowledge has investigated the extent to which the intersection of multiple characteristics
moderates the effect of youth adversity on mental health problems in youth (Havers et al., 2024). This study reported
suggestive evidence to indicate differential effects of youth adversity on change over time in depression/anxiety symptoms for
males and females dependent on SES and traits of hyperactivity/inattention.
These emergent, multiplicative,effects can be contextualized in the broad analytic framework of intersectionality,
originally developed in response to the specific systemic oppression encountered by women of color in the United States
(Crenshaw, 1991). This framework posits that the intersection of an individual's characteristics (e.g., gender identity,
ethnicity, SES, neurodivergence, sexual orientation) uniquely situates them within a complex system of structural power
hierarchies, reflective of oppression and privilege (Cole, 2009). This intersection captures a social and societal position
beyond disadvantage, or advantage, associated with individual characteristics (such as gender, or ethnicity) both in isolation
and additively (Bowleg, 2012; Crenshaw, 1991). Importantly, as a critical period for the development of social identity
(Lerner & Galambos, 1998), adolescence is a pertinent window for investigating associations between youth adversity and
mental health in the context of intersectionality (Ghavami et al., 2016).
This inquiry is important, because investigation into the differential effects of youth adversity on adolescent mental health
that may in part be dependent on an individual's intersectional position can contribute to developing a stratified framework
for explicating the structural mechanisms that connect youth adversity and adolescent mental health problems. This in turn
can inform the subsequent design of interventions tailored towards adolescents who are most vulnerable to the negative
consequences of youth adversity, and importantly, contribute to an accumulation of evidence that can inform policy changes
at a structural level (Hankivsky et al., 2014; Patil et al., 2018).
1.4 |Current study
Recognizing that the theoretical and judicial origins of intersectionality are rooted in gender and race (Crenshaw, 1991), the current
study makes use of available measures in an existing cohort study to explore intersectionality conceptualized more expansively
(Cole, 2009) in terms of gender (binary), SES, and traits of hyperactivity/inattention (as an index of neurodivergence) in a
representative sample of adolescents living in a rural/coastal region of the United Kingdom (Hosang et al., 2023).
Of note, an emerging body of literature documents the advantages of utilizing a multilevel modeling framework for
addressing research questions pertaining to intersectionality, in which intersectional identity is conceptualized as being a
cluster, or strata‐level characteristic (Evans et al., 2018; Merlo, 2018). Inherent in this framework from an empirical
standpoint is the requirement for a reasonable minimum number of clusters, balanced with the number of within‐cluster
observations and model complexity, to estimate models with desirable properties (e.g., consistency, limited bias) (McNeish &
Stapleton, 2016; Van de Schoot & Miocević,2020). Because intersectionality in the current study was represented by a
relatively low number of intersectionality profiles (i.e., 8; see Methods), we utilize a multiple group structural equation
modeling framework–with the estimation of compound parameters specified to capture moderation attributable to
intersectionality.
The current study addresses two research questions: (1) Does the effect of youth adversity on depression/anxiety vary
across intersectionality profiles? It was hypothesized that the effect of youth adversity would vary across intersectionality
profiles, and that the strongest effect would be observed for the profile reflecting multiple disadvantage (females, lower SES,
high traits of hyperactivity/inattention). (2) Is the effect of youth adversity on depression/anxiety moderated by gender, SES,
and hyperactivity/inattention, and their intersections? It was hypothesized that the effect of youth adversity would be
moderated by these characteristics and their intersections, though no specific predictions were made regarding the direction
of the intersectional moderating effects.
While the cross‐sectional nature of the study design means that the temporal impact of youth adversity on mental health
cannot be elucidated–importantly, the level of detail inherent in investigating moderation through an intersectional lens can,
even in the absence of temporal precedence, shed light on the conditions under which youth adversity and mental health
problems in adolescence are most (and least) strongly linked.
1306
|
HAVERS ET AL.
2|MATERIALS AND METHODS
2.1 |Participants
Participants were drawn from the HeadStart Cornwall study (Deighton et al., 2019; Hosang et al., 2023). Year 9 pupils in all
31 state‐maintained schools in Cornwall, United Kingdom, were invited to participate. Age data is not available for this
sample, but the average age for Year 9 pupils in the United Kingdom is 13–14‐years. The sample (N= 12,067) comprises data
pooled from collections undertaken in 2017 (N= 4,269), 2018 (N= 4,194), and 2019 (N= 3,604). The final sample used in the
main analysis was N= 11,707 individuals with youth adversity and intersectionality profile data (detailed below).
All schools participated, except for one in 2017. Pupils assented ahead of participation at their school via an online portal.
Parents were provided with information regarding the study ahead of data collection and consent was assumed unless
children were opted out. Both pupil assent and parental consent was required for participation (Deighton et al., 2019).
2.2 |Measures
Combined depression/anxiety symptoms were measured using the 5‐item self‐report emotional problems subscale of the
Strengths and Difficulties Questionnaire (SDQ) (items listed in Supplementary Table 7; note) (Goodman, 1997). The items
capture usual experiences over the last 6 months (e.g., “I am often nervous in new situations”). Responses are made on
3‐point rating scale (“Not true”,“Somewhat true”,“Certainly true”). Item‐level data was used for the main analyses (see
Statistical Analyses). Total scores (0–10) were used for the reporting of measured‐variable level descriptive statistics.
