PreprintPDF Available

Youth adversity and trajectories of depression and anxiety symptoms in adolescence in the context of intersectionality

Authors:
Preprints and early-stage research may not have been peer reviewed yet.

Abstract

Background: Youth adversity is associated with persistence of depression and anxiety symptoms over time. Evidence suggests that this association may be greater for disadvantaged societal groups (such as females) compared with advantaged groups (e.g., males). However, given that persistent symptoms are observed across a range of disadvantaged groups (e.g., low compared with high socio-economic status [SES]), the intersection of individual characteristics may be an important moderator of inequality. Methods: Data from HeadStart Cornwall (N=5,336) was used to assess the effect of youth adversity on symptoms of depression and anxiety, measured using the Strengths and Difficulties Questionnaire emotional problems subscale, at three time-points in 11-14-year-olds. Latent trajectories and regression coefficients were estimated for eight intersectionality profiles (based on gender, SES, and hyperactivity/inattention) within a multiple group structural equation model. Compound parameters were specified to estimate the moderating effects of the individual characteristics and their intersections. Results: Youth adversity, compared with an absence of such, was associated with higher average depression and anxiety symptoms at baseline (11-12-years), across all intersectionality profiles. The magnitude of the effect of youth adversity differed across profiles, and there was weak evidence to suggest that the effect of youth adversity on the average rate of change in depression and anxiety symptoms was moderated by the intersection of, i) gender and SES, and ii) gender, SES, and hyperactivity/inattention. Conclusions: Youth adversity has detrimental effects on the development of depression and anxiety symptoms that pervade across intersectionality profiles: The extent to which these effects are moderated by intersectionality are discussed in terms of operational factors and sample size. The current results provide a platform for further research, which is needed to determine whether intersectionality is important in moderating the effect of youth adversity on the development of depression and anxiety symptoms in adolescence.
For Peer Review
Youth adversity and trajectories of depression and anxiety
symptoms in adolescence in the context of intersectionality
Journal:
Journal of Child Psychology and Psychiatry
Manuscript ID
Draft
Manuscript Type:
Original Article
Date Submitted by the
Author:
n/a
Complete List of Authors:
Havers, Laura; Birkbeck University of London, Department of
Psychological Sciences
Shuai, Ruichong; Queen Mary University of London Faculty of Medicine
and Dentistry, Centre for Psychiatry and Mental Health
Fonagy, Peter; University College London, Division of Psychology and
Language Sciences
Fazel, Mina; University of Oxford, Psychiatry; Oxford University Hospitals
NHS Foundation Trust, Centre for Psychological Medicine
Morgan, Craig; King's College London, Health Service & Population
Research
Fancourt, Daisy; University College London, Behavioural Science and
Health
Smuk, Melanie; Barts and the London School of Medicine and Dentistry,
Centre for Psychiatry
Bhui, Kamaldeep ; University of Oxford, Department of Psychiatry
Shakoor, Sania; Queen Mary University of London, Centre for Psychiatry,
Wolfson Institute of Preventive Medicine
Hosang, G; Wolfson Institute of Preventive Medicine, Centre for
Psychiatry
Key Words:
Adolescence, Adversity, Anxiety, Depression, Longitudinal studies
JCPP
For Peer Review
1
Journal submission: The Journal of Child Psychology and Psychiatry
Word count including abstract: 4,908 (max 5000)
Word count abstract: 294 (max 300)
Tables: 3 (max 5)
Figures: 2 (max 5)
Style: APA 5th
Title: Youth adversity and trajectories of depression and anxiety symptoms in adolescence in
the context of intersectionality
Running head: Intersectionality, youth adversity, and trajectories of adolescent depression
and anxiety symptoms
Authors: Laura Havers1*, Ruichong Shuai1, Peter Fonagy2,3, Mina Fazel4, Craig Morgan5,6,
Daisy Fancourt7, Paul McCrone8, Melanie Smuk9, Kamaldeep Bhui10,11,12, Sania Shakoor1+ and
Georgina M. Hosang1+
Affiliations:
1 Centre for Psychiatry and Mental Health, Wolfson Institute of Population Health, Queen
Mary, University of London, London, UK
2 Anna Freud National Centre for Children and Families, London, UK
Page 1 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
2
3 Research Department of Clinical, Educational and Health Psychology, University College
London, London, UK
4 Department of Psychiatry, University of Oxford, Oxford, UK
5 Health Service and Population Research, Institute of Psychology, Psychiatry &
Neuroscience, King's College London, London, UK
6 ESRC Centre for Society and Mental Health, King's College London, London, UK
7 Department of Behavioural Science and Health, University College London, London, UK
8 Institute for Lifecourse Development, University of Greenwich, London, UK
9 Centre for Genomics and Child Health, Blizard Institute, Queen Mary, University of London,
London, UK
10 Department of Psychiatry, Nuffield Department of Primary Care Health Sciences, and
Wadham College, University of Oxford, Oxford, UK
11 Oxford Health and East London NHS Foundation Trusts, UK
12 World Psychiatric Association Collaborating Centre, Oxford, UK
*Corresponding author
+Joint last authors
Page 2 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
3
Abstract
Background: Youth adversity is associated with persistence of depression and anxiety
symptoms over time. Evidence suggests that this association may be greater for
disadvantaged societal groups (such as females) compared with advantaged groups (e.g.,
males). However, given that persistent symptoms are observed across a range of
disadvantaged groups (e.g., low compared with high socio-economic status [SES]), the
intersection of individual characteristics may be an important moderator of inequality.
Methods: Data from HeadStart Cornwall (N=5,336) was used to assess the effect of youth
adversity on symptoms of depression and anxiety, measured using the Strengths and
Difficulties Questionnaire emotional problems subscale, at three time-points in 11-14-year-
olds. Latent trajectories and regression coefficients were estimated for eight
intersectionality profiles (based on gender, SES, and hyperactivity/inattention) within a
multiple group structural equation model. Compound parameters were specified to
estimate the moderating effects of the individual characteristics and their intersections.
Results: Youth adversity, compared with an absence of such, was associated with higher
average depression and anxiety symptoms at baseline (11-12-years), across all
intersectionality profiles. The magnitude of the effect of youth adversity differed across
profiles, and there was weak evidence to suggest that the effect of youth adversity on the
Page 3 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
4
average rate of change in depression and anxiety symptoms was moderated by the
intersection of, i) gender and SES, and ii) gender, SES, and hyperactivity/inattention.
Conclusions: Youth adversity has detrimental effects on the development of depression and
anxiety symptoms that pervade across intersectionality profiles: The extent to which these
effects are moderated by intersectionality are discussed in terms of operational factors and
sample size. The current results provide a platform for further research, which is needed to
determine whether intersectionality is important in moderating the effect of youth
adversity on the development of depression and anxiety symptoms in adolescence.
Keywords: Adversity, depression, anxiety, longitudinal studies, adolescence
Page 4 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
5
Introduction
Symptoms of depression (e.g., low mood, loss of enjoyment) and anxiety (e.g.,
nervousness, worry) are among the most common mental health problems in adolescence
(Michaud & Fombonne, 2005). A global prevalence rate of ~35% has been estimated for
elevated depressive symptoms (Shorey et al., 2022) and ~10% for elevated anxiety symptoms
(Biswas et al., 2020). In clinical populations, prevalence rates of depressive and anxiety
disorders are estimated at ~3.5% and ~4.5%, respectively (World Health Organization, 2017),
and these disorders are associated with a range of maladies, including cardiovascular disease
(Tully et al., 2016) and poor quality of life (Hohls et al., 2021).
It has been estimated that one third to half of lifelong mental health disorders are
evident by the age of 15-years (Caspi et al., 2020; Kessler et al., 2005), highlighting the
importance of early to mid-adolescence when considering the emergence of mental health
problems. To understand the pathways from symptoms to disorder, it is critical to consider
the development of these symptoms over time, and the risk factors associated with this
development.
Experiencing persistently high or increasing symptoms of depression and/or anxiety
across adolescence is associated with negative outcomes, such as substance use, and school
dropout (Morin et al., 2011; Schubert et al., 2017). Importantly, research shows that some
groups in society are more likely to experience a trajectory of elevated (i.e., high, or
increasing) depression and/or anxiety symptoms. These include individuals with
neurodivergent conditions (e.g., attention deficit hyperactivity disorder [ADHD]) compared
with neurotypical individuals, those from low compared with high socio-economic status
[SES] backgrounds, and females compared with males (Leban, 2021; Schubert et al., 2017).
These findings suggest that minoritized, disadvantaged, and neurodivergent groups, are at
Page 5 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
6
heightened risk of experiencing trajectories of more elevated depression and anxiety
symptoms compared to their counterparts.
