You are on page 1of 40

ARTICLE IN PRESS

META-ANALYSIS

Systematic Review and Meta-analysis: The Science of


Early-Life Precursors and Interventions for Attention-
Deficit/Hyperactivity Disorder
Elizabeth Shephard, PhD, Pedro F. Zuccolo, PhD, Iman Idrees, MSc, Priscilla B.G. Godoy, MSc,
Erica Salomone, PhD, Camilla Ferrante, MSc, Paola Sorgato, MSc, Luís F.C.C. Cata ~o, MD candidate,
Amy Goodwin, PhD, Patrick F. Bolton, PhD, FRCPsych, Charlotte Tye, PhD,
Madeleine J. Groom, PhD, Guilherme V. Polanczyk, MD, PhD

Objective: To evaluate which early neurocognitive and behavioral precursors are associated with the development of attention-deficit/hyperactivity dis-
order (ADHD) and whether these are currently targeted in early interventions.
Method: We conducted 2 systematic reviews and meta-analyses of empirical studies to examine the following: (1) early-life (0 5 years) neurocognitive
and behavioral precursors associated with familial likelihood for ADHD, an early ADHD diagnosis/elevated ADHD symptoms, and/or the presence of
later-childhood ADHD; and (2) interventions delivered to children aged 0 to 5 years targeting the identified precursors or measuring these as outcomes.
Standardized mean differences (Hedges’ g) and pre-post-treatment change scores (SMD) were computed.
Results: A total of 149 studies (165,095 participants) investigating 8 neurocognitive and behavioral domains met inclusion criteria for part 1. Multi-
level random-effects meta-analyses on 136 studies revealed significant associations between ADHD and poorer cognitive (g = 0.46 [95% CIs: 0.59,
0.33]), motor (g = 0.35 [CIs: 0.48, 0.21]) and language (g = 0.43 [CIs: 0.66, 0.19]) development, social (g = 0.23 [CIs: 0.03, 0.43]) and
emotional (g = 0.46 [CIs: 0.33, 0.58]) difficulties, early regulatory (g = 0.30 [CIs: 0.18, 0.43]) and sleep (g = 0.29 [CIs: 0.14, 0.44]) problems, sensory
atypicalities (g = 0.52 [CIs: 0.16, 0.88]), elevated activity levels (g = 0.54 [CIs: 0.37, 0.72]), and executive function difficulties (g = 0.34 [CIs: 0.05,
0.64] to 0.87 [CIs: 1.35, 0.40]). A total of 32 trials (28 randomized, 4 nonrandomized, 3,848 participants) testing early interventions that targeted
the identified precursors met inclusion criteria for part 2. Multi-level random-effects meta-analyses on 22 studies revealed significant intervention-related
improvements in ADHD symptoms (SMD = 0.43 [CIs: 0.22, 0.64]) and working memory (SMD = 0.37 [CIs: 0.06, 0.69]).
Conclusion: Children aged 0 to 5 years with current or later-emerging ADHD are likely to experience difficulties in multiple neurocognitive/behavioral
functions. Early interventions show some effectiveness in reducing ADHD symptoms, but their effects on neurocognitive/behavioral difficulties require
further study.
Key words: attention-deficit/hyperactivity disorder (ADHD), early-life precursors, neurocognition, early intervention, prevention
J Am Acad Child Adolesc Psychiatry 2021;000(000):1−40.

ttention-deficit/hyperactivity disorder (ADHD) development.6 Nonpharmacological treatments have been

A is a neurodevelopmental condition that is usually


diagnosed in the school-age years or later, after
symptoms have clearly emerged and are causing functional
shown to produce modest or no effect at all when used
alone,7 although there is some evidence of improvements in
symptoms and associated functional impairments when
impairments.1,2 It is well established that medications pro- combined with medication.8 Therefore, new scientific and
duce clinically relevant reduction of symptom severity,3 and clinical approaches are needed to advance the therapeutics
growing evidence indicates that they also have an impact on for ADHD.1 One possible approach, inspired by research in
functional outcomes, such as reducing rates of accidents the field of autism, another neurodevelopmental condition
and injuries.4 Nevertheless, currently available treatments with substantial etiological, cognitive, and clinical overlap
for ADHD have important limitations. Long-term adher- with ADHD,9,10 is to invest in early identification and
ence to medications is low, and evidence of long-term bene- intervention as a strategy to maximize the benefits of treat-
fits is scarce.5,6 Symptoms are controlled while treatment is ment on the developmental trajectory of the disorder.
administered, but there is no evidence to indicate long-last- Indeed, increasing evidence indicates that ADHD
ing or permanent effects in ameliorating symptoms over begins very early in life and that there are identifiable

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 1


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

neurocognitive and behavioral precursors before reliable preschool), neurocognitive and behavioral atypicalities asso-
diagnosis can be established.11,12 Prospective longitudinal ciated with ADHD emerge. We investigated this question
studies have shown that atypicalities in brain function, cog- by examining whether developmental stage, defined as
nition, and behavior are present as early as the first year of infancy/toddlerhood (0 2 years) versus preschool (3 5
life and predict the later emergence of ADHD symptoms years) significantly moderated associations between neuro-
and/or are associated with increased familial likelihood for cognitive and behavioral atypicalities and ADHD in the first
ADHD.13−18 There have been previous selective, narrative meta-analysis.
reviews on this topic,11,19,20 as well as 1 systematic review
examining early motor atypicalities preceding ADHD21 and METHOD
1 meta-analysis focused on infant negative emotionality and PRISMA Statement and Preregistration
self-regulation.22 These studies indicate multiple possible Methodology and reporting for this systematic review and
precursors across different developmental domains, but the meta-analysis are consistent with the PRISMA statement
predictive ability of each remains unclear. It is important to (see PRISMA Checklist in Supplement 1, available online).
document the relative importance of early-life precursors for The protocol was preregistered on PROSPERO
understanding the neurodevelopmental pathways that lead (CRD42020165286).
to ADHD, particularly in terms of which aspects of neuro-
cognitive function are most predictive of the emergence Search Strategy and Selection Criteria
of the disorder and have potential clinical utility as early A systematic literature review was conducted to identify eli-
intervention or prevention targets.11 Furthermore, it is gible studies using 4 electronic bibliographic databases
important to understand which early-life developmental (MEDLINE, PsycINFO, EuropePMC, Scopus) and refer-
precursors are modifiable and the characteristics of interven- ence lists of eligible articles and review articles. The searches
tions that have such an effect. included full peer-reviewed journal articles accepted for
There has been no comprehensive and systematic analy- publication from the beginning of time until the final search
sis of the literature on early-life (0 5 years) precursors and date (see Supplement 2, available online). Articles published
interventions for ADHD. This is an emerging field, with no in languages other than those understood by the research
established knowledge base. A comprehensive account of team were translated using online software. For the first sys-
progress in the field to date is fundamental to organize exist- tematic review, studies were included if they met 1 of the
ing findings and to guide future work. The aims of this following inclusion criteria: empirical studies that (1) com-
study were therefore to investigate (1) which early neuro- pared neurocognition or behavior between infants with
cognitive and behavioral precursors are most robustly familial likelihood for ADHD (due to having a first-degree
associated with the onset and course of ADHD, and (2) relative with ADHD) to infants without familial likelihood
whether the precursors identified are currently targeted in for ADHD; (2) compared neurocognition/behavior
early interventions. between children aged 0 to 5 years with an early ADHD
To address each of these aims, we conducted 2 system- diagnosis or elevated ADHD symptoms to children without
atic reviews and meta-analyses of the following: (1) empiri- ADHD symptoms; and (3) examined cross-sectional or lon-
cal studies examining early-life (0 5 years) neurocognitive gitudinal associations between neurocognition/behavior
and behavioral precursors associated with familial likelihood measured at 0 to 5 years and concurrent or later-life ADHD
for ADHD, an early ADHD diagnosis or elevated ADHD symptoms. For the second systematic review, studies were
symptoms, or the presence and severity of later-life ADHD included if they met 1 of the following inclusion criteria:
symptoms; and (2) empirical studies of interventions target- reported effects of an intervention targeting 1 or more of
ing the neurocognitive and behavioral precursors identified the neurocognitive or behavioral precursors identified in the
in the first meta-analysis, delivered to infants with familial first review in (1) infants with familial likelihood for
likelihood for ADHD or preschoolers with ADHD symp- ADHD, (2) children aged 0 to 5 years with an early ADHD
toms. We discuss the findings in relation to early interven- diagnosis or elevated ADHD symptoms, or (3) ADHD
tion and prevention methods for ADHD, focusing on traits measured dimensionally in children aged 0 to 5 years.
whether currently available interventions target the most Case studies and review articles were excluded.
important early-life neurocognitive and behavioral precur-
sors associated with the disorder and are efficacious, and Data Selection, Extraction, and Coding
suggest future directions for research in this field. Titles and abstracts of retrieved studies were screened inde-
An additional aim of this work was to investigate when pendently by the first and second authors (ES and PFZ) to
in the early developmental period (infancy, toddlerhood, or identify those that potentially met inclusion criteria.
2 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

Subsequently, 2 authors (the first author ES and 1 among was assessed using funnel plots and the rank correlation test
PFZ, II, LFC, PBGG, MG, CT, AG, ES, PS, or CF) inde- for asymmetry in metafor.28 Finally, subgroup analyses were
pendently assessed the full text of potentially eligible studies conducted to test for moderating effects of developmental
and extracted data from included studies using standardized stage (infancy/toddlerhood vs preschool), study population,
forms. Extracted information included the following: study analysis type, whether effect size was or was not converted,
population/design; participant characteristics; ADHD and and study quality rating on the pooled effect size estimates
neurocognitive/behavioral outcome measures; intervention following significant findings (see Supplement 4, available
details; summary statistics; statistical model results and online, for full details). All R data and code are available at:
effect sizes (for full details, see Supplement 3, available https://osf.io/rwb3c/.
online). Data not available from publications were requested To assess whether early interventions target early neu-
from corresponding authors. Study quality and risk of bias rocognitive and behavioral precursors associated with
were rated independently during data extraction using the ADHD (aim 2), a narrative synthesis of findings was cre-
Newcastle Ottawa Scale (NOS)23 for observational stud- ated, grouped by ADHD and neurocognitive/behavioral
ies, the Cochrane Risk of Bias tool (RoB-2)24 for random- outcome measures for randomized controlled trials (RCTs)
ized controlled trials, and the Cochrane ROBINS-I25 for and nonrandomized intervention studies separately. Where
nonrandomized intervention studies. Disagreements during possible (for outcome measures reported in ≥4 studies),
screening and data extraction were resolved through discus- meta-analyses were conducted in R26 to estimate the
sion, mediated by the first author (ES). pooled effect size across studies. The difference between
pre post treatment scores in the treatment group(s) com-
Data Synthesis and Analysis pared to the control group were calculated and used as
To address the question of which early-life neurocognitive effect sizes (standardized mean difference [SMD]), repre-
and behavioral precursors are associated with ADHD (aim senting intervention-related improvements in outcome
1), a narrative synthesis of the findings was created, with measures (see Supplement 4, available online, for full
studies grouped by neurocognitive/behavioral function. details). For studies reporting outcome measures at multi-
Meta-analyses were conducted in R 4.0.226 to estimate the ple postintervention time-points, we selected those closest
pooled effect size across studies for each neurocognitive/ to the end of the intervention. For studies reporting more
behavioral function. The standardized mean difference, than 1 ADHD symptom outcome measure, the measure
Hedges’ g, was calculated for all studies and used as the with the least risk of bias (ie, completed by blinded inform-
measure of effect size in meta-analyses. For studies that ants or those most distant from the intervention context,
reported findings from analyses that were not based on eg, teacher ratings for a parent-mediated intervention) was
mean differences between groups (eg, regression), effect selected. When the degree of blinding appeared to be equal
sizes were converted to Hedges’ g using the R package esc.27 between measures, we calculated a composite score (mean
For studies that reported analyses of the same neurocogni- of effect sizes across measures). Outcome measures reflect-
tive/behavioral measure at multiple time-points in the 0- to ing total ADHD symptoms rather than inattention or
5-year period in the same participants, effect sizes were aver- hyperactivity/impulsivity dimensions were selected or
aged across time-points. For studies that reported measures computed by averaging dimension scores wherever possi-
of ADHD at multiple time-points, the analysis in relation ble. For studies that examined multiple measures of the
to ADHD symptoms at the latest developmental time-point same neurocognitive function (eg, 3 inhibition measures),
was selected. If studies reported multiple ADHD measures effect sizes were computed for each measure separately.
(eg, parent and teacher reports), effect sizes were averaged Multi-level random-effects meta-analytic models were then
across measures. For studies that examined multiple meas- fitted in metafor,28 with effect sizes nested within studies
ures of the same neurocognitive construct (eg, 3 memory for those that reported multiple effect sizes for the same
tasks), effect sizes were computed for each measure sepa- outcome measure. The REML estimator with the
rately. Multi-level random-effects meta-analytic models Knapp Hartung confidence interval adjustment were
were then fitted in metafor,28 with effect sizes nested within used.29 Heterogeneity was assessed with Q and I2 using
studies for those that reported multiple effect sizes for the the dmetar package.30 Publication bias was assessed using
same neurocognitive function to account for nonindepen- funnel plots and the rank correlation test for asymmetry in
dence of data. The Restricted Maximum-Likelihood metafor.28 Subgroup analyses were conducted to test for
(REML) estimator was used with the Knapp Hartung con- moderating effects on pooled estimates following signifi-
fidence interval adjustment.29 Heterogeneity was assessed cant findings (see Supplement 4, available online). R data
with Q and I2 using the dmetar package.30 Publication bias and code are available at: https://osf.io/rwb3c/.
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 3
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