Youth adversity was measured in two ways. First, through the SDQ peer problems subscale item on bullying‐victimization: “Other
children or young people pick on me or bully me”. Second, using local government data from the Supporting Families program
(https://www.gov.uk/government/publications/supporting-families-programme-guidance-2022-to-2025). In this program, youth
adversity is indicated where families experience any of the adversities listed in Supporting Information S1: Supplementary
Materials 1(including household substance misuse, household domestic violence, homelessness). Data regarding the number and
type of local government recorded adversity was not available for this sample, therefore a binary variable was created to indicate the
presence/absence of either bullying‐victimization and or local government‐recorded adversity.
Gender and SES data were sourced from School Census records. Gender was recorded as “female”or “male”. Free school
meals eligibility was used to index “lower SES”or “higher SES”.Hyperactivity/inattention traits were measured using the SDQ
hyperactivity/inattention subscale (five items). Total scores of 0–6 were used to indicate “low”levels, and 6–10 to indicate
“high”levels, following scoring guidelines (Goodman, 1997).
Participants were classified into one of eight intersectionality profile groups. Group assignment was based on gender
(female/male), SES (lower/higher), and traits of hyperactivity/inattention (high/low). For example, “female, lower SES, low
hyperactivity/inattention”.
2.3 |Statistical analyses
Before the main analyses, confirmatory factor analysis was used to evaluate the measurement properties of depression and
anxiety conceptualized as a common latent factor, indicated by the five SDQ emotional problems items. These analyses were
conducted for the whole sample, and separately for each intersectionality profile group. Following this, measurement
invariance analysis was conducted to assess the extent to which the measurement of depression/anxiety as a common factor is
sufficiently consistent across intersectionality profile groups (Supporting Information S1: Supplementary Materials 2).
Recommendations for modeling ordered categorical response data were used (Svetina et al., 2020; Wu & Estabrook, 2016).
To address the first research question, the constrained measurement model from the analysis above was imbedded in a
multiple group structural equation model. The necessary adjustments for scaling are detailed in Supporting Information S1:
Supplementary Table 9; note. In this model, the latent factor of depression/anxiety was regressed on the binary youth
adversity variable, and on dummy variables representing cohort effects (chosen for comparison to the 2019 cohort) (see
Supporting Information S1: Supplementary Materials 3:Mplus script; and Figure 1: schematic diagram).
A model with youth adversity regression slopes freely estimated for each intersectionality profile group was compared to a model
where they were constrained to equality. A better fit of the unconstrained model would suggest, at an omnibus level, that the relation
between youth adversity and depression/anxiety varies across intersectionality profiles. Wald statistics were used to evaluate the
significance of the difference between the unstandardized regression estimates between intersectionality groups. The false discovery
rate method was applied (at α= .05) for multiple testing correction (Benjamini & Hochberg, 1995).
Diagonally weighted least squares estimation was used to model the ordered categorical response data, using pairwise
present data. Model fit was primarily assessed using the comparative fit index (CFI), root mean square error of
JOURNAL OF ADOLESCENCE
|
1307
approximation (RMSEA), and standardized root mean square residual (SRMR). Acceptable fit was broadly indicated by
CFI > 0.95, RMSEA < 0.08, and SRMR < 0.06 (Hu & Bentler, 1999; Marsh et al., 2004). To accommodate missing data (under
the assumption that it was missing at random), the final model was rerun with multiply imputed item‐level data and youth
adversity data from 10 datasets, and the results are reported as sensitivity analyses.
To address the second research question, compound parameters of the weighted least squares parameter estimates from
the final analysis model (from research question one) were estimated. These compound parameters were specified to
approximate the extent to which the parameters reflecting the effect of youth adversity on depression/anxiety are moderated
by the individual characteristics of gender, SES, and hyperactivity/inattention, and by interactions between these
characteristics (reflecting intersectionality effects, e.g., gender by SES) (Figure 1). Standard errors and confidence intervals for
these compound parameters were estimated from 1000 bootstrapped draws.
Rsoftware (version 4.2) was used to calculate descriptive statistics. Mplus (version 8.8) was used to conduct structural
equation modeling.
3|RESULTS
Table 1details the sample characteristics. More than a third of participants (36.37%, 95% CI = 35.50, 37.25, N= 4,267)
experienced youth adversity. The sample‐wide mean for measured (nonlatent) depression/anxiety symptoms was 4.15
(SD = 2.69).
Rates of youth adversity and mean depression/anxiety symptoms for the individual characteristics, and for each
intersectionality profile, are shown in Supporting Information S1: Supplementary Tables 1and 2, respectively. Both the
highest prevalence of youth adversity and the highest mean for depression/anxiety symptoms were observed in
the intersectionality profile of female, lower SES, and high hyperactivity/inattention. Rates of youth adversity were lowest
for the intersectionality profile of female, higher SES, and low hyperactivity/inattention. The lowest mean for depression/
anxiety symptoms was observed in the intersectionality profile of male, higher SES, and low hyperactivity/inattention.