Youth adversity and adolescent depression/anxiety symptoms
In addition to the risk conferred by individual characteristics, research consistently
shows an association between youth adversity and trajectories of elevated depression
and/or anxiety symptoms (Bevilacqua et al., 2021; Desch et al., 2023; Leban, 2021). Youth
adversity encompasses stressful and potentially traumatic experiences that occur inside the
home (e.g., abuse and parental separation), often referred to as adverse childhood
experiences [ACEs] (Felitti et al., 1998; Kalmakis & Chandler, 2014), as well as outside the
home (e.g., bullying victimisation). Youth adversity is also found to be more prevalent in
disadvantaged, minoritized, and neurodivergent groups, which may contribute to why
trajectories of elevated depression and anxiety symptoms are common in these groups
(Assini-Meytin et al., 2022; Craig et al., 2020; Walsh et al., 2019).
Of the studies that have investigated the association between youth adversity and
mental health problems in young people, few have evaluated whether this association is
moderated by (dependent on) individual characteristics, such as gender. One such study
found a moderating effect of gender, observing a stronger effect of youth adversity on
depression symptoms for females compared with males, but there were no moderating
effects of race and ethnicity (Assini-Meytin et al., 2022). Another study did not detect
moderation by gender in the association between youth adversity and trajectories of
depression and anxiety symptoms (Leban, 2021). However, an individual can hold multiple
forms of minority or disadvantaged statuses (e.g., being a neurodivergent female from a low
SES background), which may lead to greater vulnerability to mental health problems in the
Page 6 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
7
face of adversity (Ghavami et al., 2016), although this has yet to be systematically
investigated.
Intersectionality and the youth adversity-depression/anxiety association
The broad analytic framework of intersectionality posits that the intersection,
reflecting an interaction, of an individual’s characteristics (e.g., gender, ethnicity, SES) has
importance beyond the additive effects of these characteristics (Bowleg, 2012; Crenshaw,
1990). Intersectional identities are considered within a complex system of societal and social
hierarchies, in the context of power, oppression, and privilege (Crenshaw, 1990).
Investigating the extent to which the effects of youth adversity on the development of
depression and anxiety symptoms may depend on intersections of individual characteristics
reflective of exclusion and marginalisation may offer a more comprehensive and better
account of how adversity leads to poor mental health.
Current study
To our knowledge, this is the first study to investigate the association between youth
adversity and trajectories of depression and anxiety symptoms in adolescence, in the context
of intersectionality (indexed by gender, SES, and hyperactivity/inattention). This study will
address two main research questions:
1) Is youth adversity associated with baseline and change over time in depression
and anxiety symptoms across intersectionality profiles? It was hypothesised that
across intersectionality profiles, youth adversity would be associated with higher
Page 7 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
8
baseline depression and anxiety symptoms that would remain higher over time,
compared to an absence of youth adversity.
2i) Does the association between youth adversity and depression and anxiety
symptoms (at baseline and their change over time) differ across intersectionality
profiles; and if so, ii) to what extent is the association moderated by gender, SES, and
hyperactivity/inattention, and their intersections? It was hypothesised i) that the
association between youth adversity and depression and anxiety symptoms (at
baseline and their change over time) would differ across intersectionality profiles; and
ii) that these associations would be moderated by gender, SES, and
hyperactivity/inattention, and their intersections. No directional predictions were
made regarding the moderating effects.
Page 8 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
9
Methods
Participants
A total of 5,336 individuals aged between 11 and 14 years were included in the current
investigation (Supplementary Table 1) and were drawn from the HeadStart Cornwall study
(Deighton et al., 2019; Hosang et al., 2023). In this study, pupils from all 31 state-maintained
secondary schools in Cornwall in the United Kingdom (UK) were invited to take part in 2017
when they were in school Year 7 (age 11-12-years, N = 4,575), and were followed up annually
in Year 8 (age 12-13-years, N = 4,600), and Year 9 (13-14-years, N = 3,604) (see Deighton et
al., 2019 for a detailed description of HeadStart). School identification data and individual-
age data was not available for this sample (reported ages reflect UK average ranges). Parental
consent was assumed unless parents opted their child out, and pupils assented prior to online
participation at their school. Ethical approval for HeadStart was obtained from the University
College London Ethics Committee (reference: 8097/003).
Measures
Depression and anxiety symptoms were measured at each school year using the five-
item emotional problems subscale of the Strengths and Difficulties Questionnaire (SDQ)
(Goodman, 1997). Items were self-rated on a 3-point Likert scale (“Not true”, “Somewhat
true”, “Certainly true”). A prorated total score (0-10) was calculated where at least three
items had response data, which was true for all observations at each school year.
Youth adversity was measured in two ways. First, using the bullying-victimisation item
from the SDQ peer problems subscale: “Other children or young people pick on me or bully
me” (Goodman, 1997). Second, using data collected from the local government Supporting
Families programme, where individuals were coded “On Family List” if they experienced any
Page 9 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
10
of the adversities listed in Supplementary Materials 1 (including risk of sexual exploitation,
homelessness, exposure to domestic violence). Data were combined to create a binary
variable indicating presence/absence of bullying (in any school year) and/or “On Family List”.
Data regarding the number and type of adversity was not available for this sample.
Gender and SES data were drawn from School Census records. Gender was recorded
as “female” or “male”, and receipt/non-receipt of free school meals was used to index “lower
SES” and “higher SES”, respectively. Hyperactivity/inattention was measured using the five
items of the SDQ hyperactivity/inattention subscale (Goodman, 1997). Individuals were
classified as “low” with scores of 0-6, and “high” with scores above 6, in line with scoring
recommendations (Goodman, 1997). Intersectionality profiles were created based on the
combination of gender, SES, and hyperactivity/inattention. Individuals were assigned to one
of eight intersectionality profiles (e.g., male, higher SES, low hyperactivity/inattention
profile).
Statistical analyses
To address research question one, trajectories of depression and anxiety symptoms
were estimated by specifying a latent growth model within a structural equation modelling
framework (Supplementary Materials 2). Depression and anxiety symptoms at each school
year were modelled as observed variables, specified as indicators of a latent intercept factor
and a latent slope factor. The latent intercept was positioned at school Year 7 (age 11-12-
years), reflecting estimated initial/baseline scores. The latent slope reflects the annual rate
of change in symptoms across school Years 7-9 (herein referred to as change over time).
School year-specific (residual) variances were freely estimated.
Page 10 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
11
An unconditional model was first run, followed by a conditional model with the latent
growth factors regressed on youth adversity. A multiple group conditional model was then
run, where parameters were freely estimated for each intersectionality profile
(Supplementary Materials 3: Mplus script; Supplementary Figure 1: path diagram). A
schematic diagram of the final model is shown in Figure 1. Incremental better fit of these
models would figuratively indicate that trajectories of depression and anxiety symptoms vary
with the presence/absence of youth adversity, and that at an omnibus level, the relations
between youth adversity and the latent growth factors vary across intersectionality profiles.
Model fit was assessed using the comparative fit index (CFI), root mean square error
of approximation (RMSEA), and standardised root mean square residual (SRMR). CFI >.95,
RMSEA <.08, and SRMR <.06 were broadly considered indicative of acceptable fit (Hu &
Bentler, 1999; Marsh et al., 2004). Bayesian information criteria (BIC) was also used, with
lower values indicative of better relative fit. Depression and anxiety symptoms were treated
as continuous data. Full information maximum likelihood estimation was used to
accommodate missing data across school years under the assumption that data was missing
at random. Robust estimation was used to accommodate multivariate nonnormality of
residuals, with adjustment to SE and test statistics. As a sensitivity analysis, the final analysis
model was rerun with multiply imputed youth adversity data from 10 datasets.
To address research question two, a parameter moderation approach was used (e.g.,
Bauer, 2017). The maximum likelihood regression estimates from the final analysis model
were specified to approximate the extent to which the effect of youth adversity on the latent
growth factors is moderated by the individual characteristics (gender, SES,
hyperactivity/inattention), and interactions between these characteristics (reflecting
intersectional effects). Using this approach, the derived compound parameters reflect the
Page 11 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
12
extent to which the regression coefficients from the final analysis model are moderated by
the characteristics and their intersections (Figure 1). Standard errors and confidence intervals
of the compound parameters were estimated from 1,000 bootstrapped draws.
Prior to the main analyses (detailed above), the five depression and anxiety items
were specified as indicators of a common latent factor in a confirmatory factor analysis, and
measurement properties of the model were assessed at each school year. This provides
model-based information regarding the extent to which modelling depression and anxiety as
a unitary construct provides an adequate representation of the sample data, although testing
models with more than one factor was beyond the scope of the current study. Longitudinal
measurement invariance of depression and anxiety across school years was also assessed.