RESULTS without elevated ADHD symptoms or an early ADHD


Aim 1: What are the early-life neurocognitive and behav- diagnosis (k = 40, 26.9%), general population/community
ioral precursors of ADHD? samples (k = 63, 42.3%), and other populations (eg, preterm
infants, infants with familial likelihood for autism) (k = 37,
SAMPLE 24.8%) (Table 1). Approximately half of the studies focused
After full-text reading, 149 non-duplicate studies (16,5095 on preschool-aged (3 5 years) samples (k = 75, 50.3%),
participants) met inclusion criteria (Figure 1, Table 113−17,31 whereas 55 (36.9%) studied infant/toddler (0 2 years)
−173 samples, and a further 19 studies (12.8%) included samples
). Eight domains of neurocognitive/behavioral function
were examined (General cognitive, language and motor abili- of both infants/toddlers and preschoolers (0 5 years). In
ties; Social and emotional functioning; Sleep; Early regulatory terms of socioeconomic diversity of the samples, 82 studies
problems (eg, excessive crying); Sensory processing; Activity (55.0%) did not report or provided insufficient information
level; Executive function; Brain structure and resting-state on socio-economic status (SES) to characterize the samples;
neurophysiological activity). Most studies (k = 97, 65.1%) a further 27 studies (18.1%) were limited to participants of
examined more than 1 of these functions. Populations stud- middle-to-high SES, and 40 studies (26.9%) included par-
ied were as follows: infants/preschoolers with familial likeli- ticipants of diverse SES backgrounds. Concerning race and
hood for ADHD (k = 9, 6.0%), preschoolers with versus ethnicity of the samples, 92 studies (61.7%) did not report

FIGURE 1 PRISMA Flowchart: Studies Investigating Early Neurocognitive Atypicalities in Attention-Deficit/Hyperactivity


Disorder (ADHD)

4 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry


Volume 00 / Number 00 / & 2021
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Summary of Studies Investigating Early Neurocognitive and Behavioral Atypicalities in Attention-Deficit/Hyperactivity Disorder (ADHD)
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Abulizi 201731 Cohort Population sample n = 1,184 »12.0 »65.0 47.1% 1 (low) − 6 (high) scale. Not reported SDQ Social and emotional Good
of infants Cohort mean = 4.26 functioning; Activity
(1.26) level
Agapitou 200832 Case-Control Community sample n = 20 ADHD, n = 20 »67.2 Same as 20% (Both groups) Not reported Not reported Unspecified General cognitive and Poor
of preschoolers TD neurocog. questionnaires language abilities
with elevated
ADHD symptoms
vs TD
Allely 201333 Cohort Population-based n = 53 ADHD, n = 106 »12.0 »91.0 ADHD: 28.3% Not reported Not reported DAWBA Social functioning Good
sample of infants TD TD: 31.3%
who later met Dx
criteria for ADHD
or not
Arnett 201234 Cohort Population samples n = 1,506 57.3 (47-70, Australian First, second, and 49% Parental education (y). % “Non-White”. ADHD factor Language ability Good
(Australia and US) sample), 58.8 (54-71, third grades Cohort Whole cohort: scores derived
of twins US sample) (»72.0 − 108.0) mean = 13.97 (2.13) 44%. No further from DBRS and
information SWAN
Arnett 201335 Cohort and Population sample n = 149 ADHD, n = 867 T1: »1.0 Third grade ADHD: 35.6%, no- Mean annual family % White. ADHD: DBRS General cognitive and Good
T2: »6.0

ARTICLE IN PRESS
case-control of infants who no-ADHD (»108.0) ADHD: 53.4% income. ADHD: 76.5%. No-ADHD: language abilities;
later had elevated T3: »15.0 $44,133 (33,205). 86.5%. No further Activity level; Sleep
ADHD symptoms T4: »24.0 No-ADHD: $57,567 information
or not T5: »36.0 (41,842)
Astbury 198736 Cohort + case- Community sample n = 23 ADHD, n = 34 T1: »12.0 »60.0 ADHD: 56.5%, no- Father’s occupation Not reported ADHD Dx General cognitive and Fair
control of preterm (very no-ADHD T2: »24.0 ADHD: 55.9% ADHD: 21.7% assigned by language abilities
low birth weight) T3: »60.0 professional/ pediatrician
infants who later managerial, 43.5% and
met Dx criteria for skilled worker, psychologist
ADHD or not 26.1% semi/un- (instruments
skilled worker/ not specified)
unemployed, 8.7%
not reported. No-
ADHD: 35.3%
professional/
managerial, 23.5%
skilled worker,
28.6% semi/un-
skilled worker/

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


unemployed, 12.6%
not reported
Auerbach 200437 Cohort + case- Community sample n = 31 EL, n = 21 TL 7.5 (1.27) Familial ADHD 0% (All-male Not reported Not reported Paternal scores on Activity level; Good
control of infants with assessed at cohort) the DSM-IV Executive function
elevated and birth ADHD Rating
typical familial Scale
likelihood for
www.jaacap.org

ADHD (EL vs TL)


Auerbach 200538 Cohort + case- Community sample n = 92 EL, n = 66 TL 0.0 − 2.0 Familial ADHD 0% (All-male Not reported Not reported Paternal scores on Activity level; Early Good
control of infants with assessed at cohort) the DSM-IV regulatory problems
elevated and birth ADHD Rating
typical familial Scale
likelihood for
ADHD (EL vs TL)
Auerbach 200813 Cohort + case- Community sample n = 36 EL, n = 22 TL T1: 7.34 (0.62) Familial ADHD 0% (All-male Not reported Not reported Paternal scores on Activity level; Good
control of infants with T2: 12.5 (0.74) assessed at cohort) the DSM-IV Executive function;
elevated and T3: 25.8 (1.31) birth ADHD Rating Social and
typical familial Scale emotional
likelihood for functioning; Sensory
ADHD (EL vs TL) processing

(continued)
5
6

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Bates 198539 Cohort Community sample n = 33 90 (ns varied T1: »6.0 »36.0 Not reported 14% Upper class, 70% Not reported PBQ Social and emotional Poor
of infants by measure) T2: »13.0 middle class, 16% functioning;
T3: »24.0 working class General cognitive
and language
abilities; Activity
level
www.jaacap.org

Becker 201040 Cohort Community sample n = 300 »3.0 »24.0 - »180.0 51.7% “Most children in Cohort described as Lifetime ADHD Early regulatory Good
of infants with and cohort from SES- “primarily Dx from problems
without obstetric disadvantaged Caucasian.” No Mannheim
and psychosocial families” further Parent
risk information Interview and
K-SADS
Bedford 201941 Cohort Community sample n = 76 »14.0 EL-ASD: 90.81 59.6% Not reported Not reported Conners Executive function Fair
of infants with (6.33)
elevated and TL-ASD: 89.34
typical familial (4.81)
likelihood for
autism (EL-ASD vs
TL-ASD)

ARTICLE IN PRESS
Begnoche 201642 Cohort Community sample n = 21 55.08 (1.56) »66.0 51.2% Annual family income; 87.5% Non-Hispanic MacArthur HBQ Executive function Poor
of preschoolers 47.5% of cohort > Caucasian, 5%
$60,000 African American,
5% Asian
American, 2.5%
Hispanic
Ben-Sasson Cohort + case- Community sample n = 38 ADHD, n = 814 T1: 18.26 (3.85) 97.0 (6.24) 52% (Whole “Sample SES similar to 69% of Whole cohort CBCL Sensory processing; Poor
201743 control of preschoolers no-ADHD T2: 30.0 (3.37) cohort) population census”; White. No further Executive function;
with and without T3: 41.14 (3.0) 90% of cohort with information Activity level
later elevated parental education
ADHD symptoms > high school.
Berdan 200844 Cohort Community sample n = 200 »54.0 »64.0 55.0% Hollingshead Index. 65% European BASC Social functioning Poor
of preschoolers Cohort mean: 43.96 American, 29%
Journal of the American Academy of Child & Adolescent Psychiatry

with and without (range: 17 65) African American,


externalizing 4% Biracial, 2%
problems Hispanic
Berger 201345 Cohort Community sample n = 60 (behavioral »60.0 Same as 0% (all male) Parental mean years of Not reported ADHD Rating Executive function Good
of infants with analysis) neurocog. education. Scale IV
elevated and n = 18 (EEG analysis) Maternal: 13.3 (1.9); (DuPaul)
typical familial paternal: 13.0 (2.4)
likelihood for
ADHD (EL and TL)
Bora 201446 Cohort + case- Community sample n = 13 ADHD, n = 67 »0.0 (Term equivalent »108.0 49.5% Not reported Not reported SDQ; DSM-IV Brain structure Good
control of preterm infants no-ADHD birth age) ADHD
who later met Dx Diagnostic
criteria for ADHD Interview
Volume 00 / Number 00 / & 2021

or not
Breaux 201647 Cohort + case- Community sample n = 75 ADHD, n = 93 T1: 44.36 (3.32) 80.80 (5.12) 45.8% Median annual family Of whole cohort, DISC Language ability; Good
control of preschoolers no-ADHD T2: 56.81 (3.66) income; cohort 53.6% European Executive function
with externalizing median: $47,108 American, 22.6%
problems who Latino, 13.7%
later met Dx Multi-ethnic
criteria for ADHD
or not
Bron 201248 Cohort Clinical sample of n = 44 25.6 (11.6) 110.4 (28.8) 36.4% Not reported (but was Not reported CBCL Early regulatory Fair
infants with used in analysis) problems
regulation
disorder

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Bundgaard 201849 Cohort + case- Preschoolers with n = 24 ADHD, n = 25 ADHD: 37.96 (5.35), ADHD: 28.13 ADHD: 20.8%, SES based on Not reported CBCL; ADHD Sleep Good
control and without no-ADHD no-ADHD: 38.56 (5.52) (4.48), no-ADHD: 20.0% maternal education. Rating Scale IV
elevated ADHD no-ADHD: 28.96 High SES: high
symptoms (5.49) school+4 y;
selected from intermediate SES:
population high school+1-3 y,
sample Low SES: high
school or less.
ADHD: 25% high,
29.2% intermediate,
33.3% low, 12.5%
not reported. No-
ADHD: 25% high,
32% intermediate,
25% low, 18% not
reported
Butcher 200950 Cohort Community sample n = 65 3.06 (0.25) 105.6 (14.4) Not reported Parental education Not reported CBCL Motor ability Poor
of preterm infants level (y). Maternal:
11.5 (3.5); paternal:

ARTICLE IN PRESS
12.1 (2.5)
Campbell 198651 Cohort + case- Community sample n = 47 (Full cohort), »35.0 »73.0 48.9% ADHD: “mainly Not reported SNAP; CBCL Activity level; Good
control of preschoolers n = 10 ADHD, n = 37 Working-Middle Executive function
with and without no-ADHD class”, no-ADHD:
externalizing “mainly Middle/
problems with Professional class”
and without later
ADHD symptoms
Carlson 199552 Cohort Community sample n = 170 T1: »0.23 »132.0 45.9% Household 80% Caucasian, 14% CBCL Motor ability; Good
of low-SES infants T2: »6.0 occupational African American, Executive function
category. 25% 6% Native or
Professional/ Hispanic
administrative/
managerial, 26%
sales/clerical, 13%
craftsperson, 5%
transport/ laborer,
13% service worker,

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


15% student/
unemployed
Caspi 199553 Cohort Population sample n = 976 T1: »36.0 »180.0 Not reported “Representative of the Not reported RBPCL Executive function Good
of infants T2: »60.0 population”
Cheung 201854 Cohort Community sample n = 113 EL-ASD, T1: »9.0 »36.0 45.0% Not reported Not reported CBCL Executive function Fair
of infants with n = 27 TL-ASD T2: »15.0
elevated and
typical familial
www.jaacap.org

likelihood for
autism (EL-ASD vs
TL-ASD)
Cunningham Case-control Preschoolers with n = 24 ADHD, ADHD: 57.4 (3.2), Same as ADHD: 54.2%, Not reported Not reported DBDS; CBCL General cognitive and Good
200255 and without n = 35 TD TD: 57.7 (6.5) neurocog. TD: 55.9% language abilities;
elevated ADHD Social functioning
symptoms
de la Osa 201356 Cohort Community sample n = 622 35.64 (1.92) Same as 50.2% 33% High SES, 31.4% 88.9% White, 7.9% SDQ Social and emotional Good
of preschoolers neurocog. medium-high SES, Hispanic, 0.3% functioning;
with and without 14.1% medium SES, African American, Executive function
elevated ADHD 15.9% medium-low 1% Asian, 1.9%
symptoms SES, 5.6% low SES Other