The common factor model for depression/anxiety showed adequate fit for the whole sample, and for each
intersectionality profile. The average weighted omega estimate across intersectionality profiles was ω= 0.80, 95%
CI = 0.80, 0.81. Model fit statistics and the reliability (omega) estimates derived from these models are reported in
Supporting Information S1: Supplementary Table 3. In the analysis of measurement invariance, threshold and loading
FIGURE 1 Schematic diagram of multiple group structural equation model. SES, socioeconomic status. Intersectional profile (depicted at the center of
the Venn diagram) is the grouping variable in the multiple group model, where all parameters inside the box are estimated separately for each
intersectionality profile (e.g., the profile of female, lower SES, low hyperactivity/inattention). Compound parameters are estimated, reflecting the extent to
which the individual characteristics (e.g., gender) and their intersections (labelled with superscript letters in the diagram: a) gender by hyperactivity/
inattention, b) hyperactivity/inattention by SES, and c) gender by SES) moderate the path from youth adversity to the depression/anxiety latent factor. The
five items of the Strengths and Difficulties Questionnaire emotional problems subscale are specified as indicators of the latent factor of depression/anxiety.
The model depicted schematically in this figure provides the framework for addressing the research questions detailed in the Methods section.
1308
|
HAVERS ET AL.
invariance of the common factor model across intersectionality profiles was supported (Supplementary Table 4). Parameter
estimates from the constrained thresholds and loadings measurement model are included in Supporting Information S1:
Supplementary Table 5.
3.1 |Research question 1: Does the effect of youth adversity on depression/anxiety vary across
intersectionality profiles?
The unconstrained multiple group model demonstrated acceptable fit (Supporting Information S1: Supplementary Table 6).
Across all intersectionality profiles, youth adversity was associated with higher average levels of depression/anxiety,
compared to an absence of youth adversity (Figure 2).
At an omnibus level, the effect of youth adversity on depression/anxiety varied across intersectionality profiles, indicated
by the better fit of the unconstrained model compared to the constrained model (Supporting Information S1: Supplementary
Table 6). Supporting Information S1: Supplementary Table 7shows the parameter estimates of the measurement model
within the main structural equation model. Specifically, in terms of parameter estimates, the effect of youth adversity on
depression/anxiety varied across intersectionality profiles (Supporting Information S1: Supplementary Table 8). The
strongest effect of youth adversity was observed in the male, higher SES, low hyperactivity/inattention profile.
After adjusting for multiple testing, the only significant difference (p< .004) between the parameter estimates reflective of
the effect of youth adversity on depression/anxiety was observed between the male, higher SES, low hyperactivity/inattention
group (where the effect was stronger) and the female, higher SES, low hyperactivity/inattention group (Wald statistic = 4.632,
p< .001) (Supporting Information S1: Supplementary Table 9).
3.2 |Research question 2: Is the effect of youth adversity on depression/anxiety moderated by
gender, SES, and hyperactivity/inattention, and their intersections?
Table 2shows the compound parameter estimates for evaluating moderation effects attributable to the individual
characteristics of gender, SES, and hyperactivity/inattention, and their intersections. Moderating effects were detected for
TABLE 1 Descriptive characteristics.
Variable N(% [95% CI])
Gender
Female 5934 (50.74% [49.84, 51.64])
Male 6112 (49.26% [48.36, 50.16])
SES
Lower 1507 (12.49% [11.90, 13.09])
Higher 10559 (87.51% [86.91, 88.09])
Hyperactivity/inattention
High 3229 (27.32% [26.52, 28.13])
Low 8590 (72.68% [71.87, 73.48])
Youth adversity
Yes 4267 (36.37% [35.50, 37.25])
No 7465 (63.63% [62.75, 64.50])
Measured depression/anxiety symptoms
a
Mean (SD) 4.15 (2.69)
Range (of 0–10) 0–10
Median (IQR) 4(2–6)
Note: Percentage of individuals with data available for each variable.
Abbreviations: IQR, interquartile range; SES, socioeconomic status.
a
N= 11,841.
JOURNAL OF ADOLESCENCE
|
1309
FIGURE 2 The effect of youth adversity on depression/anxiety by intersectionality profiles. SES, socioeconomic status. Latent depression/anxiety scaled
relative to the intercept for the male, higher SES, low hyperactivity/inattention intersectionality profile (mean‐centered at zero). Parameter estimates shown
in Supporting Information S1: Supplementary Table 8.
TABLE 2 Compound parameter estimates of the effect of youth adversity on depression/anxiety by intersectionality profiles.