This provides model-based information regarding the extent to which the measurement of
depression and anxiety can be considered equivalent across time (van de Schoot et al., 2012)
(Supplementary Materials 4). Model specifications recommended by Liu et al. (2017) for
ordered categorical response data were used. For the models described in this paragraph,
diagonally weighted least squares estimation was used, using pairwise present data.
Measurement invariance in depression and anxiety across intersectionality profiles at
age 13-14-years was observed in this sample (pooled with two other Year 9 cohorts from the
HeadStart Cornwall study) (Havers et al., in preparation).
Descriptive statistics were calculated using R (version 4.2). Structural equation
modelling was conducted using Mplus (version 8.8).
< Figure 1 about here >
Results
Page 12 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
13
A description of the sample is presented in Table 1. Mean depression and anxiety
symptoms were 3.82 (SD 2.54) at age 11-12-years, 3.89 (SD 2.64) at age 12-13-years, and 4.32
(SD 2.75) at age 13-14-years. A total of 52.94% of individuals experienced youth adversity
across the study period. Descriptive results for youth adversity and depression and anxiety
symptoms for each intersectionality profile are reported in Table 2. At each age, proportions
experiencing youth adversity and mean levels of depression and anxiety symptoms were
highest for the intersectionality profile of female, lower SES, and high
hyperactivity/inattention. Depression and anxiety symptoms were lowest for the
intersectionality profile of male, higher SES, and low hyperactivity/inattention. The rate of
youth adversity was lowest for the intersectionality profile of female, higher SES, and low
hyperactivity/inattention.
< Table 1 and Table 2 about here >
The depression and anxiety items were adequately represented by a common factor
model at each school year. Model fit statistics and reliability estimates (ω = 0.80-0.85) derived
from the models are reported in Supplementary Table 2. Longitudinal measurement
invariance at the scalar level (a model with constrained thresholds and loadings) was
supported (Supplementary Table 3).
Is youth adversity associated with baseline and change over time in depression and anxiety
symptoms across intersectionality profiles?
Page 13 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
14
The fit of the conditional multiple group model was acceptable (Supplementary Table
4). Figure 2 shows the estimated average latent trajectories for each intersectionality profile
in the presence of youth adversity, compared to an absence of youth adversity.
For all intersectionality profiles, the presence of youth adversity compared to an
absence of such, was associated with higher mean depression and anxiety symptoms at
baseline (age 11-12-years). Youth adversity was only marginally associated with a different
rate of change over time in these symptoms, except for in two profiles, where there was a
more moderate effect (Supplementary Table 5): In these profiles, there was weak evidence
to suggest that youth adversity, compared to an absence of youth adversity, was associated
with a decrease in change over time for the male, lower SES, high hyperactivity/inattention
profile, and with an increase in change over time for the female, lower SES, high
hyperactivity/inattention profile.
< Figure 2 about here >
Does the association between youth adversity and depression and anxiety symptoms (at
baseline and their change over time) differ across intersectionality profiles?
The conditional multiple group model showed an improvement in terms of BIC
compared to the conditional single group models (Supplementary Table 4). This provides
model-based information to suggest that the association between youth adversity and the
latent growth factors differ across intersectionality profiles at an omnibus level.
Parameter estimates of the conditional multiple group model are reported in
Supplementary Tables 5 and 6. The effect of youth adversity on baseline symptoms was
highest for the male, lower SES, high hyperactivity/inattention intersectionality profile (B =
Page 14 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
15
2.318, SE = 0.521), and was lowest for the female, low SES, low hyperactivity/inattention
intersectionality profile (B = 1.106, SE = 0.329). As noted for research question one,
although the evidence is notably weak, the results suggest that the effects of youth
adversity on the rate of change over time may differ across intersectionality profiles; being
associated with an increase in the rate of change over time for most, but not all, profiles.
To what extent is the association between youth adversity and depression and anxiety
symptoms (at baseline and their change over time) moderated by gender, SES, and
hyperactivity/inattention, and their intersections?
Compound parameter estimates for assessing the moderating effects of the individual
characteristics and their intersections are shown in Table 3. There was weak evidence for
moderation effects in the association between youth adversity and change over time in
depression and anxiety symptoms (discussed here for effect sizes greater than half a standard
deviation): The first was for a gender by SES interaction. The difference between the
estimates for lower SES females (B = 0.169, SE = 0.196) and males (B = -0.307, SE = 0.200),
was 0.543 SD greater (SE = 0.311, 95% CI -0.042, 1.153) compared to the difference between
the estimates for higher SES females (B = -0.008, SE = 0.069) and males (B = 0.209, SE = 0.067).
This reflects a greater moderating effect of gender at a lower level of SES than at a higher
level of SES.
The second was for a gender by SES by hyperactivity/inattention interaction. The
difference between the estimates for lower SES females (B = 0.507, SE = 0.437) and males (B
= -0.626, SE = 0.372) compared to higher SES females (B = -0.163, SE = 0.171) and males (B =
0.028, SE = 0.028), was 1.254 SD greater (SE = 0.737, 95% CI -2.189, 0.134) for high
hyperactivity/inattention, than it was for the difference between the estimates for lower SES
Page 15 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
16
females (B = 0.036, SE = 0.198) and males (B = -0.104, SE = 0.029), compared to higher SES
females (B = 0.029, SE = 0.077) and males (B = 0.029, SE = 0.077) for low
hyperactivity/inattention. This reflects a greater moderating effect of gender and SES at a
high compared to a low level of hyperactivity/inattention.
We did not observe any other potentially notable moderating effects of youth
adversity on baseline depression and anxiety symptoms, or change over time in these
symptoms, either for the individual characteristics as main effects or for their interactions
(reflecting intersectional effects).
< Table 3 about here >
Discussion
To our knowledge, this is the first study to investigate, a) the effect of youth adversity
on the trajectories of depression and anxiety symptoms in adolescence separately for
different intersectionality profiles (based on gender, SES, and hyperactivity/inattention), and
b) the role of intersectionality in differentiating the effect of youth adversity on these
trajectories. This investigation provides a novel contribution by finding evidence to suggest
that the effect of youth adversity on depression and anxiety symptoms at age 11-12-years is
most detrimental for males from lower SES backgrounds with high levels of
hyperactivity/inattention. We found evidence to suggest variability in the association
between youth adversity and the trajectories of depression and anxiety symptoms across
intersectionality profiles. Moderation effects attributable to the intersection of i) gender and
SES, and ii) gender, SES, and hyperactivity/inattention, were notable, albeit confidence
Page 16 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
17
intervals were wide, indicating uncertainty in the estimated effects and therefore a need for
caution when interpreting the results.
The findings from this study supported the hypothesis that youth adversity would be
associated with trajectories of elevated depression and anxiety symptoms across
intersectionality profiles. These findings are broadly in line with previous research that
reported that youth adversity was associated with an increased risk of being in a
high/increasing latent trajectory group for adolescent depression and anxiety symptoms
compared to a low scoring latent trajectory group (Leban, 2021). The current results extend
these findings by demonstrating that the detrimental effects of youth adversity are evident
across intersections of society indexed by gender, SES, and hyperactivity/inattention. Youth
adversity did not alter the rate of change over time in depression and anxiety symptoms with
any certainty, however, importantly average depression and anxiety symptoms started
higher and remained higher in the presence of youth adversity, compared to an absence of
youth adversity, for all intersectionality profiles.
Findings related to the second hypothesis that gender, SES, and
hyperactivity/inattention, and the intersections of these characteristics, would moderate the
relation between youth adversity and the latent growth factors were less clear. While a broad
level of moderation attributable to between group differences in intersectionality profiles can
be inferred through model comparison, moderation attributable to the characteristics that
were used to classify individuals into intersectionality profiles was only weakly evident for
two intersectional effects: i) for gender by SES, suggesting a greater moderating effect of
gender at a lower compared to a higher level of SES; and ii) for gender by SES by
hyperactivity/inattention interaction, suggesting a greater moderating effect of gender and
SES at a high compared to a low level of hyperactivity/inattention. It is possible that the broad,
Page 17 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
18
group-level differences may be reflective of some unmodelled characteristics in the data
between the groups. As a hypothetical example, differences between the intersectionality
profile groups could reflect characteristics at the school level that correlate with the
characteristics of the groups. For instance, neurotypical individuals from higher SES
backgrounds are more likely to attend grammar schools compared to nonselective schools
than neurodivergent individuals from lower SES backgrounds (Burgess et al., 2018). While
speculative, since school information is not available, this example serves to illustrate why
there could be moderation at the group-level that is not due to the variables that were
modelled.