(continued)
7
8

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Delobel-Ayoub Cohort + case- Preterm children n = 175 ADHD, n = 803 »60.0 Same as Not reported Not reported Not reported SDQ General cognitive Good
200957 control with and without no-ADHD neurocog. ability
elevated ADHD
symptoms from
population-based
cohort
www.jaacap.org

DeSantis 200458 Cohort Clinical sample of n = 28 1.0−4.0 36.0−96.0 46.4% Not reported Not reported CBCL Early regulatory Good
infants with problems
excessive crying/
colic
DeWolfe 200059 Case-control Clinical sample of n = 25 ADHD, n = 25 ADHD: 57.84 (8.28), Same as 16.0% (in each Hollingshead Index. Not reported Clinical Dx Activity level; Good
preschoolers with TD TD: 58.32 (8.04) neurocog. group) ADHD mean: 34.4 according to Executive function
ADHD Dx and TD (12.51), no-ADHD DSM-IV;
preschoolers mean: 42.48 (17.82) Conners; Dx
Interview
Doherty 201860 Cohort Community sample n = 106 EL-ASD, EL-ASD: 38.84 (1.57), Same as 43.0% Not reported Not reported CBCL Executive function Fair
of infants with n = 25 TL-ASD TL-ASD: 38.52 (2.43) neurocog.
elevated and
typical familial

ARTICLE IN PRESS
likelihood for
autism (EL-ASD vs
TL-ASD)
Dougherty 201161 Cohort + case- Community sample n = 11 ADHD, n = 530 43.2 (3.6) Same as 45.7% Parental education 86.9% White, 9.1% PAPA Social and emotional Good
control of preschoolers no-ADHD neurocog. level. 34.9% College Hispanic (no functioning;
with and without degree (1 parent), further info) Executive function
research Dx of 33.9% college
ADHD degree (both
parents)
DuPaul 200162 Case-control Community sample n = 58 ADHD, n = 36 ADHD: 47.76 (8.28), Same as ADHD: 13.8%, TD: “Primarily from middle ADHD: 79% White, ADHD Rating Social and emotional Good
of preschoolers TD TD: 49.68 (8.76) neurocog. 44.4% class background” 7% African Scale IV functioning
with and without American, 12% (DuPaul);
research Dx of Latino, 2% Other. Conners,
Journal of the American Academy of Child & Adolescent Psychiatry

ADHD No-ADHD: 97% SIDAC


White, 3% African
American
Einziger 201814 Cohort Community sample n = 64 T1: »36.0 162.72 (11.16) 0.0% (All male) Parental years of Not reported Conners Activity level; Poor
of infants with T2: »54.0 education. Maternal Emotional
elevated and mean: 13.08 (1.83); functioning;
typical familial paternal: 12.91 (1.62) Executive function
likelihood for
ADHD (EL and TL)
Elberling 201563 Cohort Population sample n = 1489 0.0−10.0 73.56 (5.40) 50.0% Not reported Not reported DAWBA Activity level; Good
of infants Emotional
functioning; Motor
ability
Volume 00 / Number 00 / & 2021

Ezpeleta 201564 Case-control Community sample n = 23 ADHD, n = 538 ADHD: 44.88 (3.96), Same as ADHD: 26.1%, ADHD: 8.7% High SES, ADHD: 91.3% DICA; SDQ Executive function; Good
of preschoolers no-ADHD no-ADHD: 45.12 (3.96) neurocog. no-ADHD: 51.7% 43.5% middle-high Caucasian, 8.7% Emotional
with and without SES, 47.8% middle- Hispanic. No- functioning
research Dx of low/low SES. No- ADHD: 89.6%
ADHD ADHD: 34.2% high Caucasian, 5.8%
SES, 46.4% middle- Hispanic, 0.9%
high SES, 19.5% Asian, 3.7% Other
middle-low/low SES
Floyd 200165 Cohort Community sample n = 34 »55.7 Same as 42.9% Not reported Not reported BASC Executive function Poor
of preschoolers neurocog.
Foulon 201566 Cohort Population sample n = 1,131 T1: »4.0 »36.0 Not reported Not reported Not reported SDQ Executive function; Poor
of infants T2: »24.0 Motor ability; Early
regulatory
problems; Sleep

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Frenkel 201767 Cohort Community sample n = 174 49.0−55.0 62.18 (2.15) 49.4% Median household 16.7% African/ Black, MacArthur HBQ Resting-state Fair
of abandoned / income range: 7.5% American / neurophysiological
institutionalized $75,000 $125,000 Alaska Native, activity
or not children 38.5% Asian, 31%
White, 3.4%
Multiracial, 3.4%
Missing
Frick 2019a68 Cohort Population-based n = 112 10.04 (0.24) 36.40 (1.18) 47.3% Cohort mean SES: 4.84 Not reported ADHD Rating Executive function; Good
sample of (0.51). Measure not Scale IV Social and
preschoolers stated. 86% of (DuPaul) emotional
Mothers and 80% of functioning
fathers had
university degree
Frick 2019b69 Cohort Population-based n = 47 T1: 12.47 (0.43) 74.78 (1.18) 57.0% 77.7% of Mothers and Not reported ADHD Rating Activity level; Good
sample of T2: 18.51 (0.73) 66.1% of fathers had Scale IV Executive function
preschoolers T3: 24.60 (0.61) university degree (DuPaul)
T4: 36.97 (1.08)
Friedman 200570 Cohort Community sample n = 13 (T1 T3) T1: »1.0 98.40 (8.0) 61.5% Not reported Not reported CBCL Activity level; Motor Poor
of infants n = 13 (T2 T3) T2: »3.0 ability; Executive

ARTICLE IN PRESS
function
Gagne 201171 Cohort Community sample n = 590 24.84 (0.05) Same age as 46.7% Hollingshead Index. 88.2% White, 3.1% CBCL Executive function Good
of twins neurocog. Cohort mean: 51.5 Black, 2.1% Asian,
(10.4) 6.6% Multiracial
Gagne 201972 Cohort Community sample n = 198 46.56 (12.48) Same age as 48.5% Average household 84% White, 11% CBCL Executive functioning Good
of siblings neurocog. income: »$70,000 Multiracial, 4%
(range $20 200k). African American,
Mean maternal 13% Hispanic or
years of education: Latino, 3% Other
15.82; mean
paternal years of
education: 15.12
Gaspardo 201873 Cohort Community sample n = 62 24.0 (5.0) Same as 51.0% 38% Class B, 59% class Not reported CBCL Social and emotional Good
of preterm infants neurocog. C, 3% class D. Mean functioning;
maternal years of Executive function
education: 10 (3)
Girouard 199874 Cohort Community sample n = 76 18.06 (0.52) 69.96 (SD = 10.8) 44.7% “Mainly low-middle 87.1% White (no DISC General cognitive and Good
of preterm infants class”. Mean = 45.71 further info) language ability

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


(15.10). Maternal
education (y); 12.56
(2.96); paternal
education (y): 13.19
(3.16)/
Gliga 201875 Cohort + case- Community sample n = 20 EL-ASD T1: »8.0 T3: »36.0 EL-ASD+ADHD: Not reported Not reported SDQ in older Executive function Fair
control of infants with +ADHD, T2: »14.0 50.0%, siblings with
elevated and n = 26 EL-ASD-ADHD, EL-ASD-ADHD: autism
www.jaacap.org

typical familial n = 54 TL-ASD 61.5%,


likelihood for TL-ASD: 53.7%
autism, with and
without elevated
familial likelihood
for ADHD (EL-
ASD+ADHD vs
EL-ASD-ADHD vs
TL-ASD)
Gould 201976 Cohort Community sample n = 577 »18.0 »84.0 46.0% Maternal education. 46% White (no Conners General cognitive and Good
of preterm infants 60.3% completed further info) motor ability
secondary
education, 49.7%
completed tertiary
education
9

(continued)
10

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Guedeney 201477 Cohort Population sample n = 1,250 11.8 (1.0) 36.0−60.0 59.9% Family income during Not reported SDQ Social functioning Good
of infants pregnancy (euros/
mo): 13.5% <1,500,
30.7% 1,501 2,300,
27.8% 2,301 3,000,
28% >3,000.
www.jaacap.org

Maternal education
at birth: 25.7%
<high school, 18.5%
high school, 22.5%
high school+2 y,
33.3% university
Gunn 200978 Cohort Community sample n = 179 51.60 (6.52) Same age as 46.0% 71% of Families had Not reported CBCL Sensory processing Good
of preschoolers neurocog. annual incomes <
$25,000; 2% had
incomes >$75,000
Gurevitz 201479 Retro-spective Clinical sample of n = 58 ADHD, n = 58 Retrospective data ADHD: 98.04 ADHD: 31.0%, “High-average SES”. Not reported Community Language and motor Good
case-control children Dx with TD collected from: (17.16), TD: 34.5% ADHD: 25.9% of clinical Dx of abilities; Early
ADHD and T1: »0.0 − 1.0 TD: 93.24 (15.72) mothers and 26.9% ADHD regulatory

ARTICLE IN PRESS
community T2: »3.0 of fathers had recorded in problems; Sleep
sample of TD T3: »9.0 academic medical (from medical
children T4: »18.0 profession. TD: records records)
54.5% of parents
had academic
profession
Gusdorf 201180 Cohort Community sample n = 89 »36.0 Same as 49.4% Parental education. Not reported SDQ Executive function; Good
of preschoolers neurocog. 23.9% of mothers Motor ability
and 30.7% of fathers
university, 32.6% of
mothers and 36% of
fathers college
Handen 200781 Case-control Community sample n = 52 ID+ADHD, ID+ADHD: 61.65 Same as ID+ADHD: 26.9%, Hollingshead Index ID+ADHD: 67.3% Conners Activity level; Good
Journal of the American Academy of Child & Adolescent Psychiatry

of preschoolers n = 21 ID (7.50), ID: 59.0 (6.68) neurocog. ID: 57.1% Class. ADHD: 7.7% White, Non- Executive function
with ID with and level 1 (highest), 32.7% Latino, 32.7%
without elevated level 2, 11.5% level Other. ID: 81%
ADHD symptoms 3, 23.1% level 4, 25% White, Non-
level 5. TD: 19% Hispanic, 19%
level 1, 38.1% level Other
2, 19% level 3, 11.5%
level 4, 14.3% level 5
Hatch 201482 Cohort Community sample n = 197 51.84 (5.27) Same as 25.9% Nakao-Treas SE 40.1% White, 19.3% ADHD Rating Executive function Good
of preschoolers neurocog. Prestige Index White Hispanic, Scale IV
with and without whole cohort mean: 10.7% Black, 1.5% (DuPaul)
ADHD symptoms 54.90 (16.77) Black Hispanic,
10.2% Asian,
Volume 00 / Number 00 / & 2021

18.3% Multiracial
or Other
Healey 201183 Cohort Community sample n = 216 51.13 (5.69) Same as 27.3% female Nakao-Treas SE 40.2% White, 18.2% ADHD Rating Emotional functioning; Good
of preschoolers neurocog. Prestige Index White Hispanic, Scale IV Executive function
with and without whole cohort mean: 10.8% Black, 1.3% (DuPaul)
elevated ADHD 55.47 (15.14) Black Hispanic,
symptoms 10.8% Asian,
18.7% Other
Houwen 201784 Cohort Community sample n = 153 49.20 (9.60) Same as 51.0% Maternal education Not reported SDQ Motor ability; Good
of preschoolers neurocog. level: low (primary Executive function;
and lower Emotional
secondary school), functioning
intermediate (upper
secondary school or

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
vocational school),
high (higher
vocational training
or university). 5%
Low SES, 37%
intermediate SES,
58% high SES
Hutchison 201785 Cohort Community sample n = 50 2.32 (1.0) 40.9 (0.7) 44.0% Not reported 36% Caucasian, 26% CBCL Sensory processing Good
of infants Hispanic, 18%
Other
Hwang-Gu 201986 Case-control Clinical sample of n = 65 ADHD, ADHD: 56.52 (6.24), Same as ADHD: 24.6%, Parental education. Not reported SNAP-IV and Executive function Good
preschoolers with n = 98 TD TD: 57.72 (7.80) neurocog. TD: 51.0% ADHD: 76.9% clinician-
elevated ADHD mothers and 67.7% assigned
symptoms and fathers college, 20% suspected
community mothers and 27.4% ADHD Dx
sample of TD fathers senior high,
preschoolers 3.1% mothers and
4.8% fathers junior
high. TD: 75.5%

ARTICLE IN PRESS
mothers and 76.3%
fathers college,
19.4% mothers and
20.6% fathers senior
high, 5.1% mothers
and 3.1% fathers
junior high
Ilot 201087 Cohort Community sample n = 622 24.84 (0.6) Same as Not reported Hollingshead Index. 85.4% Caucasian, CBCL; Revised Activity level Good
of twins neurocog. Cohort mean: 50.9 3.2% Black, 2% Rutter Scale
(14.1) Asian, 7.3%
Multiracial, 2.2%
Other
Jacobson 2018a88 Case-control Community sample n = 52 ADHD, n = 38 ADHD: 60.12 (6.96), Same as ADHD: 34.6%, TD: No group differences Whole sample: DISC/DICA-IV; Brain structure Good
of preschoolers TD TD: 59.04 (6.96) neurocog, 36.8% in Hollingshead 83.3% Caucasian, Conners
with research Dx Index, but data not 11.1% African
of ADHD or TD reported American, 3.3%
Asian, 1.1%
Multiracial