Parameter description
Effect of youth adversity on depression/anxiety
Unstandardized estimate (SE) 95% CI Standardized estimate (SE) 95% CI
Weighted averages
Males 0.982 (0.039) 0.901, 1.053 0.938 (0.036) 0.865, 1.002
Females 0.754 (0.041) 0.672, 0.835 0.705 (0.035) 0.639, 0.778
Higher SES 0.903 (0.032) 0.834, 0.970 0.861 (0.027) 0.809, 0.914
Lower SES 0.626 (0.075) 0.488, 0.756 0.558 (0.066) 0.433, 0.676
Low hyperactivity/inattention 0.867 (0.033) 0.799, 0.936 0.851 (0.030) 0.794, 0.912
High hyperactivity/inattention 0.874 (0.056) 0.756, 0.983 0.756 (0.045) 0.659, 0.841
Weighted main effects
Gender (male ‐female) 0.228 (0.052) 0.123, 0.323 0.215 (0.049) 0.114, 0.36
SES (higher ‐lower) 0.277 (0.079) 0.144, 0.420 0.262 (0.075) 0.136, 0.397
Hyperactivity/inattention (low ‐high) −0.007 (0.060) −0.123, 0.119 −0.006 (0.057) −0.115, 0.120
Weighted interaction effects
Gender × SES 0.064 (0.158) −0.234, 0.370 0.060 (0.150) −0.221, 0.356
Gender × hyperactivity/inattention 0.221 (0.120) −0.023, 0.470 0.209 (0.113) −0.019, 0.442
SES × hyperactivity/inattention 0.110 (0.177) −0.230, 0.469 0.104 (0.167) −0.219, 0.449
Gender × SES × hyperactivity/inattention 0.117 (0.353) −0.649, 0.738 0.110 (0.333) −0.619, 0.618
Note:N= 11,707 with intersectionality profile data and youth adversity data. Weighted pooled standard deviations used for calculation of standardized estimates. Compound
parameter estimates specified using the weighted least squares estimates derived from the main multiple group model of latent depression/anxiety regressed on youth adversity,
weighted by intersectionality profile sample size. SE and bias‐corrected bootstrapped CI from 1000 draws. The effect of youth adversity on depression/anxiety reflects the average
unit change in latent depression/anxiety with exposure to youth adversity, compared to an absence of youth adversity. Results were substantively unchanged where missing item‐
level data and youth adversity data was imputed (10 datasets, N= 11,797). Standardized estimates for main and interaction effects with non‐zero CI shown in bold typeset
Abbreviation: SES, socioeconomic status.
1310
|
HAVERS ET AL.
gender, and for SES, with a stronger effect of youth adversity on depression/anxiety for males compared to females (β= 0.215,
95% CI = 0.114, 0.360), and for individuals with higher compared to lower SES (β= 0.262, 95% CI = 0.136, 0.397).
There was also evidence suggesting moderation attributable to the intersectional (interaction) effect between gender and
hyperactivity/inattention. Although the confidence intervals for this effect included zero, it warrants discussion as its effect
size (β= 0.209) was comparable to the effects discussed above (i.e., for gender, and for SES). Specifically, the difference
between the estimates for low hyperactivity/inattention females (b= 0.723, SE = 0.048) and males (b= 1.010, SE = 0.043), was
0.209 SD greater (95% CI = −0.019, 0.442) compared to the difference between high hyperactivity/inattention females
(b= 0.840, SE = 0.078) and males (b= 0.906, SE = 0.078). This suggests a more pronounced moderating effect of gender
(where the effect is stronger for males compared to females) at a low level of hyperactivity/inattention compared to a high
level of hyperactivity/inattention.
4|DISCUSSION
This study stands among the first to explore the intersectional role of gender, SES, and hyperactivity/inattention, in
moderating the effect of youth adversity on depression/anxiety during adolescence. It adds to existing findings showing that
youth adversity adversely affects adolescent mental health—further providing evidence to suggest that these negative effects
pervade across the different societal intersections under consideration in the current study. The findings from this study
suggest a stronger effect of youth adversity and depression/anxiety for males than for females, and for individuals from
higher than from lower SES backgrounds. Additionally, there was some indication of moderation reflecting the intersection
between gender and hyperactivity/inattention. Though the confidence intervals for this estimate included zero, the effect size
was noteworthy, prompting discussion and further investigation in other samples to evaluate the robustness of this finding.
Contrary to our hypothesis, the most substantial effect of youth adversity on depression/anxiety was seen in males from
higher SES backgrounds with low hyperactivity/inattention. While this effect was significantly greater only compared to the
effect in the female, higher SES, low hyperactivity/inattention group, it aligns broadly with the moderating effects of gender
and SES that were observed (i.e., stronger effects for males, and for higher SES). The direction of these effects contrast with
findings from other studies that have reported greater effects of youth adversity in females (Assini‐Meytin et al., 2022), and in
individuals from lower SES backgrounds (Walsh et al., 2019). Comparative studies are required to ascertain the extent to
which our differing results may reflect, for example, the inclusion of bullying‐victimization as contributing to youth adversity
rather than the sole inclusion of within‐household ACEs (see Introduction), and the current focus on narrow symptoms of
depression/anxiety specifically, rather than considering a broader range of mental health problems.
The current results may also be interpreted from the perspective of stress inoculation theory (Compton & Pfau, 2005).
From this theoretical perspective, the findings may be understood to reflect that females and individuals from lower SES
backgrounds may exhibit more resilience to depression and anxiety in the face of youth adversity due to resistance, or
enhanced inoculation, to stress. For example, it could be speculated that females who experience youth adversity are more
robust to depression and anxiety, compared to males, because of their development under inherent structural conditions of
oppression (similarly, individuals from lower SES compared to higher SES backgrounds) (Banyard & Graham‐
Bermann, 1993). Further regarding SES, there may also be specific pressures affecting higher SES individuals that have
detrimental effects on their mental health and wellbeing (Luthar et al., 2018; Luthar et al., 2020). How this may manifest
further in the context of youth adversity and intersectionality stands as an interesting avenue for future research. However, as
outlined in the introduction of the current paper, importantly –the intersection of an individual's characteristics may define
their social/societal position within a system of structural hierarchies, beyond the effects of individual characteristics in
isolation (Crenshaw, 1991).