In terms of moderation attributable to intersectionality, the uncertainty in the
estimates prompts caution in interpreting the observed effects. The interaction
(intersectional) effects in the current study were specified as compound parameters. While
these estimates cannot be used in the same way that maximum likelihood estimates could be
utilised, for example, to estimate sample size requirements for estimating the effects with
greater certainty in future studies (Hancock & French, 2013), it was nonetheless of interest to
investigate the effects of increasing the current sample size. The results of post-hoc analyses
suggested that doubling the sample size would result in greater certainty in the two specific
interaction effects that were discussed above. These pseudo power analyses give a crude
estimate of the extent to which sample sizes may need to be increased in future studies, in
samples with similar characteristics, in order to observe more certainty of the moderating
effects of intersectionality as they were conceptualised in the current study.
Despite the lack of certainty regarding moderation attributable to intersectionality,
the results allude to an intersectional profile characterised by male gender, low SES, and high
hyperactivity/inattention as being most vulnerable to experiencing an increase in depression
Page 18 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
19
and anxiety symptoms at age 11-12-years in the face of youth adversity, compared to an
absence of such. From a clinical perspective, while this may be motivation to pay particular
attention to adversity-exposed individuals meeting these profile characteristics, the current
findings invite further research to test the robustness of this result in other samples and using
other measures. Further, the results highlight the negative effects of youth adversity across
intersectionality profiles.
In addition to the effects of sample size, there are several other factors to consider
when interpreting the findings of uncertainty in terms of moderation due to intersectionality.
The current results only pertain to specific societal/social identities reflected by the
intersection of gender, SES, and hyperactivity/inattention, from many possible
characteristics. Of note, trans, nonbinary, and gender diverse identification was not measured
in this sample, but may be an important component of an individual’s intersectional identity
(e.g., Kidd et al., 2021). Other individual characteristics, such as, sexual orientation (Jonas et
al., 2022), immigration status (Kern et al., 2020), and ethnicity (Mersky et al., 2021) may
further be important factors to consider in the context of youth adversity and mental health.
These could not be considered in the current study due to absence of or insufficient data.
Specifically regarding ethnicity, intersectionality theory was developed in the context of black
feminism (Crenshaw, 1990). Thus, ethnicity is important theoretically, and as well,
preliminary empirical findings suggest that ethnicity may contribute to intersectional
differences in the development of depression symptoms in young people (Chen & Tung,
2023). Future work planned in our group will build on these findings by investigating the
effects of youth adversity in addition to including ethnicity as an intersectional characteristic
(Hosang et al., 2023).
Page 19 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
20
Of note, other recent work in our group found evidence to suggest that the cross-
sectional association between youth adversity and depression and anxiety at age 13-14-years
in the current sample (pooled with two other Year 9 cohorts from the HeadStart Cornwall
study) was moderated by gender, and SES, as individual (non-interacting) characteristics
(Havers et al., in preparation). This pattern of results was not observed at age 11-12-years
(baseline) in the current study. Collectively, these results could suggest that age 13-14-years
but not age 11-12-years (where the latent intercept factor was positioned in the current
study), represents a specific developmental window of vulnerability for differentiating the
effect of youth adversity on depression and anxiety symptoms in terms of gender and SES.
However, in repositioning the latent intercept factor to 13-14-years in post-hoc analysis of
the current data, moderating effects of gender and SES were not detected. Several factors
could be contributing to this divergence in findings. For example, youth adversity in the
current study included bullying victimisation across the study period, rather than solely at age
13-14-years as was the case in the cross-sectional study. More work in this area is required to
facilitate an in-depth evaluation of and discussion around the source/s of these discrepant
findings.
The results of this study have implications for research and practice since they expand
our understanding of the association between youth adversity and the development of
depression and anxiety symptoms in adolescence. The current findings should spur future
research in this area to explore the intersectionality of individual characteristics, which can
contribute towards a stratified approach to investigating mechanisms linking youth adversity
and mental health problems. In turn, this can inform theory development, with the ultimate
goal of informing clinical and/or community-level interventions that aim to reduce and
mitigate the negative impacts of youth adversity.
Page 20 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
21
Methodological considerations
The current study has a number of strengths including the use of both self-report and
objective (i.e., from local government, and School Census records) youth adversity data from
a large sample of adolescents. Utilising repeated measures of depression and anxiety
symptoms, this study was able to explore trajectories of mental health in the face of youth
adversity. However, there are several limitations that need to be considered when
interpreting the findings. First, the temporal impact of youth adversity on the trajectories of
depression and anxiety symptoms could not be ascertained, since adversities were those
reported across the study period, not prior to the reporting of symptoms at baseline. It is
important for future studies on this topic to adopt prospective longitudinal data collection so
that temporal relations between adversity and the development of mental health problems
can be delineated. Notwithstanding, our results show that individuals exposed to youth
adversity experience higher levels of depression and anxiety symptoms that remain higher
across ages 11-14-years compared to those that have not experienced adversity. Second,
information about the experience of specific forms of adversity was not available in this study.
Instead, the presence or absence of a range of different adversities was provided. More
granular detail would allow for the identification of specific youth adversities that may pose
a greater risk for the development of depression and anxiety symptoms. For example, neglect
and emotional abuse have been found to have an especially strong association with mental
health problems in adolescence and adulthood (Kisely et al., 2018; Mills et al., 2013). Future
research should expand on existing findings to explore the impact of specific types of youth
adversity on trajectories of depression and anxiety symptoms in young people in the context
of intersectionality.
Page 21 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
22
Conclusion
In summary, our results indicate that youth adversity has detrimental effects on the
development of depression and anxiety symptoms across ages 11-14-years, and that these
effects pervade across intersectionality profiles. Our findings add to a growing body of
literature that point to the negative impact of youth adversity on adolescent mental health
and underscore the pervasiveness of these effects across the societal intersections under
consideration in the current study. Although the certainty around intersectional moderation
was limited, our findings indicate that the intersectionality profile characterised by male
gender, lower SES, and high hyperactivity/inattention may be at a heightened risk of elevated
levels of depression and anxiety symptoms at age 11-12-years in the face of youth adversity.
An accumulation of research in this area is fundamental for drawing conclusions regarding
the extent to which intersectional identity is, or is not, an important contextual condition for
differentiating the effects of youth adversity on trajectories of depression and anxiety
symptoms in adolescence.
Page 22 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
23
Financial disclosures and acknowledgements
The ATTUNE project is funded by a cross council UK Research and Innovation [UKRI]
award (MR/W002183/1). The HeadStart Cornwall data was collected as part of the
HeadStart learning programme and supported by funding from The National Lottery
Community Fund. The content is solely the responsibility of the authors, and it does not
reflect the views of The National Lottery Community Fund. LH thanks Gregory Hancock
for providing such valuable guidance and support with the statistical modelling used in this
study. The authors are grateful to all the research teams and participants who have
contributed to the data that was analysed in this paper.
Competing interests
The authors have declared that no competing interests exist.
Correspondence
Dr Laura Havers. Address: Centre for Psychiatry & Mental Health, Wolfson Institute of
Population Health, Queen Mary, University of London, Yvonne Carter Building, 58 Turner
Street, London E1 2AB. Tel: 0207 882 2017. Email: l.havers@qmul.ac.uk
Page 23 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
24
Key points
Youth adversity is associated with persistent/increasing symptoms of depression and
anxiety over time in adolescence. This association may be greater for minoritized, or
disadvantaged groups
The intersection of individual characteristics may be important for moderating the
association between youth adversity and trajectories of depression and anxiety symptoms,
but this has not yet been tested
The current study found that youth adversity, compared to an absence, was
associated with higher depression and anxiety symptoms at baseline that remained higher
across ages 11-14-years, across intersectionality profiles (defined by gender, socio-economic
status, and hyperactivity/inattention). Weak evidence suggested some degree of
moderation attributable to the intersection between, i) gender and SES, and ii) gender, SES,
and hyperactivity/inattention
Future research should seek to assess the extent to which different
operationalisations of intersectionality may impact on detecting moderation
An accumulation of results from across studies is necessary to determine the extent
to which individuals’ intersectional profiles may provide a meaningful basis on which to
focus prevention and intervention efforts
Page 24 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
25
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.
https://doi.org/10.1007/s40653-021-00424-3
Bauer, D. J. (2017). A more general model for testing measurement invariance and
differential item functioning. Psychological Methods, 22, 507–526.
https://doi.org/10.1037/met0000077
Bevilacqua, L., Kelly, Y., Heilmann, A., Priest, N., & Lacey, R. E. (2021). Adverse childhood
experiences and trajectories of internalizing, externalizing, and prosocial behaviors
from childhood to adolescence. Child Abuse & Neglect, 112, 104890.
https://doi.org/10.1016/j.chiabu.2020.104890
Biswas, T., Scott, J. G., Munir, K., Renzaho, A. M. N., Rawal, L. B., Baxter, J., & Mamun, A. A.