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


Jacobson 2018b89 Case-control Community sample n = 54 ADHD, n = 51 ADHD: 60.6 (6.96), Same as ADHD: 40.7%, Not reported Whole sample: 84% DISC/DICA-IV; General cognitive, Good
of preschoolers TD TD: 59.04 (6.72) neurocog. TD: 37.3% Caucasian, 11% Conners language and motor
with research Dx African American, abilities; Executive
of ADHD or TD 3% Asian, 1% function
Multiracial, 1%
Other
Jacobvitz 198790 Cohort + case- Community sample n = 34 ADHD, n = 34 T1: »0.25 »72.0 ADHD: 29.4%, no- All families of low SES 80% of mothers CBCL Motor ability; Social Fair
control of low-SES infants no-ADHD T2: »3.0 ADHD: 29.4% background. SES White (no info on and emotional
www.jaacap.org

with and without T3: »6.0 measured but not infants or fathers) functioning; Early
later elevated T4: »30.0 reported regulatory
ADHD symptoms T5: »42.0 problems; Activity
level; Executive
function
Jaspers 201391 Cohort with Population-based n = 419 ADHD, »0.0−48.0 »132.0− »193.0 ADHD: 39.6%, Not reported Not reported CBCL Activity level; Motor Good
retro- sample of n = 1,245 no-ADHD no-ADHD: 56.4% and language
spective + adolescents with abilities; Sleep;
case- and without Social and
control research Dx of emotional
aspects ADHD, with data functioning
retrospectively
collected from
early childhood

(continued)
11
12

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Johnson 201492 Cohort + case- Population-based n = 16 ADHD, n = 120 12.43 (0.25) »91.0 ADHD: 12.5%, Not reported Not reported DAWBA Activity level Good
control sample of infants no-ADHD no-ADHD: 31.7%
comparison who later met Dx
criteria for ADHD
or not
Johnson 201693 Cohort Population-based n = 371 »24.0 130.0 (4.3) PT: 55.0%, Parental occupational Not reported ADHD Rating General cognitive and Fair
www.jaacap.org

sample of (n = 219 PT, n = 153 FT) FT: 58.0% status (1 = high to Scale IV motor abilities
preterm (PT) and 4 = lowest). 43% (DuPaul)
full-term (FT) Level 1, 24% level 2,
infants 27% level 3, 6% level
4.
Kidwell 201794 Cohort Community sample n = 271 T1: »36.0 121.20 (4.32) 49.4% 46% SES 73.4% European Conners Sleep; Executive Good
of preschoolers T2: »54.0 disadvantaged American, 7.38% function
(below Federal Hispanic/Latino,
Poverty Line); 53.5% 5.17% African
not SES American, 14.02%
disadvantaged Multiracial
Landau 2010a95 Cohort + case- Community sample n = 47 EL, EL: 6.88 (0.63), Familial ADHD 0.0% (All male) Parental y of Not reported Paternal scores on Social and emotional Good
control of infants with n = 31 TL TL: 7.03 (0.57) assessed at education. EL: the DSM-IV functioning

ARTICLE IN PRESS
elevated and birth maternal 12.13 ADHD Rating
typical familial (1.88), paternal 12.24 Scale
likelihood for (1.90); TL: maternal
ADHD (EL vs TL) 13.0 (1.84), paternal
13.0 (1.82)
Landau 2010b96 Cohort + case- Community sample n = 26 EL, EL: 1.54 (0.56), Familial ADHD 0.0% (All male) Parental y of Not reported Paternal scores on Sleep Good
control of infants with n = 18 TL TL: 1.42 (0.42) assessed at education. EL: the DSM-IV
elevated and birth maternal 13.29 ADHD Rating
typical familial (2.13), paternal 12.42 Scale
likelihood for (1.84); TL: maternal
ADHD (EL vs TL) 13.61 (2.35), paternal
13.33 (2.40)
Landis 202097 Cohort Community sample n = 249 59.40 (6.36) Same as 22.0% Hollingshead Index 82% Hispanic/Latino BASC-2 Executive function; Good
Journal of the American Academy of Child & Adolescent Psychiatry

of preschoolers neurocog. cohort mean: 43.63 (no further info) Emotional


with elevated (12.63) functioning
externalizing
problems
Lavigne 201698 Cohort Community sample n = 344 53.04 (3.96) Same as 48.3% Hollingshead Index. 63.1% White Non- ECI Emotional functioning; Good
of preschoolers neurocog. 81.1% Classes 1 and Hispanic, 13.4% Executive function;
2, 18.9% classes 3 African American, Language ability;
and 4 15.7% Hispanic, Sensory processing
2.3% Asian, 4.1%
Multiracial, 1.4%
not reported
Lawson 2004a99 Cohort Community sample n = 75 T1: 12.14 (0.31) 42.27 (0.46) 42.8% Maternal mean years Not reported Conners Executive function; Fair
of infants T2: 24.13 (0.40) of education: 13.5 Emotional
Volume 00 / Number 00 / & 2021

(2.5) functioning
Lawson 2004b100 Cohort Community sample n = 45 T1: 7.6 (0.3) 61.0 (2.0) 56.4% Hollingshead Index Not reported Conners; CBCL Executive function Fair
of preterm infants T2: 24.1 (0.8) mean: 31.3 (11.6)
T3: 36.1 (0.9)
T4: 48.4 (0.8)
Lemcke 2016101 Cohort + case- Population sample n = 2034 ADHD, T1: »6.0 ADHD: 136.0 ADHD: 20.1%, Maternal education Not reported Community ICD- Motor and language Good
control of infants who n = 76286 no-ADHD ADHD: 6.4 (0.89), (15.6), no-ADHD: 48.8% level (high, medium, 10 ADHD Dx in abilities; Activity
later met Dx no-ADHD: 6.4 (0.80) no-ADHD: 135.0 low). Whole cohort: medical levels; Early
criteria or not (16.2) 53.4% high, 38% records regulatory problems
medium, 8.6% low
Levine 2018102 Cohort Community sample n = 190 »24.0 »54.0 MT: 43.3%, % Low SES: 92.5% MT, % European SDQ Executive function Poor
of infants (n = 87 MT, n = 103 no- No-MT: 54.5% 25% No-MT. % ethnicity: 75% MT,
prenatally MT) Mothers without 78.4% No-MT
exposed to formal education

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
methadone or not qualification: 80.6%
(MT and no-MT) MT, 19.3% No-MT
Lyons-Ruth Cohort Community sample n = 62 T1: »12.0 59.0 (49 71) 40.3% “Mainly low-SES 80% Caucasian (no PBQ Social functioning Poor
2009103 of low-SES infants T2: »18.0 families” further info)
Mahone 2007104 Case-control Community sample n = 25 ADHD, n = 25 ADHD: 58.3 (10.0), Same as 20.0% (Both Maternal y of ADHD: 60% Community Executive function; Poor
of preschoolers TD TD: 58.2 (10.0) neurocog. groups) education. ADHD: Caucasian, 36% Clinical DSM-IV Emotional
with clinical 14.3 (2.4); TD: 15.3 African American, ADHD Dx; functioning
ADHD Dx and TD (2.5) 4% Biracial. TD: Conners
preschoolers 72% Caucasian,
28% African
American
Mahone 2011105 Case-control Community sample n = 11 ADHD, ADHD: 60.0 (6.0), Same as ADHD: 38.0%, Not reported Whole sample: 85% YC-DISC; Conners Brain structure Good
of preschoolers n = 12 TD TD: 60.0 (7.2) neurocog. TD: 62.0% Caucasian, 8%
with research Dx African American,
of ADHD and TD 4% Biracial, 3%
preschoolers Asian
Mariani 1997106 Case-control Clinical sample of n = 34 ADHD, n = 30 ADHD: 60.1 (7.5), Same as 0.0% (All boys) Hollingshead Index. 100% White Clinical interview General cognitive, Good
preschoolers with TD TD: 61.0 (6.3) neurocog. ADHD: 62.4 (23.1). for DSM-III-TR motor and language
clinical ADHD Dx TD: 69.7 (23.4) ADHD; abilities; Executive

ARTICLE IN PRESS
and community Conners function
sample of TD
preschoolers
Marks 2005107 Case-control Community sample n = 22 ADHD, n = 50 ADHD: 53.52 (5.52), Same as ADHD: 18.2%, Family income: 47% of Full sample: 35% ADHD Rating Activity level; Good
of preschoolers no-ADHD no-ADHD: 50.76 (8.28) neurocog. no-ADHD: 54.0% sample earning > Caucasian, 3% Scale IV Executive function
with and without $70,000 and 90% of African American, (DuPaul)
elevated ADHD sample earning > 21% Latino, 24%
symptoms $25,000 per year; Asian American,
44% of fathers and 18% Other or
48% of mothers had Multiracial
undergraduate
degree
Mathieson Cohort + case- Population sample n = 19 ADHD, T1: »18.0 Same as 40.0% (Whole Maternal education. Not reported Behavioral Activity level; Social Good
2000108 control of toddlers with n = 671 no-ADHD T2: 30.0 neurocog. cohort) 9% <9 y, 16% Checklist and emotional
and without college/university functioning
ADHD symptoms degree
McGee 1991109 Cohort + case- Population sample n = 21 ADHD, T1: »36.0 Same as ADHD: 38.1%, Family Adversity Not reported Behavior Profile General cognitive, Good
control of children with n = 977 no-ADHD T2: »60.0 neurocog. no-ADHD: 49.0% Index. 24% ADHD -Attention language and motor

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


and without and 9% no-ADHD abilities
elevated ADHD scored >3
symptoms (indicative of
adversity)
McLaughlin Cohort Community sample n = 166 9.0−30.0 »54.0 Not reported Not reported Not reported PAPA Resting-state Good
2010110 of abandoned / neurophysiological
institutionalized activity
infants and
www.jaacap.org

nonabandoned
community
controls
Meeuwsen Cohort Community sample n = 321 T1: »6.0 »84.0 43.0% Sociodemographic Not reported CBCL; Conners Activity level; Good
2019111 of infants T2: »12.0 Adversity Index Executive function
T3: »21.0 used in analysis but
T4: 36.0 means/range not
reported
Melegari 2015112 Case-control Community sample n = 30 ADHD, ADHD: 52.7 (12.0), Same age as ADHD: 10.0%, Not reported Not reported CBCL; PAPA Social functioning; Good
of preschoolers n = 30 TD TD: 57.3 (7.5) neurocog. TD: 20.0% girls Executive function
with research Dx
of ADHD or TD

(continued)
13
14

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Mendez 2002113 Cohort Community sample T1: n = 113, T1: 54.8 (7.0) Same age as T1: 55.7%, Parental education 87.9% African Conners Social functioning; Fair
of low-income T2: n = 34 T2: »62.0 neurocog. T2: not reported level. 24% Some American, 12.1% Language ability
preschoolers high school, 28% not reported
high school
graduate, 35% some
college, 10%
www.jaacap.org

college degree, 3%
vocational school.
All low income
Mikoteit 2018114 Cohort Community sample n = 19 T1: 64.08 (4.08) T2: 75.12 (3.12), 47.4% Not reported Not reported SDQ Sleep Good
of preschoolers T3: 109.80 (1.56)
115
Miller 2013 Cohort Community sample n = 180 T1: 51.60 (1.2) T1: same as 23.9% (elevated Nakao Treas SE 40.2% Non-Hispanic ADHD Rating Executive function Good
of preschoolers neurocog. symptoms Prestige Index Caucasian, 10.7% Scale IV
with and without T2: »62.0 group), mean: 55.47 (15.14) African American, (DuPaul)
elevated ADHD 32.9% no-ADHD 10.7% Asian,
symptoms group 19.6% Hispanic,
18.7% Other or
Multiracial
Miller 2018116 Cohort + case- Community sample n = 17 ADHD, T1: »3.0 »96.0−132.0 ADHD: 41.0%, Household annual ADHD: 27% non- CASI-5; NICHQ Executive function Fair

ARTICLE IN PRESS
control of infants with n = 30 no-ADHD T2: »6.0 no-ADHD: 57.0% income. ADHD: 7% White. No-ADHD: Vanderbilt
elevated or T3: 12.0 <$25,000, 25% 18% non-White scale; K-SADS
typical familial T4: »24.0 $50,000 75,000,
likelihood for 32%
autism, who later $75,000 100,000,
met Dx criteria for 36% >$100,000. No-
ADHD or not ADHD: 29%
$50,000 75,000,
24%
$75,000 10,000,
47% >$100,000
Miller 2019a117 Cohort Community sample n = 291 T1: »4.0 T2: »84.0 53.6% Maternal education. Parental. Caucasian: SNAP-IV Activity level; Good
of infants T3: »108.0 16.2% High school, 69.4% mothers, Emotional
Journal of the American Academy of Child & Adolescent Psychiatry