Before discussing the suggestive evidence for moderation attributable to intersectionality, it is important to highlight the
absence of moderation attributable to hyperactivity/inattention (in isolation). This finding could imply that this characteristic
does not affect the relation between youth adversity and depression/anxiety in mid‐adolescence. While previous findings
indicate higher depression and anxiety symptoms in youths with elevated levels of hyperactivity/inattention (Meinzer
et al., 2014), whether this characteristic, or trait, has a specific role in moderating the effect of youth adversity on mental
health has not been extensively investigated. Further research is required to unpack the role of hyperactivity/inattention, and
neurodivergence more broadly, in the context of youth adversity and mental health. However, even though a moderating
effect of hyperactivity/inattention in isolation was not detected in the current study, its intersectional relations with other
characteristics such as gender may be of potential importance, as is discussed below.
With regard to the intersectional moderation effect of gender and hyperactivity/inattention, the lack of statistical
confidence reflected the interval for the estimate prompts caution in interpretation. Nonetheless, because the estimated effect
size was comparable to those estimated for gender and for SES, it is discussed in terms of providing preliminary evidence to
suggest that the effect of youth adversity on adolescent depression/anxiety may partly depend on the intersection between
gender and hyperactivity/inattention. Specifically, this result indicates a stronger differential effect of gender (i.e., a greater
JOURNAL OF ADOLESCENCE
|
1311
effect for males than for females) in individuals with low compared to high levels of hyperactivity/inattention. While
replication of this effect in other samples and using other measures is necessary, through a lens of social and societal
oppression and privilege –the results may tentatively suggest that youth adversity has a particularly pronounced impact on
depression/anxiety in neurotypical males, due to their relatively privileged position in society (Banyard & Graham‐
Bermann, 1993). Notwithstanding, an accumulation of findings from across further studies is needed to draw conclusions
that can meaningfully contribute towards theoretical and clinical advancements.
To this point, it should be considered that the current findings may at least in part be reflective of specific factors such as
the binary operationalization of youth adversity, the conceptualization of depression/anxiety as a single construct, and the
operationalization of the individual characteristics contributing to the intersectional profiles (e.g., using free school meals
eligibility as an index of SES, using a cut‐offscore for hyperactivity/inattention symptoms as an index of neurodivergence).
Importantly, the results of our measurement invariance analyses indicate that the current findings are unlikely to be solely
due to noninvariance in the measurement of depression/anxiety across intersectional profiles. This contributes to recent
intersectionality research findings demonstrating invariance in the measurement of depression across intersectional groups
in adulthood (Cintron et al., 2023). It is also crucial to consider that other demographic and individual‐level characteristics
such as ethnicity (Mersky et al., 2021), immigration status (Kern et al., 2020), gender diverse identification and sexual
orientation (Jonas et al., 2022)–all of which were not measured in the current sample, are likely to be important components
of intersectionality (Ghavami et al., 2016). As intersectionality theory was developed specifically in terms of the intersectional
position of women of color (Crenshaw, 1991), this absence is particularly poignant for ethnicity. The inclusion of ethnicity as
an individual component of intersectional identity should be prioritized in future studies addressing research questions in
this emerging line of inquiry and is planned in future studies within our research group (Hosang et al., 2023).
Given the use of compound parameters in our study to approximate the moderating effects of intersectionality, these estimates
cannot be used for sample size planning in future studies in the same way for example that maximum likelihood estimates could be
used (Hancock & French, 2013). Nonetheless, pseudo power analyses indicated that doubling the current sample size would lead to
non‐zero overlapping confidence intervals for the intersectional moderating effect of gender by hyperactivity/inattention (b= 0.221,
95% CI = 0.051, 0.392). This gives a very crude indication of the extenttowhichfuturesamplesizesmayneedtobeincreasedto
detect moderation effects for intersectionality as conceptualized in the current study.
5|STRENGTHS AND LIMITATIONS
The current study has multiple strengths, such as the use of both self‐report and objective (local government) measures of
youth adversity, and data from a large sample of adolescents. However, interpretations of our results should consider several
limitations. Only self‐reported data regarding depression/anxiety was available in the current study (indexed by the
emotional problems subscale of the SDQ). It is noted that the SDQ was developed to incorporate multi‐informant reports
(self, parent, teacher). It is therefore highlighted that the current findings pertain specifically to individuals' perceptions of
their own emotional problems. Future studies incorporating multi‐informant reports would provide an interesting basis for
investigating how self, parent, and teacher interpretations may differ by intersectionality groups. Also related to the specific
use of the SDQ emotional problems subscale in our analyes, further incorporating other subscales of the SDQ would facilitate
a more comprehensive evaluation regarding, for example, the extent to which the observed gender differences in the effect of
youth adversity may also manifest differentially across other dimensions of psychopathology (e.g., prosocial behavior,
conduct problems).