(2020). Global variation in the prevalence of suicidal ideation, anxiety and their
correlates among adolescents: A population based study of 82 countries.
eClinicalMedicine, 24, 100395. https://doi.org/10.1016/j.eclinm.2020.100395
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. https://doi.org/10.2105/AJPH.2012.300750
Page 25 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
26
Burgess, S., Crawford, C., & Macmillan, L. (2018). Access to grammar schools by socio-
economic status. Environment and Planning A: Economy and Space, 50(7), 1381–
1385.
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.
https://doi.org/10.1001/jamanetworkopen.2020.3221
Chen, E. Y.-J., & Tung, E. Y.-L. (2023). Similarities and Differences in the Longitudinal
Trajectories of Depressive Symptoms from Mid-Adolescence to Young Adulthood:
The Intersectionality of Gender, Race/Ethnicity, and Levels of Depressive Symptoms.
Journal of Racial and Ethnic Health Disparities, 1–16.
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. https://doi.org/10.1007/s11920-
020-01193-w
Crenshaw, K. (1990). 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. The British Journal of Psychiatry, 215(3), 565–
567. https://doi.org/10.1192/bjp.2019.19
Page 26 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
27
Desch, J., Mansuri, F., Tran, D., Schwartz, S. W., & Bakour, C. (2023). The association
between adverse childhood experiences and depression trajectories in the Add
Health study. Child Abuse & Neglect, 137, 106034.
https://doi.org/10.1016/j.chiabu.2023.106034
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: The Adverse
Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4),
245–258. https://doi.org/10.1016/S0749-3797(98)00017-8
Ghavami, N., Katsiaficas, D., & Rogers, L. O. (2016). Chapter Two - 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 (Eds.), Advances in Child
Development and Behavior (Vol. 50, pp. 31–73). JAI.
https://doi.org/10.1016/bs.acdb.2015.12.001
Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A Research Note. Journal
of Child Psychology and Psychiatry, 38(5), 581–586. https://doi.org/10.1111/j.1469-
7610.1997.tb01545.x
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.
Hohls, J. K., König, H.-H., Quirke, E., & Hajek, A. (2021). Anxiety, depression and quality of
life—A systematic review of evidence from longitudinal observational studies.
International Journal of Environmental Research and Public Health, 18(22), 12022.
Page 27 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
28
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.
https://doi.org/10.1371/journal.pone.0289438
Hu, L., & Bentler, P. M. (1999). Cutoff criteria 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.
https://doi.org/10.1002/jcv2.12079
Kalmakis, K. A., & Chandler, G. E. (2014). Adverse childhood experiences: Towards a clear
conceptual meaning. Journal of Advanced Nursing, 70(7), 1489–1501.
https://doi.org/10.1111/jan.12329
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, Supplement), S12–S20.
https://doi.org/10.1016/j.jadohealth.2020.02.013
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
Page 28 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
29
National Comorbidity Survey Replication. Archives of General Psychiatry, 62(6), 593–
602. https://doi.org/10.1001/archpsyc.62.6.593
Kidd, K. M., Sequeira, G. M., Douglas, C., Paglisotti, T., Inwards-Breland, D. J., Miller, E., &
Coulter, R. W. S. (2021). Prevalence of Gender-Diverse Youth in an Urban School
District. Pediatrics, 147(6), e2020049823. https://doi.org/10.1542/peds.2020-
049823
Kisely, S., Abajobir, A. A., Mills, R., Strathearn, L., Clavarino, A., & Najman, J. M. (2018). Child
maltreatment and mental health problems in adulthood: Birth cohort study. The
British Journal of Psychiatry, 213(6), 698–703. https://doi.org/10.1192/bjp.2018.207
Leban, L. (2021). The effects of adverse childhood experiences and gender on
developmental trajectories of internalizing and externalizing outcomes. Crime &
Delinquency, 67(5), 631–661.
Liu, Y., Millsap, R. E., West, S. G., Tein, J.-Y., Tanaka, R., & Grimm, K. J. (2017). Testing
measurement invariance in longitudinal data with ordered-categorical measures.
Psychological Methods, 22(3), 486.
Marsh, H. W., Hau, K.-T., & Wen, Z. (2004). In search of golden rules: Comment on
hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in
overgeneralizing Hu and Bentler’s (1999) findings. Structural Equation Modeling,
11(3), 320–341.
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.
https://doi.org/10.1016/j.chiabu.2021.105066
Page 29 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
30
Michaud, P.-A., & Fombonne, E. (2005). Common mental health problems. BMJ, 330(7495),
835–838. https://doi.org/10.1136/bmj.330.7495.835
Mills, R., Scott, J., Alati, R., O’Callaghan, M., Najman, J. M., & Strathearn, L. (2013). Child
maltreatment and adolescent mental health problems in a large birth cohort. Child
Abuse & Neglect, 37(5), 292–302. https://doi.org/10.1016/j.chiabu.2012.11.008
Morin, A. J. S., Maïano, C., Nagengast, B., Marsh, H. W., Morizot, J., & Janosz, M. (2011).
General Growth Mixture Analysis of Adolescents’ Developmental Trajectories of
Anxiety: The Impact of Untested Invariance Assumptions on Substantive
Interpretations. Structural Equation Modeling: A Multidisciplinary Journal, 18(4),
613–648. https://doi.org/10.1080/10705511.2011.607714
Schubert, K. O., Clark, S. R., Van, L. K., Collinson, J. L., & Baune, B. T. (2017). Depressive
symptom trajectories in late adolescence and early adulthood: A systematic review.
Australian & New Zealand Journal of Psychiatry, 51(5), 477–499.
Shorey, S., Ng, E. D., & Wong, C. H. J. (2022). Global prevalence of depression and elevated
depressive symptoms among adolescents: A systematic review and meta-analysis.
British Journal of Clinical Psychology, 61(2), 287–305.
https://doi.org/10.1111/bjc.12333
Tully, P. J., Harrison, N. J., Cheung, P., & Cosh, S. (2016). Anxiety and Cardiovascular Disease
Risk: A Review. Current Cardiology Reports, 18(12), 120.
https://doi.org/10.1007/s11886-016-0800-3
van de Schoot, R., Lugtig, P., & Hox, J. (2012). A checklist for testing measurement
invariance. European Journal of Developmental Psychology, 9(4), 486–492.
https://doi.org/10.1080/17405629.2012.686740
Page 30 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
31
Walsh, D., McCartney, G., Smith, M., & Armour, G. (2019). Relationship between childhood
socioeconomic position and adverse childhood experiences (ACEs): A systematic
review. J Epidemiol Community Health, 73(12), 1087–1093.
https://doi.org/10.1136/jech-2019-212738
World Health Organization. (2017). Depression and Other Common Mental Disorders: Global
Health Estimates.