41.9% college, 68.7% fathers; functioning


35.7% African American:
postgraduate, 5.5% 16.5% mothers,
other, 0.7% missing 18.6% fathers;
Hispanic: 7.2%
mothers, 5.5%
fathers; Asian:
3.1% mothers,
2.7% fathers,
Other: 3.4%
mothers, 3.1%
fathers; Missing:
0.3% mothers,
Volume 00 / Number 00 / & 2021

1.4% fathers
Miller 2019b118 Cohort Community sample n = 291 T1: »9.0 T2: »84.0 53.6% As above for Miller As above for Miller SNAP-IV Emotional functioning; Good
of infants T2: »60.0 T3: »108.0 2019a116 2019a116 Executive function
Miller 2020119 Cohort + case- Community sample n = 43 EL, T1: »12.0 0.0−12.0 EL: 32.6%, Household annual Not reported ADHD Dx in Activity level; Good
control of infants with n = 40 TL T2: »18.0 TL: 42.5% income. EL: 35% < parents/ Executive function;
elevated and T3: »24.0 $100,000, 52.5% > siblings with General cognitive
typical familial $100,000, 12.5% DSM-5 and language
likelihood for Missing. TL: 53.5% < Checklist for abilities
ADHD (EL vs TL) $100,000, 32.6% > ADHD, CAARS,
$100,000, 13.9% CASI-5 and
Missing NICHQ
Vanderbilt
scale

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Miyahara 2014120 Case-control Clinical sample of n = 93 ADHD, T1: »36.0 Same age as ADHD: 25.8%, Not reported Not reported Community Activity level Poor
preschoolers with n = 76 TD T2: »48.0 neurocog. TD: 33.0% Clinical ADHD
clinical ADHD Dx Dx; ADHD
and community Rating Scale IV
sample of TD pre- (DuPaul)
schoolers
Nesayan 2018121 Cohort Community sample n = 155 61.44 (9.60) Same age as 51.0% Not reported Not reported Conners Sensory processing Good
of preschoolers neurocog.
O’Callaghan Cohort + case- Community sample n = 108 ADHD, T1: »6.0 T3: »60.0 48.0% (Whole Not reported Not reported CBCL Sleep Good
2010122 control of infants who n = 1,701 no-ADHD T2: 24.0−48.0 T4: »168.0 cohort)
later had elevated
ADHD symptoms
or not
Olson 2000123 Cohort Community sample n = 168 T1: »6.0 T4: 84.0−120.0 43.0% 70% Middle class, 16% Not reported CBCL Executive function; Good
of infants (Follow-up ns 90 136) T2: »13.0 T5: »204.0 upper class, 13% Early regulatory
T3: »24.0 working class problems
Otterman 2019124 Cohort Population sample n = 4,450 T1: »36.0 T3: 60.0−72.0 Not clear Maternal education. ADHD: 62.4% Dutch, CBCL; DISC-YC Executive function; Good
of infants with and T2: »48.0 T4: »84.0 ADHD: 5.2% low, 10.3% non-Dutch Emotional
without later 38% medium, 56.8% Western, 27.3% functioning

ARTICLE IN PRESS
research Dx of high. No-ADHD non-Western. No
ADHD part of cohort not data for no-ADHD
reported part of cohort
Overgaard Cohort + case- Population sample n = 248 ADHD, »18.0 ADHD: 41.7 (1.3), ADHD: 43.0%, “Representative of the “Mostly Caucasian” CBCL; PAPA Emotional functioning Good
2014125 control of infants with and n = 82 no-ADHD no-ADHD: 41.6 no-ADHD: 49.0% population”
without later (1.2)
research Dx of
ADHD
Papageorgiou Cohort Community sample n = 120 T1: 7.69 (6.34) T2: »41.0 54.0% “Majority middle SES” “Mostly Caucasian” Parent-rated Executive function; Good
2014126 of infants T2: »41.0 Revised Rutter Social functioning
Scale
Pappa 2014127 Cohort Population sample n = 581 37.5 (1.5) 70.8 (2.4) 48.0% Not reported Not reported except CBCL Social functioning Good
of infants that of all-Dutch
origin
Pauli-Pott 2018128 Cohort Community sample n = 120 T1: 54.4 (3.2) Same age as 43.2% Parental occupation. Not reported ADHD Rating Executive function Good
of preschoolers T2: 66.9 (3.4) neurocog. 20.8% of mothers Scale
with and without and 83.2% of fathers (Dopfner);
elevated ADHD working full-time, PACS

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


symptoms 45.6% of mothers
and 3.2% of fathers
working part-time,
32.8% of mothers
and 12.2% of fathers
not working, no
data for 0.8% of
mothers and 2.4% of
www.jaacap.org

fathers. Parental
education: 10.4% of
mothers and 18.4%
of fathers with basic
education, 32% of
mothers and 20.8%
of fathers with work
qualification, 19.2%
mothers and 20.8%
fathers high school,
38.4% of mothers
and 38.4% of fathers
college

(continued)
15
16

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Pauli-Pott 2019129 Cohort Community sample n = 120 T1: 54.4 (3.2) Same age as 43.2% As above for Pauli-Pott Not reported ADHD Rating Executive function Good
of preschoolers T2: 66.9 (3.4) neurocog. 2018127 Scale
with and without (Dopfner);
elevated ADHD PACS
symptoms
Perren 2007130 Cohort Community sample n = 160 62.28 (6.48) 74.04 (6.70) 42.5% Not reported Not reported SDQ Social functioning Good
www.jaacap.org

of preschoolers

Perrin 2019131 Case-control Community sample n = 45 ADHD, ADHD: 61.0 (6.6), Same age as ADHD: 38.0%), Hollingshead Index. ADHD: 68% White, ECI-4 General cognitive, Good
of preschoolers n = 48 no-ADHD No-ADHD: 58.0 (6.2) neurocog. No-ADHD: 42.0% ADHD: 52.0 (12.4), 61% Non-Hispanic language and motor
with and without no-ADHD: 60.0 (6.0) or non-Latino. No- abilities; Social
elevated ADHD ADHD: 57% functioning;
symptoms White, 91% non- Executive function
Hispanic or non-
Latino
Petersen 2018132 Cohort Community sample n = 153 T1: »30.0 Same age as 46.0% Hollingshead Index. 90% Non-Hispanic CBCL Executive function Good
of preschoolers T2: »36.0 neurocog. Mean: 48.99 (13.29) Caucasian, 3%
T3: »42.0 Hispanic, 3%
African American,

ARTICLE IN PRESS
3% Asian
American, <1%
Multiracial
Peterson 2018133 Cohort Population sample n = 6,067 8.91 (0.89) 24.74 (1.97) 48.2% 27.6% Low 56.6% European SDQ Executive function; Good
of infants deprivation, 36.9% descent, 13.3% Social and
medium Maori, 13.6% emotional
deprivation, 35.5% Asian descent, functioning
high deprivation 12.9% Pacifika,
3.5% MELAA, New
Zealander or
Other
Peyre 2016134 Cohort Population sample n = 1,459 T1: 38.0 (0.8) Same age as 47.9% Mean years of parental Not reported SDQ Language ability Good
of infants T2: 67.8 (1.8) neurocog. education: 13.6 (2.3)
Journal of the American Academy of Child & Adolescent Psychiatry

Prior 1983135 Case-control Clinical sample of n = 20 ADHD, n = 20 ADHD: 48.0 (9.0), Same age as 15.0% (Both Parental occupation. Not reported Community General cognitive and Good
preschoolers with TD TD: 48.0 (6.0) neurocog. groups) ADHD: 10% Clinical ADHD motor abilities;
clinical Dx of mothers and 30% Dx; Conners Social and
ADHD and fathers professional/ emotional
community managerial, 5% functioning; Activity
sample of TD mothers and 45% level; Executive
preschoolers fathers clerical/ function
armed services/
police, 85% mothers
and 25% fathers
manual workers. TD:
5% mothers and
35% fathers
Volume 00 / Number 00 / & 2021

professional/
managerial, 6%
fathers clerical/
armed services/
police, 95% mothers
and 30% fathers
manual workers
Rabinovitz 2016136 Cohort + case- Community sample n = 96 ADHD, 51.24 (5.88) 90.96 (3.84) 29.2% (Whole Nakao Treas SE Total sample: 59% ADHD Rating Emotional functioning; Good
control of preschoolers n = 65 TD (Whole cohort) (Whole cohort) cohort) Prestige Index. White, 13% Asian, Scale IV Executive function
with research Dx Total sample mean: 11.8% Black, (DuPaul); K-
of ADHD or TD 64.42 (17.48) 15.5% Multiracial SADS
or Other, 29.8%
Hispanic

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Rajendran 2013137 Cohort + case- Community sample n = 140 ADHD, T1: 51.72 (5.64) T2: »60.0 27.3% (Whole Nakao Treas SE Total sample: 58.3% ADHD Rating General cognitive, Good
control of preschoolers n = 76 TD T2: »60.0 T3: »72.0 cohort) Prestige Index. White, 12.5% Scale IV language and motor
with research Dx (Whole cohort) T4: »84.0 ADHD: 61.28 (17.87), Black, 10.6% (DuPaul); abilities; Executive
of ADHD or TD (Whole cohort) TD: 69.16 (14.89) Asian, 18.6% BASC-2; K- function
Multiracial, 31.5% SADS
Hispanic
Rajendran 2015138 Cohort + case- Community sample n = 138 ADHD, T1: 51.60 (5.64), T1 and T2 same ADHD: 23.9%, Nakao Treas SE ADHD: 59.4% White, ADHD Rating Executive function Good
control of preschoolers n = 76 TD T2: 63.48 (6.0) age as TD: 32.9% Prestige Index. 15.9% Black, 5.8% Scale IV
with research Dx neurocog. ADHD: 60.06 (18.35), Asian, 18.8% (DuPaul); K-
of ADHD or TD T3: 75.48 (5.88), TD: 68.60 (15.68) Multiracial or SADS
T4: 87.60 (6.24), Other, 35.5%
T5: 99.48 (5.88), Hispanic. TD:
T6: 111.48 (6.12) 55.3% White, 6.6%
Black, 19.7%
Asian, 18.4%
Multiracial or
Other, 25%
Hispanic
Rints 2015139 Cohort Community sample n = 36 43.12 (3.69) Same age as 55.6% “Most children of “Mostly White” SWAN Executive function; Fair

ARTICLE IN PRESS
of preschoolers neurocog. parents with Language ability
university degree”
Roberts 2019140 Cohort Community sample n = 74 27.31 (15.12) 51.81 (13.22) 0.0% (All male) Not reported 77% Caucasian, CBCL Social functioning Good
of infants and 5.12% African
preschoolers with American, 1.3%
Fragile X Hispanic or Latino,
Syndrome 1.3% Native
American or
Alaskan, 15.4%
Biracial
Robson 1997141 Cohort Community sample n = 85 T1: »7.0 T3: 68.7 (2.7) 63.5% Bishen Index of Not reported McCarthy scales; Executive function; Good
of low- T2: »12.0 paternal CBCL General cognitive
birthweight occupation. Mean: ability
infants 42.6 (14.4). 11%
unskilled labor, 69%
semi-skilled
occupation, 20%
skilled trade or

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


occupation. Mean
paternal years of
education: 12.7 (2.6)
Rogers 2012142 Cohort Community sample n = 165 9.23 (0.28) »60.0 48.0% SES risk measured and Not reported SDQ Brain structure Good
of preterm infants used in analysis but
data not reported
Rohrer- Cohort Population sample n = 1,181 42.0 (1.2) Same age as 47.7% Parental education (y). Not reported PAPA Executive function; Good
Baumgartner143 of infants neurocog. Maternal: 15.2 (2.4); General cognitive
www.jaacap.org

paternal: 14.5 (2.6). and language


abilities
Romano 2006144 Cohort Population sample n = 2,946 0.0−23.0 »95.0 49.0% Ratio of household 88.5% White, 2.1% CBCL Early regulatory Fair
of infants annual income to Black, 6.4% Asian, problems
low-income cut-off. 3% Other
16.6% <1, 83.4% >1
Rosch 2018145 Case-control Community sample n = 47 ADHD, ADHD: 61.20 (7.20), Same age as ADHD: 36.2%, Hollingshead Index. Total sample: 85.7% YC-DISC / DICA- Language ability; Brain Good
of preschoolers n = 40 TD TD: 60.0 (7.20) neurocog. TD: 37.5% ADHD: 56.8 (10.9), Caucasian, 9.9% IV; Conners structure
with research Dx TD: 56.8 (11.1) African American,
of ADHD or TD 3.3% Asian, 1.1%
Missing
Ruff 1990146 Cohort Community sample n = 63 PT, T1: »12.0 T3: »42.0 PT: 52.4%, Not reported Not reported Conners Executive function Good
of preterm (PT) n = 91 FT PT: 12.29 (0.69), PT: 42.90 (1.13), FT: 44.0%
FT: 12.13 (0.30); FT: 42.28 (0.48)

(continued)
17
18

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
and full-term (FT) T2: »24.0
infants PT: 24.33 (0.76),
FT: 24.12 (0.39)
Sanson 1993147 Cohort + case- Community sample n = 65 ADHD, T1: 4.0−8.0 T4: 60.0−72.0 ADHD: 40.0%, Composite measure Included in Rutter Child Social and emotional Good
control of infants with and n = 70 no-ADHD T2: 32.0−36.0 T5: 84.0−96.0 no-ADHD: 30.0% including parental composite SES Behavior functioning; Early
without later T3: 44.0−48.0 occupation and measure only Questionnaire regulatory
www.jaacap.org

elevated ADHD education and problems; Activity


symptoms ethnicity. ADHD: level; Executive
3.95 (1.11), no- function
ADHD: 3.60 (1.16)
Santo 2009148 Cohort Community sample n = 80 »12.0−24.0 »48.0−71.0 50.0% All children of low SES Not reported Conners General cognitive and Good
of preterm infants (no further info) motor ability
149
Saudino 2018 Cohort Population sample n = 618 »24.0 »36.0 Not reported Hollingshead Index. 85.4% White, 3.2% CBCL Activity level Good
of twins Mean: 50.9 (14.1) Black, 2% Asian,
7.3% Multiracial,
2.2% Other
Schmid 2014150 Cohort Population sample n = 1,120 T1: »5.0 T4: »102.0 49.4% 34% Lower class, 99% German Mannheim Parent Early regulatory Good
of infants with T2: »20.0 36.6% middle class, Interview problems
neonatal risk T3: »56.0 29.4% upper class.