Details regarding the specific types of adversities that were experienced were not available. Recent research suggests that
patterns of specific types of adversity (e.g., reflected in latent classes) may be important for investigating associations with
characteristics such as gender and SES (Lacey et al., 2022). Additionally, as youth adversity was recorded concurrently to
mental health, the temporal impact of youth adversity cannot be inferred. Future work planned in our group will incorporate
prospective records of adversity in investigating mechanisms linking youth adversity and later metal health problems across
intersectional identities, in other geographic regions of the UK (Hosang et al., 2023).
6|CONCLUSIONS
In summary, our findings indicate that youth adversity has detrimental effects on the mental health of 13–14‐year‐olds living
in Cornwall in the United Kingdom, and that these detrimental effects transcend the intersectional identities considered in
this study. Specifically, however, male gender and a higher SES background are linked to higher average levels of depression/
anxiety in the face of youth adversity, compared to female gender, and lower SES, respectively. Importantly, because this may
be due to effects such as stress inoculation, caution should be taken in concluding that these (latter) demographic subgroups
do not require support.
1312
|
HAVERS ET AL.
While intersectionality is an increasingly recognized theoretical framework for situating inquiry into adolescent mental
health (Patil et al., 2018), investigating the links between youth adversity and adolescent mental health within a framework of
intersectionality is an emerging field. Our findings are among the first to contribute to this specific line of inquiry and
provide a preliminary platform for subsequent research. An accumulation of research in this area can contribute to better
understanding how structural conditions of oppression and privilege may act together to influence the differential effects of
youth adversity. Insights from this growing research area could lead to a more stratified approach in exploring mechanisms
of change, thereby informing and enhancing interventions aimed at mitigating the adverse effects of youth adversity.
ACKNOWLEDGMENTS
The authors are grateful to all the research teams and participants who have contributed to the data that was analyzed in this
paper, and to all individuals involved in the ATTUNE project young people advisory groups. LH is grateful to Dubravka
Svetina Valdivia and Gabriela Roman; and Gregory R. Hancock, for providing valuable guidance and support with the
statistical modeling used in this study. The ATTUNE project is funded by a Cross Council UK Research and Innovation
[UKRI] award (MR/W002183/1). Additionally, this research was funded by the National Institute for Health Research
(NIHR) Applied Research Collaboration Oxford and Thames Valley at Oxford Health NHS Foundation Trust. KB is part
supported by Oxford Health NIHR Biomedical Research Centre. The HeadStart Cornwall data was collected as part of the
HeadStart learning program and supported by funding from The National Lottery Community Fund (HeadStart). The
content is solely the responsibility of the authors. The views expressed are those of the authors and not necessarily those of
the NHS, NIHR, Department of Health, or The National Lottery Community Fund (HeadStart).
CONFLICT OF INTEREST STATEMENT
The authors have declared that no competing interests exist
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from Cornwall Council. Restrictions apply to the availability of
these data, which were used under licence for this study. Data are available with the permission of Cornwall Council (www.
cornwall.gov.uk). Mplus code for the main analysis model is provided in Supplementary Materials 3. Additional analytic code
can be requested from the corresponding author.
ETHICS STATEMENT
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national
and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All
procedures involving human subjects/patients were approved by the University College London Ethics Committee for the
HeadStart study (approval number: 8097/003).
ORCID
Laura Havers http://orcid.org/0000-0003-2529-4669
REFERENCES
Assini‐Meytin, L. C., Fix, R. L., Green, K. M., Nair, R., & Letourneau, E. J. (2022). Adverse childhood experiences, mental health, and risk behaviors in
adulthood: Exploring sex, racial, and ethnic group differences in a nationally representative Sample. Journal of child & adolescent trauma,15(3),
833–845.
Banyard, V. L., & Graham‐Bermann, S. A. (1993). Can women cope?: A gender analysis of theories of coping with stress. Psychology of Women Quarterly,
17(3), 303–318.
Bellis, M. A., Hughes, K., Ford, K., Ramos Rodriguez, G., Sethi, D., & Passmore, J. (2019). Life course health consequences and associated annual costsof
adverse childhood experiences across Europe and North America: A systematic review and meta‐analysis. The Lancet Public Health,4(10), e517–e528.
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal
Statistical Society Series B: Statistical Methodology,57(1), 289–300.
Blakemore, S. J., Burnett, S., & Dahl, R. E. (2010). The role of puberty in the developing adolescent brain. Human Brain Mapping,31(6), 926–933.
Bowleg, L. (2012). The problem with the phrase women and minorities: Intersectionality—an important theoretical framework for public health. American
Journal of Public Health,102(7), 1267–1273.
Campbell, O. L. K., Bann, D., & Patalay, P. (2021). The gender gap in adolescent mental health: A cross‐national investigation of 566,829 adolescents across
73 countries. SSM ‐Population Health,13, 100742.
Caspi, A., Houts, R. M., Ambler, A., Danese, A., Elliott, M. L., Hariri, A., Harrington, H., Hogan, S., Poulton, R., Ramrakha, S., Rasmussen, L. J. H.,
Reuben, A., Richmond‐Rakerd, L., Sugden, K., Wertz, J., Williams, B. S., & Moffitt, T. E. (2020). Longitudinal assessment of mental health disorders
and comorbidities across 4 decades among participants in the Dunedin Birth Cohort Study. JAMA Network Open,3(4), e203221.