Page 31 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
32
Table 1. Sample Description
Year 7
(11-12-years)
Year 8
(12-13-years)
Year 9
(13-14-years)
Variable
N = 4,575
N = 4,600
N = 3,604
Gender
Female
Male
No data
2,270 (49.62%)
2,303 (50.34%)
2 (< 1%)
2,335 (49.24%)
2,265 (51.76%)
0 (0%)
1,785 (49.53%)
1,819 (50.47%)
0 (0%)
SES
Lower
Higher
No data
674 (14.73%)
3,901 (85.27%)
0 (0%)
668 (14.52%)
3,932 (85.48%)
0 (0%)
555 (15.40%)
3,048 (84.57%)
1 (<1%)
Hyperactivity/inattention
High
Low
No data
1,170 (25.57%)
3,280 (71.69%)
125 (2.73%)
1,199 (26.07%)
3,291 (71.54%)
110 (2.39%)
1,002 (27.80%)
2,529 (70.17%)
73 (2.03%)
Youth adversity
Yes
2,444 (53.42%)
2,435 (52.93%)
1,958 (54.33%)
Page 32 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
33
No
No data
2,108 (46.08%)
23 (<1%)
2,151 (46.76%)
14 (<1%)
1,637 (45.42%)
9 (<1%)
Depression and anxiety symptoms (0-10)
N with total score data
Mean (SD)
Range
Median
4,462
3.82 (2.54)
0-10
4
4,500
3.89 (2.63)
0-10
4
3,537
4.32 (2.75)
0-10
4
Note. N: number of individuals. SES: socio-economic status
Page 33 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
34
Table 2. Youth Adversity and Depression and Anxiety Symptoms by Intersectionality Profiles
Depression and anxiety symptoms by school year
Mean (SD) c
Intersectionality profile
N (%) in
intersectionality
profile a
N (%) of
intersectionality
profile reporting
youth adversity b
Year 7 (age 11-12-years)
Year 8 (age 12-13-years)
Year 9 (age 13-14-years)
Males, Higher SES, Low Hyperactivity/inattention
1,327 (29.84%)
623 (47.02%)
2.86 (2.21), N = 1,327
2.77 (2.24), N = 1,185
2.92 (2.39), N = 883
Females, Higher SES, Low Hyperactivity/inattention
1,519 (34.18%)
692 (45.59%)
4.01 (2.45), N = 1,519
4.43 (2.59), N = 1,364
5.06 (2.56), N = 1,024
Males, Lower SES, Low Hyperactivity/inattention
188 (4.23%)
143 (76.06%)
3.42 (2.49), N = 188
3.26 (2.38), N = 144
3.20 (2.55), N = 128
Females, Lower SES, Low Hyperactivity/inattention
245 (5.47%)
167 (68.72%)
4.39 (2.53), N = 245
4.88 (2.55), N = 203
5.84 (2.51), N = 162
Males, Higher SES, High Hyperactivity/inattention
591 (13.29%)
356 (60.34%)
3.93 (2.47), N = 591
3.69 (2.58), N = 507
3.63 (2.55), N = 383
Females, Higher SES, High Hyperactivity/inattention
362 (8.15%)
226 (62.43%)
5.34 (2.59), N = 362
5.14 (2.68), N = 296
5.88 (2.59), N = 235
Males, Lower SES, High Hyperactivity/inattention
121 (2.70%)
98 (81.67%)
4.52 (2.51), N = 121
3.60 (2.59), N = 99
3.50 (2.56), N = 68
Females, Lower SES, High Hyperactivity/inattention
95 (2.14%)
82 (86.32%)
6.11 (2.50), N = 95
6.24 (2.71), N = 69
6.31 (2.31), N = 54
Note. N: number of individuals. SES: socio-economic status
a N = 4,448 with intersectionality profile data
b percentage of individuals in intersectionality profile with youth adversity data across the study period (N = 4,441)
c individuals in intersectionality profile with depression and anxiety symptoms data at each time-point
Page 34 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
35
Table 3. Compound Parameter Estimates of Youth Adversity as a Predictor of the Latent Growth Factors of Depression and Anxiety Symptoms
Latent intercept factor regressed on youth adversity a
Latent slope factor regressed on youth adversity b
Parameter description
Parameter
estimate (SE)
95% CI
Standardised
estimate (SE)
95% CI
Parameter
estimate (SE)
95% CI
Standardised
estimate (SE)
95% CI
Weighted averages
Males
Females
Higher SES
Lower SES
Low hyperactivity/inattention
High hyperactivity/inattention
1.635 (0.093)
1.518 (0.105)
1.564 (0.072)
1.647 (0.211)
1.521 (0.077)
1.730 (0.151)
1.445, 1.816
1.298, 1.716
1.424, 1.718
1.185, 2.075
1.372, 1.676
1.426, 2.022
0.928 (0.065)
0.746 (0.060)
0.823 (0.044)
0.858 (0.141)
0.825 (0.050)
0.840 (0.085)
0.794, 1.054
0.630, 0.860
0.742, 0.915
0.574, 1.107
0.737, 0.925
0.674, 1.006
-0.018 (0.063)
0.019 (0.066)
0.010 (0.048)
-0.058 (0.135)
0.022 (0.051)
-0.059 (0.099)
-0.144, 0.111
-0.115, 0.142
-0.095, 0.104
-0.329, 0.227
-0.081, 0.124
-0.259, 0.138
-0.020 (0.073)
0.019 (0.069)
0.011 (0.051)
-0.066 (0.200)
0.024 (0.056)
-0.057 (0.101)
-0.165, 0.130
-0.119, 0.146
-0.099, 0.110
-0.427, 0.273
-0.088, 0.133
-0.247, 0.144
Weighted main effects
Gender (male - female)
SES (higher - lower)
Hyperactivity/inattention (low - high)
0.117 (0.145)
-0.083 (0.230)
-0.209 (0.170)
-0.165, 0.396
-0.517, 0.389
-0.564, 0.126
0.061 (0.067)
-0.044 (0.122)
-0.110 (0.090)
-0.084, 0.208
-0.269, 0.199
-0.296, 0.068
-0.037 (0.091)
0.068 (0.143)
0.081 (0.111)
-0.208, 0.155
-0.232, 0.349
-0.122, 0.297
-0.039 (0.100)
0.072 (0.159)
0.086 (0.119)
-0.223, 0.166
-0.248, 0.380
-0.137, 0.314
Weighted interaction effects
Gender x SES
Gender x hyperactivity/inattention
-0.606 (0.461)
0.298 (0.352)
-1.523, 0.261
-0.358, 0.997
-0.319 (0.243)
0.157 (0.186)
-0.820, 0.121
-0.187, 0.533
0.512 (0.291)
0.041 (0.232)
-0.038, 1.068
-0.422, 0.492
0.543 (0.311)
0.043 (0.248)
-0.042, 1.153
-0.449, 0.532
Page 35 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
36
SES x hyperactivity/inattention
Gender x SES x hyperactivity/inattention
0.670 (0.524)
0.015 (1.052)
-0.367, 1.731
-1.902, 2.155
0.352 (0.277)
0.008 (0.556)
-0.190, 0.920
-1.026, 1.115
-0.025 (0.350)
-1.183 (0.687)
-0.711, 0.627
-2.655, 0.127
-0.027 (0.374)
-1.254 (0.737)
-0.772, 0.659
-2.819, 0.134
Note. N = 4,441 (with intersectionality profile and youth adversity data). N: number of individuals. SES: socio-economic status. Parameter estimates from maximum likelihood estimation, with
SE and bias-corrected bootstrapped CI from 1,000 draws. Weighted pooled standard deviations used for calculation of standardised estimates. Compound parameter estimates specified using
the maximum likelihood estimates derived from the multiple group model of youth adversity as a predictor of the latent growth factors of depression and anxiety symptoms, weighted by
intersectionality profile sample size. Results were substantively unchanged where missing youth adversity data was imputed (10 datasets, N = 4,448)
a average effect of youth adversity on depression and anxiety symptoms at baseline (age 11-12-years)
b average effect of youth adversity on change over time in depression and anxiety symptoms
Page 36 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
37
Figure 1. Schematic Diagram of the Multiple Group Conditional Latent Growth Model
Page 37 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
38
Note. SES: socio-economic status. Intersectional profile (depicted at the centre of the Venn diagram, above left) is used as the grouping variable in a multiple group model, where everything
inside of the box (above right) is estimated for each intersectionality profile group. Compound parameters are further estimated, reflecting the extent to which the individual characteristics,
as well as the intersections between them, moderate the paths from youth adversity to the latent growth factors. The observed depression/anxiety symptoms scores are indicators of the
latent growth factors. A non-schematic, labelled path diagram is shown in Supplementary Figure 1
Page 38 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
39
Figure 2. Effect of Youth Adversity on Average Depression and Anxiety Symptoms Trajectories by Intersectionality Profiles
Page 39 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
40
Note. SES: socio-economic status. Y axis: depression and anxiety symptoms total observed score. School year corresponds to the following average ages: 11-12-years (Year 7), 12-13-years
(Year 8), 13-14-years (Year 9)
Page 40 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Materials 1. Supporting Families Programme Eligibility Criteria (for “On Family List” classification)
1.1: An adult or child who has committed a proven offence in the previous 12 months
1.2: An adult or child who has received an anti-social behaviour intervention in the previous 12 months
1.3: An individual in the household is known to the Anti-Social Behaviour Team for incidents of anti-social behaviour but has not
received a formal intervention
2.1: A child whose school attendance is <90% across the last 3 terms excluding authorised absences
2.2: A child with at least 3 fixed term exclusions in the last 3 terms
2.3: A child who has been permanently excluded in the last 3 school terms
2.4: A child who is in an alternative education provision to improve their behaviour (not SEN pupils)
2.5: A child who is known to the Education Welfare Service as a ‘Child Not In School’ (CNIS)
2.6: A child identified as having a score below threshold in communication skills in the 2-2.5-year-old health check or Primary
School assessment (school readiness).