ARTICLE IN PRESS
(birth
complications;
preterm)
Schneider 2020151 Case-control Community sample n = 49 ADHD, ADHD: 60.0 (7.20), Same age as ADHD: 40.1%, Hollingshead Index. Total sample: 90% YC-DISC / DICA- Emotional functioning; Good
of preschoolers n = 35 TD TD: 58.80 (6.0) neurocog. TD: 48.6% ADHD: 56.7 (10.3), Caucasian, 5% IV; Conners Executive function
with research Dx TD: 59.1 (9.0) African American,
of ADHD or TD 5% Other
Scott 2013152 Cohort + case- Population sample n = 173 ADHD, T1: »6.0 T5: »84.0 ADHD: 15.6%, Paternal occupation % Non-White. DAWBA Sleep Good
control of infants with and n = 8,022 no-ADHD T2: »18.0 no-ADHD: 49.4% (% unskilled). ADHD: 5.7%; no-
without later T3: »30.0 ADHD: 13.5; no- ADHD: 3.8%
research Dx of T4: »42.0 ADHD: 10.7%.
ADHD Maternal education
(achieved less than
Journal of the American Academy of Child & Adolescent Psychiatry

expected at age 16
y). ADHD: 14.1%;
no-ADHD: 14.4%
Seguin 2009153 Cohort Population sample n = 1,693 40.6 (0.58) Same age as 51.0% Not reported Not reported PBQ General cognitive and Poor
of infants neurocog. language abilities;
Executive function
Shephard 2019a15 Cohort Community sample n = 104 T1: »7.0 T4: »90.0 59.6% (Whole Not reported Not reported Conners Activity level; Fair
of infants with (n = 54 EL-ASD, n = 50 EL-ASD: 7.31 (1.19), TL- EL-ASD: 90.81 cohort) Executive function;
elevated and TL-ASD) ASD: 7.38 (1.24); (6.33), TL-ASD: Social and
typical familial T2: »14.0 89.34 (4.81) emotional
likelihood for EL-ASD: 13.68 (1.57), functioning
autism (EL-ASD TL-ASD: 13.92 (1.33);
and TL-ASD) T3: »24.0
Volume 00 / Number 00 / & 2021

EL-ASD: 23.92 (1.15),


TL-ASD: 23.87 (0.68)
Shephard 2019b16 Cohort Community sample n = 31 6.15 (0.26) Familial likelihood 48.4% 61.3% mothers Class C Not reported Maternal ASRS Resting-state Good
of infants of low- for ADHD (low SES), 38.7% scores neurophysiological
SES adolescent assessed in first Classes D/E (very activity
mothers trimester of low SES). 6.5% of
pregnancy Mothers completed
primary school only,
71% of mothers
completed middle
school, 9.7%
mothers completed
high school

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Silberg 2015154 Cohort Population-based n = 312 »12.0 »36.0 Not reported 63% of Families living Not reported DISC-YC Early regulatory Good
sample of twins below the median problems; Social
annual household functioning;
income level; 51% Executive function
receiving public
assistance. 82% of
mothers and 77% of
fathers with high
school education
Silverman 1992155 Cohort Community sample n = 69 24.0 (1.0) »48.0 49.3% Duncan SE Index. 100% White Yale Children’s Executive function; Fair
of infants Mean: 53.6 (21.6) Inventory Activity level; Social
and emotional
functioning
owall 2017156
Sj€ Cohort Community sample n = 128 T1: 63.0 (1.12) T3: 71.0 (2.54), 51.0% Measured by parental Not reported ADHD Rating Executive functioning Good
of preschoolers T3: 78.0 (2.54), education level but Scale IV
with and without T4: 213.0 (2.59) data not reported (DuPaul);
ADHD symptoms Conners
Skogan 2014157 Cohort + case- Preschoolers with n = 150 ADHD, ADHD: 41.5 (1.2), Same age as ADHD: 46.7%, Not reported Not reported PAPA Executive function Good
control research Dx of n = 455 TD TD: 41.7 (1.3) neurocog. TD: 47.5%

ARTICLE IN PRESS
ADHD or TD
selected from
population-based
cohort
Skogan 2015158 Cohort + case- Preschoolers with n = 1,045 Full cohort, 41.8 (1.3) (Whole Same age as Full cohort: 48.0%, Maternal education (y). Not reported. PAPA Emotional functioning; Good
control research Dx of n = 104 ADHD, cohort) neurocog. ADHD: 36.5%, Whole cohort mean: Executive function
ADHD or TD n = 117 TD TD: 44.4% 15.3 (2.3)
selected from
population-based
cohort
Slopen 2012159 Cohort Community sample n = 114 6.0−30.0 »54.0 Not reported Not reported Not reported PAPA Social and emotional Good
of abandoned / functioning
institutionalized
infants and
nonabandoned
community
controls
Smith 2017160 Cohort + case- Community sample n = 170 ADHD, »36.0 Same age as ADHD: 52.4%, Carstairs Index of Not reported Weiss-Peters Emotional functioning; Good

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


control of preschoolers n = 88 no-ADHD neurocog. no-ADHD: 54.5% Deprivation. ADHD Activity Scale Sleep
with and without boys: 1.06 (1.78);
elevated ADHD ADHD girls: 0.92
symptoms (1.97); no-ADHD
boys: 0.85 (2.03);
no-ADHD girls:
0.13 (2.38).
Sullivan 201517 Case-control Community sample n = 38 EL, »6.7 (whole cohort) Familial likelihood EL: 36.8%, Parental education. Not reported Parental Social and emotional Fair
www.jaacap.org

of infants with n = 10 TL for ADHD TL: 50.0% EL: 60.5% of symptoms on functioning;
elevated and measured in mothers and 63.6% the Barkley Executive function
typical familial second of fathers had AARS
likelihood for trimester of undergraduate
ADHD (EL vs TL) pregnancy degree or higher.
TL: 90% of mothers
and 75% of fathers
had undergraduate
degree or higher

(continued)
19
20

SHEPHARD et al.
TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
Tamm 2016161 Cohort Community sample n = 61 »48.0 Same age as 16.4% Maternal education. “Predominantly non- PAPA; SWAN General cognitive and Good
of preschoolers neurocog. 3.3% Some high Hispanic White” language abilities;
with EF problems school, 11.7% high Executive function;
school graduate, Emotional
18.3% some functioning
college, 25%
www.jaacap.org

postgraduate
degree or
professional training
Towe-Goodman Cohort Community sample n = 636 T1: 7.21 (1.04), T3: 36.8 (1.5) 48.0% 48% of Families with 77% White, 23% DSM-IV ADHD Executive function Good
2011162 of low-income T2: 15.5 (0.94) household incomes Black Rating Scale;
infants <200% below SDQ
Federal Poverty
threshold. Mean
income-needs ratio
of sample: 2.51
(1.87)
Treyvaud 2012163 Cohort Community sample n = 170 PT, »24.0 »60.0 PT: 47.0%, Social Risk Score Not reported SDQ Early regulatory Good
of preterm (PT) n = 64 FT FT: 55.0% (0 = low risk to problems

ARTICLE IN PRESS
and full-term (FT) 12 = high risk). PT
infants mean: 2.0, FT mean:
1.0. Primary
caregiver education.
PT: 12% high school
incomplete, 58%
high school
complete, 30%
Tertiary education.
FT: 1% high school
incomplete, 40%
high school
complete, 59%
Journal of the American Academy of Child & Adolescent Psychiatry

tertiary education
Von Stauffenberg Cohort Community sample n = 776 T1: »54.0 T2: First grade 51.0% Mean income-to- 84% Caucasian, 10% CBCL; DBDRS Executive function Good
2007164 of preschoolers (»84.0), needs ratio: 4.03 African American,
T3: third grade (2.91). 22% Poor, 6% Other
(»108.0) 78% not poor
Walcott 2009165 Cohort Community sample n = 47 »60.0 »72.0 53.2% 73% of Children Not reported BASC-2 Language ability Good
of preschoolers eligible for free or
reduced-price
lunches
Waller 2015166 Cohort Community sample n = 240 41.41 (2.09) Same age as 49.2% Median family annual 86% European CBCL Social and emotional Good
of preschoolers neurocog. income: $52,000 American, 5% functioning;
with and without African American, Executive function
externalizing 8% Biracial
Volume 00 / Number 00 / & 2021

problems
Wang 2018167 Cohort Population sample n = 12,634 T1: »18.0 T3: »60.0 49.1% Not reported Not reported CBCL Language ability Good
of infants T2: »36.0
Wichstrøm Cohort Two cohorts n = 1042 Norway, Norway: 56.40 (3.6), Norway: 126.0 Norway: 50.9%, Norway (occupational Norway: 91% Norway: PAPA/ Social and emotional Good
2018168 (Norway; n = 622 Barcelona Barcelona: 45.6 (3.96) (1.92), Barcelona: 50.0% status): 5.7% leader, Norwegian, 5.8% CAPA, functioning;
Barcelona) of Barcelona: 116.4 25.7% professional Western Barcelona: DSM Executive function
preschoolers (4.20) lower level, 26% countries, 3.2% Dx Interview (both cohorts)
formally skilled non-Western
worker, 0.5% skilled Countries.
worker, 3.1% Barcelona: 91.1%
unskilled worker. Caucasian, 4.7%
Barcelona (SES): Latino, 1% Asian,
35.6% high, 31.6% 3.2% Other

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 1 Continued
First author, Sample size (for Age at neurocog., Age at ADHD, ADHD outcome Neurocognitive NOS quality
year, reference Study design Population analysis) mo (SD) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) domain(s) rating
medium-high,
14.4% medium,
13.7% medium-low,
4.7% low

Williams 2016169 Cohort + case- Population sample n = 112 ADHD Dx, T1: 0.0−12.0 T3: 96.0−108.0 ADHD Dx: 25.0%, SE position. ADHD: “Representative of Parent-report Emotional functioning; Good
control of infants with and n = 648 ADHD ADHD Dx: 8.4 (2.7), ADHD symptoms: 0.09 (1.1); ADHD Australia” Community Executive function;
without later symptoms, ADHD symptoms: 8.9 37.0%, symptoms: 0.08 Clinical Dx; Sleep
ADHD clinical Dx/ n = 3,349 no-ADHD (2.6), no-ADHD: 52.0% (0.90); no-ADHD: SDQ
elevated no-ADHD: 8.5 (2.5); 0.14 (1.0)
symptoms T2: 24.0−36.0
Willoughby Cohort Community sample n = 1074 T1: »6.0 T5: First grade 50.0% 78% Low-income 44% African Pelham ADHD Activity level; Good
201718 of low-income T2: »15.0 (»72.0−84.0) American (no Scale Emotional
infants T3: 24.0 further info) functioning;
T4: »36.0 Executive function
Winsler 1999170 Case-control Community sample n = 18 ADHD, ADHD: 45.5 (6.5), Same age as 22.0% (Both Parental occupation. ADHD: 67% DSM-III-R ADHD Social functioning; Good
of preschoolers n = 22 no-ADHD no-ADHD: 46.7 (5.1) neurocog. groups) ADHD: 11% fathers Caucasian, 11% Rating Scale Executive function
with and without and 22% mothers African American,
elevated ADHD unemployed, 6% 5% Asian

ARTICLE IN PRESS
symptoms mothers unskilled American, 17%
workers, 28% fathers Other. No-ADHD:
and 22% mothers 64% Caucasian,
skilled workers, 61% 9% Latino/
fathers and 51% Hispanic, 9%
mothers Asian-American,
professional. No- 18% Other
ADHD: 41%
mothers
unemployed, 5%
fathers and 5%
mothers unskilled
workers, 18% fathers
and 18% mothers
skilled workers, 77%
fathers and 36%
mothers
professional

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


Woodward Cohort Community sample n = 100 PT, T1: »24.0 T3: »108.0 PT: 49.0%, Parental occupation. PT: 86% Maternal DAWBA Executive function Good
2016171 of preterm (PT) n = 107 FT T2: »48.0 FT: 45.7% PT: 26% European origin;
and full-term (FT) Professional/ FT: 87.9%
infants managerial, 45% maternal
technical/skilled, European origin
29% semi-/
unskilled/
unemployed. FT:
www.jaacap.org

35.5% professional/
managerial, 54.2%
technical/skilled,
10.3% semi-/
unskilled/
unemployed
Yew 2017172 Cohort Population sample n = 3,236 T1: 48.0−60.0 T1: Same age as 49.7% Not reported for full Not reported SDQ Language ability; Good
of preschoolers neurocog. sample Social and
T2: 72.0−84.0, emotional
T3: 96.0−108.0, functioning;
T4: 120.0−132.0 Executive function

(continued)
21
ARTICLE IN PRESS
SHEPHARD et al.