Cintron, D. W., Matthay, E. C., & McCoach, D. B. (2023). Testing for intersectional measurement invariance with the alignment method: Evaluation of the
8‐item patient health questionnaire. Health Services Research,58(S2), 248–261.
Cole, E. R. (2009). Intersectionality and research in psychology. American Psychologist,64(3), 170–180.
JOURNAL OF ADOLESCENCE
|
1313
Compton, J. A., & Pfau, M. (2005). Inoculation theory of resistance to influence at maturity: Recent progress in theory development and application and
suggestions for future research. Annals of the International Communication Association,29(1), 97–146.
Craig, S. G., Bondi, B. C., O'Donnell, K. A., Pepler, D. J., & Weiss, M. D. (2020). ADHD and exposure to maltreatment in children and youth: A systematic
review of the past 10 years. Current Psychiatry Reports,22(12), 79.
Crenshaw, K. (1991). Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stanford Law Review,43(6), 1241–1300.
Deighton, J., Lereya, S. T., Casey, P., Patalay, P., Humphrey, N., & Wolpert, M. (2019). Prevalence of mental health problems in schools: Poverty and other
risk factors among 28 000 adolescents in England. British Journal of Psychiatry,215(3), 565–567.
Evans, C. R., Williams, D. R., Onnela, J. P., & Subramanian, S. V. (2018). A multilevel approach to modeling health inequalities at the intersection of
multiple social identities. Social Science & Medicine (1982),203,64–73.
Felitti, V. J., Anda, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., Koss, M. P., & Marks, J. S. (1998). Relationship of childhood abuse
and household dysfunction to many of the leading causes of death in adults. American Journal of Preventive Medicine,14(4), 245–258.
Ghavami, N., Katsiaficas, D., & Rogers, L. O. (2016). Toward an Intersectional approach in developmental science: The role of race, gender, sexual
orientation, and immigrant status. In S. S. Horn, M. D. Ruck & L. S. Liben, editors, Advances in child development and behavior (pp. 31–73). JAI.
Available from https://www.sciencedirect.com/science/article/pii/S006524071530001X
Goodman, R. (1997). The strengths and difficulties questionnaire: A research note. Journal of Child Psychology and Psychiatry,38(5), 581–586.
Griner, D., & Smith, T. B. (2006). Culturally adapted mental health intervention: A meta‐analytic review. Psychotherapy: Theory, Research, Practice, Training.
43, 531–548.
Hancock, G. R., & French, B. F. (2013). Power analysis in structural equation modeling, In: Structural equation modeling: A second course (2nd ed.,
pp. 117–159). IAP Information Age Publishing.
Hankivsky, O., Grace, D., Hunting, G., Giesbrecht, M., Fridkin, A., Rudrum, S., Ferlatte, O., & Clark, N. (2014). An intersectionality‐based policy analysis
framework: Critical reflections on a methodology for advancing equity. International Journal for Equity in Health,13(1), 119.
Havers, L., Shuai, R., Fonagy, P., Fazel, M., Morgan, C., Fancourt, D., McCrone, P., Smuk, M., Bhui, K., Shakoor, S., & Hosang, G. M. (2024). Youth adversity
and trajectories of depression/anxiety symptoms in adolescence in the context of intersectionality in the United Kingdom. Psychological Medicine,
1–11.
Hosang, G. M., Havers, L., Shuai, R., Fonagy, P., Fazel, M., Morgan, C., Karamanos, A., Fancourt, D., McCrone, P., Smuk, M., Bhui, K., & Shakoor, S. (2023).
Protocol for secondary data analysis of 4 UK cohorts examining youth adversity and mental health in the context of intersectionality. PLoS One,18(8),
e0289438.
Van de Schoot, R., & Miocević, M. (2020). Small sample size solutions: A guide for applied researchers and practitioners (p. 284). Taylor & Francis.
Hu, L., & Bentler, P. M. (1999). Cutoffcriteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural
Equation Modeling: A Multidisciplinary Journal,6(1), 1–55.
Jonas, L., Salazar de Pablo, G., Shum, M., Nosarti, C., Abbott, C., & Vaquerizo‐Serrano, J. (2022). A systematic review and meta‐analysis investigating the
impact of childhood adversities on the mental health of LGBT+ youth. JCPP Advances,2(2), e12079.
Kalmakis, K. A., & Chandler, G. E. (2014). Adverse childhood experiences: Towards a clear conceptual meaning. Journal of Advanced Nursing,70(7),
1489–1501.
Kern, M. R., Duinhof, E. L., Walsh, S. D., Cosma, A., Moreno‐Maldonado, C., Molcho, M., Currie, C., & Stevens, G. W. J. M. (2020). Intersectionality and
adolescent mental well‐being: A cross‐nationally comparative analysis of the interplay between immigration background, socioeconomic status and
gender. Journal of Adolescent Health,66(6, Suppl. ment), S12–S20.
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E. E. (2005). Lifetime prevalence and age‐of‐onset distributions of DSM‐IV
disorders in The National Comorbidity Survey Replication. Archives of General Psychiatry,62(6), 593–602.