3.1: A child with a ‘Common Assessment Frameworkor ‘Early Help Plan’ in the previous 12 months
3.2: A ‘Child In Need’ under section 17 of The Children Act 1989 in the previous 12 months
3.3: A child which has been listed as missing from home in the previous 12 months
3.4: A child is identified as at risk of sexual exploitation
3.5: A young person aged under 19 became a parent in the past 12 months
3.6: A child who is a young carer
4.1: An adult in receipt of out-of-work benefits (or Universal Credit, if relevant), except those claiming carers allowance only,
where worklessness is not considered a problem for the family
4.2: A young person aged 16 – 19 who is not in employment, education, or training
4.3: The family have problematic or unmanageable levels of debt
4.4: The family are homeless
4.5: The family are threatened with or at risk of homelessness
5.1: An individual who has experienced or is currently experiencing domestic abuse and has been engaged with specialist
services in the past 12 months
5.2: An individual in the household discloses domestic abuse to a key worker or other professional and is not engaged with
specialist services
6.1: An individual currently undergoing or who has undergone treatment for problem use of alcohol and/or other drugs in the
last 12 months
6.2: An individual in the household discloses problem use of alcohol and/or other drugs to a key worker or other professional
and is not engaged with specialist services
6.3: There is unmanaged physical or mental illness or disability within the household
6.4: A child on Universal Plus Higher or Universal Partnership Plus pathways
Page 41 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Materials 2. Additional information on latent growth modelling
In a linear growth model, a latent intercept and a latent slope factor are specified to capture the two components of a latent
trajectory (latent growth). With more than three time-points of data, nonlinear models can also be specified. The intercept and
slope factors are latent because they are not directly observed but are estimated based on the relations among the variables
that have been observed. The repeated measures are specified as indicators of the latent growth factors. In this modelling
framework, the average within-person trajectory is estimated (represented by the means of the latent intercept and the latent
slope), as well as the between-person variability around the averages (represented by the variances of the latent intercept and
the latent slope). Note that in a model where the latent growth factors are regressed on another/other variable/s and are thus
endogenous (dependent), the averages are represented as intercepts, and the variances are represented as residual variances.
Residual variances of the repeated measures are also estimated, reflecting time-specific variance not explained by the
(conditional or unconditional) latent growth factors.
Page 42 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Materials 3. Mplus script for conditional multiple group latent growth model
TITLE: Conditional multiple group latent growth model
VARIABLE:
USEVARIABLES =
Yr7_SDQ
Yr8_SDQ
Yr9_SDQ
ACES;
GROUPING =
IS_profile
(0 = 0 1 = 1
2 = 2 3 = 3
4 = 4 5 = 5
6 = 6 7 = 7);
ANALYSIS:
ESTIMATOR = MLR;
MODEL:
!! Specify latent intercept (i) and slope (s) factors
i s | Yr7_SDQ@0 Yr8_SDQ@1 Yr9_SDQ@2;
!! This specification estimates the following:
i s; ! residual variances of latent intercept and slope factors
i WITH s; ! residual covariance of latent intercept and slope factors
Yr7_SDQ Yr8_SDQ Yr9_SDQ; ! residual variances of repeated measures (observed SDQ)
[i s]; ! intercepts of latent intercept and slope factors
!! Specify structural paths
i ON ACES; ! latent intercept factor regressed on observed youth adversity measure
s ON ACES; ! latent slope factor regressed on observed youth adversity measure
!! All parameters estimated separately for each intersectionality profile group (defined under grouping command)
Page 43 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Materials 4. Additional information on longitudinal measurement invariance analysis
One approach for assessing measurement invariance is to specify a multiple group confirmatory factor analysis. Using this
approach, a series of models are sequentially specified to assess the extent to which different parameters for a measurement
model can be considered sufficiently equivalent (invariant) across measurement occasions. First, a configural model tests for
invariance in the general configuration of items to factors. Second, a metric (or ‘weak invariance’) model tests for invariance in
the factor loadings. Third, a scalar (or ‘strong invariance’) model tests for invariance in the item thresholds (for ordered
categorical data, item intercepts for continuous data), in addition to the factor loadings. A unique factor (or ‘strict invariance’)
model can be used to test for invariance in the time-specific residual variances, though this level of invariance is not typically
assessed
Page 44 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 1. Data Across School Years
Total
Complete
Year 7 and year
8 only
Year 7 and
year 9 only
Year 8 and
year 9 only
Year 7 only
Year 8 only
Year 9 only
No data
N
5,336
2,614 (48.99%)
1,262 (23.65%)
330 (6.18%)
366 (6.86%)
256 (4.80%)
258 (4.84%)
227 (4.25%)
23 (< 1%)
Note. N: number of individuals. Exact age data is not available for this sample. Based on averages in the United Kingdom, year 7 = age 11-12-years, year 8 = age 12-13-years, and year 9 = age
13-14-years
Page 45 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 2. Model Fit Information for Confirmatory Factor Analysis of Depression and Anxiety as a Common Factor Model at Each School Year
N
Par.
Test Statistic (df)
CFI
RMSEA [90% CI]
SRMR
Omega [95% CI]
Year 7 (age 11-12-years)
4,262
15
71.096 (5), p < .001
0.991
0.054 [0.044, 0.066]
0.020
0.800 [0.791, 0.811]
Year 8 (age 12-13-years)
4,500
15
139.048 (5), p < .001
0.986
0.077 [0.066, 0.089]
0.026
0.826 [0.815, 0.835]
Year 9 (age 13-14-years)
3,537
15
98.370 (5), p < .001
0.990
0.073 [0.061, 0.086]
0.023
0.850 [0.839, 0.859]
Note. N: number of individuals. Par: number of parameters. CFI: comparative fit index. RMSEA: root mean square error of approximation. SRMR: standardised root mean square residual.
Diagonally weighted least squares (DWLS) estimation with mean and variance adjustment (WLSMV), using pairwise present data. Omega estimates with bias-corrected bootstrapped CI from
1,000 draws
Page 46 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 3. Model Fit Information for Longitudinal Measurement Invariance Analysis of Depression and Anxiety as a Common Factor Model
Model
Model fit
Chi-square difference test
Change values
Par.
Test statistic (df)
CFI
RMSEA [90% CI]
SRMR
Test statistic (difference df)
CFI
RMSEA
SRMR
Configural
63
458.794 (72), p < .001
0.990
0.032 [0.029, 0.035]
0.026
-
-
-
-
Metric (constrained loadings)
55
571.113 (80), p < .001
0.987
0.034 [0.031, 0.037]
0.028
104.101 (8), p < .001
0.003
-0.002
-0.002
Scalar (constrained thresholds and
loadings)
47
656.960 (88), p < .001
0.985
0.035 [0.032, 0.0327]
0.028
92.948 (8), p < .001
0.002
-0.001
0.000
Note: N = 5,313. N: number of individuals. Par: number of parameters. CFI: comparative fit index. RMSEA: root mean square error of approximation. SRMR: standardised root mean square
residual. Diagonally weighted least squares (DWLS) estimation with mean and variance adjustment (WLSMV), using pairwise present data
Page 47 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 4. Model Fit Information for Latent Growth Models of Depression and Anxiety Symptoms
Model
Par.