NOS quality

Adolescents; DISC = Diagnostic Interview Schedule for Children; DSM = Diagnostic and Statistical Manual of Mental Disorders; Dx = diagnosis; ECI = Early Childhood Inventory; ICD = Interna-

and Behavior Questionnaire; neurocog. = neurocognitve asses; PAPA = Preschool Age Psychiatric Assessment; PBQ = Preschool Behavior Questionnaire; SDQ = Strengths and Difficulties
Note: ASRS = Adult Self-Report ADHD Rating Scale; Barkley AARS = Barkley Adult ADHD Rating Scale; BASC = Behavior Assessment System for Children; CAARS = Conners Adult ADHD Rat-

DAWBA = Development and Well-Being Assessment; DBDS = Disruptive Behavior Disorders Scale; DBRS = Disruptive Behavior Rating Scale; DICA = Diagnostic Interview for Children and

Questionnaire; SES = socio-economic status; SIDAC = Structured Interview for Diagnostic Assessment of Children; SNAP = Swanson, Nolan and Pelham Rating Scale. RBPCL = Revised Behav-
ing Scale; CAPA = Child and Adolescent Psychiatric Assessment; CASI = Child and Adolescent Symptom Inventory; CBCL = Child Behavior Checklist; Conners = Conners ADHD Rating Scale;

tional Classification of Diseases; K-SADS = Schedule for Affective Disorders and Schizophrenia for School-age Children−Present and Lifetime Version; MacArthur HBQ = MacArthur Health
or provided insufficient information on racial/ethnic back-
rating

grounds to characterize the samples, 46 studies (30.9%)


Good

were conducted with mainly White participants, and 11


studies (7.4%) included participants from diverse racial or
Neurocognitive

Executive function;
Language and
motor abilities
domain(s)

ethnic backgrounds. NOS quality ratings were good for 117


studies, fair for 18 studies, and poor for 14 studies. All 149
studies were included in narrative synthesis; 136 studies

ior Problem Checklist; SWAN = Strengths and Weaknesses of ADHD-symptoms and Normal-behavior; YC-DISC = Diagnostic Interview for Children−Young Children.
were included in meta-analyses (Figure 1). Findings in each
ADHD outcome

neurocognitive/behavioral domain are reported in the fol-


BASC; DBDRS;
measure(s)

lowing sections. The forest plot in Figure 2 provides an


DISC-IV

overview of the pooled effect sizes for each domain.


7.3% Puerto Rican,
African American,

African American,
Multi-ethnic. No-

13% Multi-ethnic
American, 14.3%

American, 11.4%
Race/ ethnicitya

General Cognitive, Language and Motor Abilities


ADHD: 68.3%
28.6% Puerto
Rican, 17.9%
ADHD: 39.2%

Meta-analysis conducted on 15 effect sizes from 14 studies


European

European

(Table S1, available online) showed that poorer general cog-


nitive ability was significantly associated with ADHD
(g = 0.46, p < .0001; Figure S1, available online,
$55.60k (46.29k); no-
Mean annual family
income. ADHD:

Figure 2), with significant heterogeneity across studies/


ADHD: $65.80k

measures [Q(14) = 66.66, p < .0001, I2 = 78.44%] but no


SES

(50.34k)

evidence of publication bias (Tau = 0.1619, p = .4351)


(Figure S2, available online). Subgroup analyses indicated
that effect size conversion and type of analysis were signifi-
Sex (% female)

no-ADHD: 45.5%
ADHD: 39.3%,

cant moderators, with larger effect sizes for nonconverted


effect sizes and for analyses comparing cognitive ability
between preschoolers with and without ADHD (Supple-
ment 5 and Figures S3 and S4, available online).
Age at ADHD,
mo (SD)

Meta-analysis on 33 effect sizes from 23 studies


neurocog.
Same age as

(Table S1, available online) showed that poorer language


ability was significantly associated with ADHD (g = 0.42,
p = .0006, Figure S1, available online, Figure 2), with signif-
Age at neurocog.,

Racial and ethnic categories are reported as described in the original studies.

icant heterogeneity [Q(32) = 1707.37, p < .0001,


mo (SD)

(Whole cohort)

I2 = 99.44%] and publication bias (Tau = 0.3302,


56.58 (3.68)

p = .0070, Figure S2, available online). Subgroup analyses


were nonsignificant (Supplement 5, available online).
Meta-analysis on 26 effect sizes from 18 studies
Sample size (for

(Table S1, available online) indicated that poorer motor


n = 123 no-ADHD
analysis)
Community sample n = 28 ADHD,

ability was significantly associated with ADHD (g = 0.35,


p < .0001, Figure S1, available online, Figure 2), with sig-
nificant heterogeneity [Q(25) = 175.98, p < .0001,
I2 = 95.86%] but no evidence of publication bias
with and without
elevated ADHD
of preschoolers
Population

(Tau = 0.2191, p = .1174, Figure S2, available online).


symptoms

Subgroup analyses were nonsignificant (Supplement 5,


available online).
Study design

Social and Emotional Functioning


Case-control
TABLE 1 Continued

Meta-analysis on 49 effect sizes from 33 studies (Table S2,


available online) indicated that greater social problems and
increased surgency/approach behaviors were significantly
year, reference
First author,

associated with ADHD (g = 0.23, p = .0242, Figure S5,


Youngwirth
2007173

available online, Figure 2), with significant heterogeneity


[Q(42) = 1095.25, p < .0001, I2 = 98.51%] but no
a

22 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

FIGURE 2 Summary of Findings From the Meta-analyses

Note: Forest plots present a summary of the pooled effect sizes and 95% confidence intervals from the meta-analyses investigating associations between ADHD and early-
life neurocognitive and behavioral precursors (left panel; the pooled effect size is shown for each neurocognitive and behavioral domain investigated) and the effects of
early interventions on those precursors (right panel; the pooled effect size for each neurocognitive and behavioral outcome measure is shown).
a
Significant effects.

publication bias (Tau = 0.0033, p = .9750, Figure S6, avail- significant heterogeneity [Q(10) = 92.93, p < .0001,
able online). Age at social functioning was a significant I2 = 93.13%] but no publication bias (Tau = 0.0909,
moderator of the association with ADHD in subgroup anal- p = .7612, Figure S9, available online). Subgroup analyses
ysis (Supplement 6, available online); poorer social func- indicated that age was a significant moderator (Supple-
tioning was only significantly associated with ADHD in 3- ment 8, available online), although only 1 study investigated
to 5-year-olds and not in 0- to 2-year-olds (Figure S7, avail- regulatory problems in 3- to 5-year-olds, so no further
able online). investigation was conducted (Table S4, available online).
Meta-analysis on 55 effect sizes from 38 studies
(Table S2, available online) showed that poorer emotional Sensory Processing
processing/regulation was significantly associated with Meta-analysis on 10 effect sizes from 6 studies (Table S5,
ADHD (g = 0.46, p < .0001, Figure S5, available online, available online) showed that greater sensory processing aty-
Figure 2), with significant heterogeneity [Q(54) = 1218.68, picalities were significantly associated with ADHD
p < .0001, I2 = 97.30%] but no publication bias (g = 0.52, p < .0001, Figure S10, available online, Figure 2),
(Tau = 0.1612, p = .0827, Figure S6, available online). Sub- with significant heterogeneity [Q(9) = 84.06, p = .0096,
group analyses were nonsignificant (Supplement 6, available I2 = 89.52%] but no publication bias (Tau = 0.0455,
online). p = .8566, Figure S10, available online). Subgroup analyses
were nonsignificant (Supplement 9, available online).
Sleep
Meta-analysis on 26 effect sizes from 11 studies (Table S3, Activity Level
available online) showed that greater sleep problems were Meta-analysis of 36 effect sizes from 30 studies (Table S6,
significantly associated with ADHD (g = 0.29, p = .0005, available online) showed that higher activity levels were sig-
Figure S8, available online, Figure 2), with significant het- nificantly associated with ADHD (g = 0.54, p < .0001,
erogeneity [Q(25) = 430.83, p < .0001, I2 = 93.83%] and Figure S11, available online, Figure 2), with significant het-
publication bias (Tau = 0.2994, p = .0324, Figure S8, avail- erogeneity [Q(35) = 474.51, p < .0001, I2 = 97.95%] but
able online). Subgroup analyses were nonsignificant (Sup- no publication bias (Tau = 0.0667, p = .5797, Figure S11,
plement 7, available online). available online). Subgroup analyses were nonsignificant
(Supplement 10, available online).
Early Regulatory Problems
Meta-analysis on 11 effect sizes from 11 studies (Table S4, Executive Function: Inhibition
available online) indicated that greater early regulatory prob- Meta-analysis on 47 effect sizes from 31 studies (Table S7,
lems were significantly associated with ADHD (g = 0.30, available online) showed that poorer inhibition was signifi-
p = .0002, Figure S9, available online, Figure 2) with cantly associated with ADHD (g = 0.63, p < .0001,
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 23
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

Figure S12, available online, Figure 2), with significant het- significantly associated with ADHD (g = 0.34, p = .0351,
erogeneity [Q(46) = 1065.55, p < .0001, I2 = 95.96%] and Figure S17, available online, Figure 2), without significant
publication bias (Tau = 0.3137, p = .0015, Figure S12, heterogeneity [Q(3) = 2.34, p = .5057, I2 = 0.00%] or publi-
available online). Subgroup analyses were nonsignificant cation bias (Tau = 0.0000, p = 1.0000, Figure S17, available
(Supplement 11, available online). online). Subgroup analyses were not conducted for reaction
time because of insufficient variability in subgroup factors.
Executive Function: Attention Meta-analysis on 10 effect sizes from 7 studies (Table
Meta-analysis on 28 effect sizes from 24 studies (Table S8, S12, available online) showed that greater intraindividual
available online) showed that poorer attention was signifi- variability was significantly associated with ADHD
cantly associated with ADHD (g = 0.39, p < .0001, (g = 0.73, p = .0068, Figure S18, available online, Figure 2),
Figure S13, available online, Figure 2) with significant het- with significant heterogeneity [Q(9) = 48.65, p < .0001,
erogeneity [Q(27) = 139.67, p < .0001, I2 = 81.04%] but I2 = 90.65%] but no publication bias (Tau = 0.4222,
no publication bias (Tau = 0.1164, p = .3997, Figure S13, p = .1083, Figure S18, available online). Subgroup analyses
available online). Subgroup analyses were not significant were nonsignificant (Supplement 16, available online).
(Supplement 12, available online).
Executive Function: Impulsivity
Executive Function: Cognitive/Behavioral Flexibility Meta-analysis on 5 effect sizes from 5 studies (Table S13,
Meta-analysis on 11 effect sizes from 10 studies (Table S9, available online) indicated that increased impulsivity was
available online) indicated that poorer cognitive/behavioral significantly associated with ADHD (g = 0.55, p = .0220,
flexibility was significantly associated with ADHD Figure S19, available online, Figure 2), with significant het-
(g = 0.54, p = .0025, Figure S14, available online, erogeneity [Q(4) = 15.21, p = .0043, I2 = 72.72%] but no
Figure 2), with significant heterogeneity [Q(10) = 198.91, p publication bias (Tau = 0.4000, p = 0.4833) (Figure S19,
< .0001, I2 = 95.38%] but no publication bias ADHD available online). Subgroup analyses were nonsignificant
(Tau = 0.0182, p = 1.0000, Figure S14, available online). (Supplement 17, available online).
Subgroup analyses were nonsignificant (Supplement 13,
available online).
Executive Function: Global Executive Function
Executive Function: Planning/Organization Meta-analysis on 55 effect sizes from 44 studies (Table S14,
Meta-analysis on 9 effect sizes from 8 studies (Table S10, available online) showed that poorer global executive func-
available online) revealed that poorer planning/organization tion was significantly associated with ADHD (g = 0.63, p
was significantly associated with ADHD (g = 0.87, < .0001, Figure S20, available online, Figure 2), with signif-
p = .0027, Figure S15, available online, Figure 2), with sig- icant heterogeneity [Q(54) = 2127.48, p < .0001,
nificant heterogeneity [Q(8) = 124.81, p < .0001, I2 = 98.09%] but no publication bias (Tau = 0.0620,
I2 = 97.34%] but no publication bias (Tau = 0.2778, p = .5042, Figure S20, available online). Subgroup analyses
p = .3585, Figure S15, available online). Subgroup analyses showed that population type and effect size conversion were
were nonsignificant (Supplement 14, available online). significant moderators (Supplement 18, available online).
Effects sizes were larger for studies that investigated differen-
Executive Function: Working Memory ces between preschoolers with and without ADHD symp-
Meta-analysis on 26 effect sizes from 16 studies (Table S11, toms (Figure S21, available online) and for studies for
available online) indicated that poorer working memory was which effect sizes were not converted (Figure S22, available
significantly associated with ADHD (g = 0.69, p = .0002, online).
Figure S16, available online, Figure 2), with significant het-
erogeneity [Q(25) = 674.04, p < .0001, I2 = 97.72%] and Other Executive Function Measures
publication bias (Tau = 0.4462, p = .0011, Figure S16, Eight studies investigated other forms of executive function,
available online). Subgroup analyses were nonsignificant including error monitoring, visual search, reward process-
(Supplement 15, available online). ing, and visuospatial abilities (Table S15, available online).
These studies could not be included in meta-analysis
Executive Function: Reaction Time and Intraindividual because of the variability in the functions examined and
Variability individual measures analyzed. No clear pattern of alteration
Meta-analysis on 4 effect sizes from 4 studies (Table S12, in these functions in association with ADHD was evident
available online) indicated that slower reaction times were (Table S15, available online).
24 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