Lacey, R. E., Howe, L. D., Kelly‐Irving, M., Bartley, M., & Kelly, Y. (2022). The clustering of adverse childhood experiences in the avon longitudinal study of
parents and children: Are gender and poverty important? Journal of Interpersonal Violence,37(5–6), 2218–2241.
Lai, M. C., Kassee, C., Besney, R., Bonato, S., Hull, L., Mandy, W., Szatmari, P., & Ameis, S. H. (2019). Prevalence of co‐occurring mental health diagnoses in
the autism population: A systematic review and meta‐analysis. The Lancet Psychiatry.6(10), 819–829.
Lerner, R. M., & Galambos, N. L. (1998). Adolescent development: Challenges and opportunities for research, programs, and policies. Annual Review of
Psychology,49(1), 413–446.
Luthar, S. S., & Kumar, N. L. (2018) Youth in high‐achieving schools: Challenges to mental health and directions for evidence‐based interventions. In:
A. W. Leschied, D. H. Saklofske & G. L. Flett, editors., Handbook of school‐based mental health promotion: An evidence‐informed framework for
implementation. (pp. 441–458). Springer International PublishingAvailable from https://doi.org/10.1007/978-3-319-89842-1_23
Luthar, S. S., Kumar, N. L., & Zillmer, N. (2020). High‐achieving schools connote risks for adolescents: Problems documented, processes implicated, and
directions for interventions. American Psychologist,75(7), 983–995.
Marsh, H. W., Hau, K. T., & Wen, Z. (2004). In search of golden rules: Comment on hypothesis‐testing approaches to setting cutoffvalues for fit indexes and
dangers in overgeneralizing Hu and Bentler's (1999) findings. Structural Equation Modeling: A Multidisciplinary Journal,11(3), 320–341.
McNeish, D., & Stapleton, L. M. (2016). Modeling clustered data with very few clusters. Multivariate Behavioral Research,51(4), 495–518.
Meinzer, M. C., Pettit, J. W., & Viswesvaran, C. (2014). The co‐occurrence of attention‐deficit/hyperactivity disorder and unipolar depression in children
and adolescents: A meta‐analytic review. Clinical Psychology Review,34(8), 595–607.
Merlo, J. (2018). Multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) within an intersectional framework. Social Science
& Medicine (1982),203,74–80.
Mersky, J. P., Choi, C., Plummer Lee, C., & Janczewski, C. E. (2021). Disparities in adverse childhood experiences by race/ethnicity, gender, and economic
status: Intersectional analysis of a nationally representative sample. Child Abuse & Neglect,117, 105066.
Pantazakos, T., & Vanaken, G. J. (2023). Addressing the autism mental health crisis: The potential of phenomenology in neurodiversity‐affirming clinical
practices. Frontiers in Psychology,14, 1225152. Available from https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.
1225152/full
Patil, P. A., Porche, M. V., Shippen, N. A., Dallenbach, N. T., & Fortuna, L. R. (2018). Which girls, which boys? The intersectional risk for depression by race
and ethnicity, and gender in the U.S. Clinical Psychology Review,66,51–68.
Scully, C., McLaughlin, J., & Fitzgerald, A. (2020). The relationship between adverse childhood experiences, family functioning, and mental health problems
among children and adolescents: A systematic review. Journal of Family Therapy,42(2), 291–316.
Slopen, N., Shonkoff, J. P., Albert, M. A., Yoshikawa, H., Jacobs, A., Stoltz, R., & Williams, D. R. (2016). Racial disparities in child adversity in the U.S.
American Journal of Preventive Medicine,50(1), 47–56.
1314
|
HAVERS ET AL.
Svetina, D., Rutkowski, L., & Rutkowski, D. (2020). Multiple‐group invariance with categorical outcomes using updated guidelines: An illustration using M
plus and the lavaan/semtools packages. Structural Equation Modeling: A Multidisciplinary Journal,27(1), 111–130.
Udry, J. R. (2000). Biological limits of gender construction. American Sociological Review,65(3), 443–457.
Walsh, D., McCartney, G., Smith, M., & Armour, G. (2019). Relationship between childhood socioeconomic position and adverse childhood experiences
(ACEs): A systematic review. Journal of Epidemiology and Community Health,73(12), 1087–1093.
Williams, D. R., Neighbors, H. W., & Jackson, J. S. (2003). Racial/ethnic discrimination and health: Findings from community studies. American Journal of
Public Health,93(2), 200–208.
Wu, H., & Estabrook, R. (2016). Identification of confirmatory factor analysis models of different levels of invariance for ordered categorical outcomes.
Psychometrika,81(4), 1014–1045.
SUPPORTING INFORMATION
Additional supporting information can be found online in the Supporting Information section at the end of this article.
How to cite this article: Havers, L., Bhui, K., Shuai, R., Fonagy, P., Fazel, M., Morgan, C., Fancourt, D., McCrone, P.,
Smuk, M., Hosang, G. M., & Shakoor, S. (2024). A cross‐sectional investigation into the role of intersectionality as a
moderator of the relation between youth adversity and adolescent depression/anxiety symptoms in the community.
Journal of Adolescence,96, 1304–1315. https://doi.org/10.1002/jad.12347
JOURNAL OF ADOLESCENCE
|
1315