Test statistic (df)
CFI
RMSEA [90% CI]
SRMR
BIC
AIC
Unconditional model
8
15.324 (1), p < .001
0.994
0.057 [0.034, 0.083]
0.013
50,443.819
50,418.399
Conditional (youth adversity) model
10
15.841 (2), p < .001
0.996
0.039 [0.023, 0.059]
0.013
49,838.734
49,774.747
Conditional (youth adversity) multiple group model
80
40.975 (16), p < .001
0.992
0.053 [0.033, 0.073]
0.021
49,413.122
48,901.231
Note. N = 4,441 (with intersectionality profile and youth adversity data). N: number of individuals. Par: number of parameters. CFI: comparative fit index. RMSEA: root mean square error of
approximation. SRMR: standardised root mean square residual. BIC: Bayesian information criterion. AIC: Akaike information criterion. Full information maximum likelihood estimation, with
robust adjustment (MLR). In the conditional model, the latent growth factors are regressed on the observed youth adversity variable. In the multiple group model, separate parameters are
estimated for each of the intersectionality profiles
Page 48 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 5. Parameter Estimates from the Multiple Group Model of Youth Adversity as a Predictor of the Latent Growth Factors of Depression and Anxiety Symptoms
Intersectionality profile
N
Intercepts of latent growth factors a
Residual variances of latent
growth factors b
Regression paths of latent growth factors
regressed on youth adversity c
Latent intercept-slope
residual covariance d
Males, Higher SES, Low
Hyperactivity/inattention
1,325
Intercept
Slope
2.093 (0.067), p < .001
0.020 (0.020), p = .647
Intercept
Slope
2.667 (0.282), p < .001
0.718 (0.149), p < .001
Intercept on YA
Slope on YA
1.591 (0.113), p < .001
0.029 (0.077), p = .709
-0.574 (0.173), p = .001
Females, Higher SES, Low
Hyperactivity/inattention
1,518
Intercept
Slope
3.308 (0.076), p < .001
0.504 (0.050), p < .001
Intercept
Slope
4.040 (0.274), p < .001
0.979 (0.140), p < .001
Intercept on YA
Slope on YA
1.496 (0.121), p < .001
0.029 (0.077), p = .705
-0.878 (0.168), p < .001
Males, Lower SES, Low
Hyperactivity/inattention
188
Intercept
Slope
2.053 (0.246), p < .001
0.005 (0.156), p = .976
Intercept
Slope
3.218 (0.818), p < .001
0.840 (0.426), p = .049
Intercept on YA
Slope on YA
1.766 (0.322), p < .001
-0.104 (0.029), p = .618
-0.700 (0.501), p = .162
Females, Lower SES, Low
Hyperactivity/inattention
243
Intercept
Slope
3.577 (0.263), p < .001
0.668 (0.154), p < .001
Intercept
Slope
3.574 (0.744), p < .001
0.368 (0.376), p = .328
Intercept on YA
Slope on YA
1.106 (0.329), p = .001
0.036 (0.198), p = .854
-0.477 (0.440), p = .278
Males, Higher SES, High
Hyperactivity/inattention
590
Intercept
Slope
2.996 (0.143), p < .001
-0.213 (0.092), p = .021
Intercept
Slope
4.042 (0.502), p < .001
1.057 (0.260), p < .001
Intercept on YA
Slope on YA
1.551 (0.192), p < .001
0.028 (0.128), p = .827
-1.142 (0.316), p < .001
Females, Higher SES, High
Hyperactivity/inattention
362
Intercept
Slope
4.183 (0.211), p < .001
0.313 (0.136), p = .021
Intercept
Slope
4.565 (0.633), p < .001
1.115 (0.319), p < .001
Intercept on YA
Slope on YA
1.770 (0.264), p < .001
-0.163 (0.171), p = .340
-0.907 (0.367), p = .013
Males, Lower SES, High
Hyperactivity/inattention
120
Intercept
Slope
2.539 (0.462), p < .001
-0.068 (0.338), p = .840
Intercept
Slope
3.108 (0.965), p = .001
0.600 (0.569), p = .292
Intercept on YA
Slope on YA
2.318 (0.521), p < .001
-0.626 (0.372), p = .092
-0.336 (0.589), p = .569
Females, Lower SES, High
Hyperactivity/inattention
95
Intercept
Slope
4.422 (0.671), p < .001
-0.301 (0.400), p = .452
Intercept
Slope
5.655 (1.413), p < .001
1.779 (0.594), p = .003
Intercept on YA
Slope on YA
1.955 (0.719), p = .007
0.507 (0.437), p = .246
-2.475 (0.798), p = .002
Note. N = 4,441 (with intersectionality profile and youth adversity data). N: number of individuals. SES: socio-economic status. YA: youth adversity. Par: number of parameters. Full
information maximum likelihood estimation, with robust adjustment (MLR). SE in parentheses. Results were substantively unchanged where missing youth adversity data was imputed (10
datasets, N = 4,448). School year-specific residual variances shown in Supplementary Table 6
a average level of depression and anxiety symptoms at age 11-12-years, (intercept), and change over time in these symptoms (slope) in the absence of youth adversity
b variance of depression and anxiety symptoms at age 11-12-years (intercept) and change over time in these symptoms (slope). Variances are residual because the latent growth factors are
regressed on youth adversity
c average effect of youth adversity on depression and anxiety symptoms at age 11-12-years (intercept on YA), and on change over time in these symptoms (slope on YA)
b covariance between depression and anxiety symptoms at age 11-12-years (intercept) and change over time in these symptoms (slope). The covariance is residual because the latent growth
factors are regressed on youth adversity
Page 49 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Table 6. School Year-Specific Residual Variance Parameter Estimates from the Multiple Group Model of Youth Adversity as a Predictor of the Latent Growth Factors of
Depression and Anxiety Symptoms
Intersectionality profile
N
Residual variances
Males, Higher SES, Low Hyperactivity/inattention
1,325
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
1.591 (0.272), p < .001
2.285 (0.149), p < .001
1.694 (0.328), p < .001
Females, Higher SES, Low Hyperactivity/inattention
1,518
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
1.432 (0.254), p < .001
2.748 (0.138), p < .001
1.603 (0.288), p < .001
Males, Lower SES, Low Hyperactivity/inattention
188
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
2.344 (0.744), p = .002
2.469 (0.434), p < .001
2.030 (0.426), p = .049
Females, Lower SES, Low Hyperactivity/inattention
243
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
2.550 (0.755), p = .001
2.978 (0.409), p < .001
2.929 (0.853), p = .001
Males, Higher SES, High Hyperactivity/inattention
590
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
1.464 (0.465), p = .002
3.189 (0.248), p < .001
2.181 (0.582), p < .001
Females, Higher SES, High Hyperactivity/inattention
362
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
1.332 (0.576), p = .021
3.157 (0.324), p < .001
0.875 (0.668), p = .191
Males, Lower SES, High Hyperactivity/inattention
120
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
2.221 (0.998), p = .026
3.220 (0.548), p < .001
2.163 (1.401), p = .123
Females, Lower SES, High Hyperactivity/inattention
95
Year 7 (11-12-years)
Year 8 (12-13-years)
Year 9 (13-14-years)
0.093 (1.201), p = .938
3.724 (0.615), p < .001
1.094 (1.147), p = .340
Note. N = 4,441 (with intersectionality profile and youth adversity data). N: number of individuals. SES: socio-economic status. Full information maximum likelihood estimation, with robust
adjustment (MLR). SE in parentheses. Results were substantively unchanged where missing youth adversity data was imputed (10 datasets, N = 4,448)
Page 50 of 51JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Supplementary Figure 1. Path Diagram of the Multiple Group Model of Youth Adversity as a Predictor of the Latent Growth Factors of Depression and Anxiety Symptoms
Note. Simplified path diagram of youth adversity predicting the latent growth factors of depression and anxiety symptoms measured at school years 7 (11-12-years), 8 (12-13-years), and 9
(13-14-years). Mean structure omitted for simplification. In this diagram, the repeated measures are labelled as V1-V3, representing depression and anxiety symptoms observed total scores
at school years 7-9. The two latent factors of the linear growth trajectory are labelled as LF1 and LF2, representing the latent intercept and latent slope components, respectively. The latent
factors are regressed on the observed youth adversity variable (labelled as V4), and the regression paths are labelled as bF1V4 and bF2V4, for the latent intercept and latent slope,
respectively. The residual variance parameters of the repeated measures are not labelled but are depicted in the curved arrows of the residuals, labelled as R7-R9. The residual covariance of
the latent growth factors is reflected in the curved arrow between the latent residuals, RInt and Rslo (the residual variance parameters of the latent growth factors are not labelled). The
conditional latent growth model within the box is estimated separately for each intersectionality profile, schematically represented by a grouping variable, labelled as G
Page 51 of 51 JCPP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Article
Full-text available
Background: Individuals from disadvantaged groups in society are at heightened risk of experiencing low self-esteem. However, limited research has explored self-esteem at the intersection of sex, socioeconomic status (SES), and immigrant origin. Method: This study examines self-esteem among 91,560 adolescents aged 14 to 17 years in Finland, drawing on data from the nationwide School Health Promotion Study. The analysis focuses on how sex, SES, and immigrant origin intersect to influence self-esteem. Results: An analysis of variance revealed large effect sizes for sex, medium effect sizes for SES, and very small effect sizes for immigrant origin on self-esteem. Girls, adolescents from lower SES backgrounds, and 1st generation immigrant and multicultural students exhibited lower self-esteem compared to their peers. A three-way interaction analysis showed that while both boys and girls who were 1st generation immigrants with low SES had the lowest self-esteem, 1st generation immigrant boys from low SES backgrounds were particularly vulnerable to low self-esteem relative to other boys. Conclusion: The independent and interactive effects of sex, SES, and immigrant origin are discussed within the context of Finland, a country known for its relatively low economic and gender inequality. These findings highlight specific groups of adolescents who may be at greater risk of low self-esteem.
-12-years) Year 8 (12-13-years) Year 9
  • Males
  • Ses Lower
  • High Hyperactivity
Males, Lower SES, High Hyperactivity/inattention 120 Year 7 (11-12-years) Year 8 (12-13-years) Year 9 (13-14-years) 2.221 (0.998), p =.026 3.220 (0.548), p <.001 2.163 (1.401), p =.123
-12-years) Year 8 (12-13-years) Year 9
  • Males
  • Ses Higher
  • High Hyperactivity
Males, Higher SES, High Hyperactivity/inattention 590 Year 7 (11-12-years) Year 8 (12-13-years) Year 9 (13-14-years) 1.464 (0.465), p =.002 3.189 (0.248), p <.001 2.181 (0.582), p <.001
-12-years) Year 8 (12-13-years) Year 9
  • Females
  • Ses Higher
  • High Hyperactivity
Females, Higher SES, High Hyperactivity/inattention 362 Year 7 (11-12-years) Year 8 (12-13-years) Year 9 (13-14-years) 1.332 (0.576), p =.021 3.157 (0.324), p <.001 0.875 (0.668), p =.191