Brain Structure and Resting-State Neurophysiological Sample


Activity A total of 32 nonduplicate studies met inclusion criteria
Five studies investigated brain structure in relation to (3,848 participants; k = 28 RCTs, 3,685 participants;
ADHD (Table S16, available online). Meta-analysis of these k = 4 nonrandomized trials, 163 participants) (Figure 3,
studies was not possible because of variability in brain Table 2174−205). Populations studied were preschoolers
regions examined, but most studies (k = 4) found that with ADHD symptoms or diagnoses (k = 19, 59.4%),
smaller brain volumes were significantly associated with general community samples (k = 9, 28.1%), preschoolers
ADHD. Four studies investigated resting-state neurophysio- with internalizing, externalizing, or executive function
logical activity (Table S16, available online). Inconsistency difficulties (k = 3, 9.4%), and preterm infants (k = 1,
in specific neurophysiological measures precluded meta- 3.1%). Most studies (k = 26, 81.3%) included preschool-
analysis, but there was some consistency in the findings, age (age 3 5 years) samples, whereas 2 studies (6.25%)
with elevated theta power significantly associated with focused on infants or toddlers (0 2 years) and 4 studies
ADHD in 2 studies and alpha asymmetry not significantly (12.5%) included samples of both infants/toddlers and
associated with ADHD in 2 studies (Table S16, available preschoolers. Most studies did not report or provided
online). insufficient information on SES (k = 14, 43.8%) or stud-
Aim 2: Do early interventions target early-life neurocog- ied mainly middle-to-high SES participants (k = 6,
nitive and behavioral precursors associated with ADHD? 18.7%); the remaining 12 studies (37.5%) included

FIGURE 3 PRISMA Flowchart: Studies Investigating Neurocognitive Functions in Early Interventions for Attention-Deficit/
Hyperactivity Disorder (ADHD)

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 25


Volume 00 / Number 00 / & 2021
26

SHEPHARD et al.
TABLE 2 Summary of Studies Investigating Neurocognitive Functions in Early Interventions for Attention-Deficit/Hyperactivity Disorder
Neurocognitive
domains for
First author, Sample size (for Age at baseline, ADHD outcome outcome Quality rating
year, reference Trial design Intervention Population analysis) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) measure(s) (Risk of Bias)b
Abikoff 2015174 Parallel groups New Forest Community sample n = 67 NFPP, n = 63 NFPP: 42.96 (6.00), NFPP: 26.9%, Employment status NFPP: 64.6% White, Parent and Executive Low
RCT Parenting of preschoolers HNC, n = 34 WLC HNC: 42.84 (6.00), HNC: 23.8%, (whole sample): 67.7% 20% African Teacher Function
Programme with research Dx WLC: 42.72 (6.00) WLC: 29.4% mothers, 82.7% American, 29.2% Conners IA, HI (Inhibition,
www.jaacap.org

(NFPP) vs Helping of ADHD fathers working. Hispanic, 15.4% and Total Attention)
the Non- Parental education Other. HNC: ADHD
Compliant Child (whole sample). 4.3% 69.4% White,
(HNC) vs Waitlist Mothers and 14.1% 12.9% African
Control (WLC) fathers high school, American, 26.2%
20.9% mothers and Hispanic, 17.7%
28.8% fathers some Other. WLC:
college, 29.4% 78.1% White,
mothers and 30.1% 15.6% African
fathers college American, 21.9%
graduate, 42.2% Hispanic, 6.3%
mothers and 30.1% Other
fathers advanced/

ARTICLE IN PRESS
professional degree
Barkley 2000175 Parallel groups Parent Training (PT) General community n = 39 PT, n = 37 PT: 57.60 (4.80), PT: 24.0%, Hollingshead Index. PT: Not reported Parent and Activity level; Some concerns
RCT vs Special sample of STC, n = 40 PT STC: 57.60 (6.00), STC: 27.0%, 34.5 (25.6) mother, Teacher CBCL Social
Treatment preschoolers with +STC, n = 42 NT PT+STC: 58.80 (6.00), PT+STC: 35.0%, 42.5 (19.8) father; STC: Attention; functioning;
Classroom (STC) elevated ADHD NT: 57.60 (6.00) NT: 45.0% 35.3 (26.5) mother, Parent DISC Executive
vs PT+STC vs No- symptoms 46.5 (22.6) father; PT function
treatment (NT) +STC: 27.9 (19.8) (Inhibition,
mother, 39.0 (18.9) Attention,
father; NT: 30.3 (24.3) Global)
mother, 45.9 (27.7)
father
Bierman 2014176 Parallel groups Research-based General community n = 190 REDI-HS, Full sample: 55.08 Full sample: 54.0% All SES disadvantaged. Full sample: 57% Parent and Language ability; Some concerns
RCT Developmentally sample of low- n = 161 U-HS (3.84) SES Measure: REDI- European Teacher ADHD Social
Journal of the American Academy of Child & Adolescent Psychiatry

Informed Head SES preschoolers HS 22.42 (9.75); U-HS American, 25% Rating Scale functioning
Start (REDI-HS) vs 21.47 (10.17). Income- African American, (DuPaul)
Usual Head Start to-needs ratio: REDI- 17% Latino.
(U-HS) HS 0.88 (0.67); U-HS
0.88 (0.53)

Capodieci 2018177 Parallel groups Working Memory Community sample n = 18 ADHD WMT, ADHD WMT: 65.88 ADHD WMT: 22.2%, Groups matched on Not reported Parent and Executive High
RCT Training program of preschoolers n = 20 non-ADHD (3.81), Non-ADHD WMT: socio-cultural level Teacher function
(WMT) vs No- with and without WMT, Non-ADHD WMT: 50.0%, but no data reported IPDDAI / (Inhibition,
training (NT) elevated ADHD n = 16 ADHD NT, 66.55 (3.63), ADHD NT: 12.5%, IPDDAG IA and WM,
symptoms n = 20 non-ADHD ADHD NT: 65.88 Non-ADHD NT: HI Impulsivity)
NT (3.77), 40.0%
Non-ADHD NT:
Volume 00 / Number 00 / & 2021

66.63 (4.72)
Christakis 2007178 Parallel groups Block Play (BP) vs Community sample n = 88 BP, BP: 21.6 (18.0 - 28.8), BP: 41.0%, Annual Parental Income. BP: 16% Non- Parent CBCL Language ability High
RCT Non-Intervention of toddlers from n = 88 NP NP: 21.3 (18.0 - 28.8) NP: 52.0% BP: 23% <$10,000, Hispanic White, Attention
(Normal Play, NP) high- and low- 27% $10,000 24,999, 14% Hispanic,
income 36% $25,000 49,999, 61% Black, 9%
backgrounds 7% $50,000 74,999, Asian. NP: 16%
7% >$75,000. NP: 25% Non-Hispanic
<$10,000, 21% White, 10%
$10,000 24,999, 32% Hispanic, 64%
$25,000 49,999, 16% Black, 9% Asian
$50,000 74,999, 6% >
$75,000
Cohen 2018179 Cross-over Yoga vs Waitlist Community sample n = 12 Yoga, n = 11 Yoga: 52.00 (7.00), Yoga: 33.0%, % receiving preschool Yoga: 42% Parent and Early regulatory Some concerns
groups RCT Control (WLC) of preschoolers WLC WLC: 46.00 (10.00) WLC: 36.0% grant/voucher. Yoga: Caucasian, 42% Teacher SDQ problems;
55%; WLC: 55%. African American, and ADHD Executive

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 2 Continued
Neurocognitive
domains for
First author, Sample size (for Age at baseline, ADHD outcome outcome Quality rating
year, reference Trial design Intervention Population analysis) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) measure(s) (Risk of Bias)b
with elevated Parental education. 8% Multiracial, 8% Rating Scale IV function
ADHD symptoms Yoga: 27% high Unknown. WLC: Preschool (Attention,
School or less, 73% 46% Caucasian, Inhibition,
Secondary education; 36% African Flexibility, RT,
WLC: 36% High American, 9% IIV)
School or less, 64% Asian, 9%
Secondary education Multiracial
DuPaul 2018180 Parallel groups Face-to-face Community sample n = 14 F2F, F2F: 54.12 (7.56), F2F: 50.0%, Mean annual family F2F: 81.25% White, Parent Conners Early regulatory High
RCT Behavioral Parent of preschoolers n = 13 ON, ON: 54.24 (6.60), ON: 40.0%, income. F2F: 6.25% Hispanic, Early problems;
Training (F2F) vs with research Dx n = 15 WLC WLC: 51.24 (8.16) WLC: 18.75% $47,083.33 (32,508.22); 6.25% Other. ON: Childhood Social
Online Behavioral of ADHD ON: $46,818.18 66.67% White, Rating Scale IO, functioning
Parent Training (25,771.37); WLC: 20% Black, 6.67% RI and Total
(ON) vs Waitlist $58,200.0 (28,007.14) Hispanic, 6.67% ADHD
Control (WLC) Multiracial. WLC:
87.5% White,
6.25% Other,
6.25% Other

ARTICLE IN PRESS
Feil 2016181 Cluster parallel Preschool First Step Community sample n = 26 PFS, PFS: 49.20 (4.80), PFS: 38.5%, NT: Not reported PFS: 34.6% African Teacher Social Social functioning Some concerns
groups RCT to Success (PFS) of preschoolers n = 19 NT NT: 49.20 (4.80) 31.6% American, 53.8% Skills
vs Normal with elevated Caucasian. NT: Improvement
Treatment (NT) ADHD symptoms 36.8% African System Rating
American, 42.1% Scale (SSiS:RS)
Caucasian Hyp/IA
Franke 2020182 Cross-over RCT Triple P Online Community sample n = 27 PPP, n = 26 Full sample: 48.00 Full sample: 28.3% Third of families with 79.2% of New Parent Conners Social functioning Some concerns
Positive Parenting of preschoolers WLC (36.0−48.0) annual income <NZ Zealand Early
Program (PPP) vs with research Dx $75k (US $50k). 55.7% European Childhood
Waitlist Control of ADHD of Mothers had ancestry Rating Scale
(WLC) university degree. and Teacher
SDQ
Halperin 2020183 Parallel groups TEAMS vs Parental Community sample n = 26 TEAMS, TEAMS: 59.40 (5.52), TEAMS: 29.1%, PES: Annual household TEAMS: 3.8% Asian, K-SADS, Parent Language and Low
RCT Education and of preschoolers n = 26 PES PES: 58.56 (7.20) 26.9% income. TEAMS: 4% < 69.2% White, and Teacher motor abilities,
Support (PES with research Dx $10,000, 8% 26.9% Multiracial. ADHD Rating Executive
−active control) of ADHD $10,000 24,999, 12% PES: 7.7% Asian, Scale (DuPaul) function
$25,000 39,999, 20% 3.8% African (Inhibition,
$40,000 69,999, 32% American, 69.2% WM)

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


$70,000 99,999, 24% White, 19.2%
>$100,000. PES: 4.2% Multiracial
<$10,000, 8.3%
$10,000 24,999, 8.3%
$25,000 39,999,
12.5%
$40,000 69,999,
29.2%
www.jaacap.org

$70,000 99,999,
37.5% >$100,000.
Healey 2019184 Parallel groups ENGAGE (ENG) vs Community sample n = 29 ENG, n = 31 ENG: 45.55 (6.74), ENG: 24.1%, Mean household 83% European Parent and Language and Some concerns
RCT Triple P (PPP) of preschoolers PPP PPP: 45.42 (6.52) PPP: 22.6% income: 1 (low) to 10 descent, 11% Teacher BASC- motor abilities,
with elevated (high). ENG: 5.73 European/Maori 2 HI and IA Executive
ADHD symptoms (2.59); PPP: 6.00 (2.65) New Zealand or function
Asian descent (Inhibition,
Global EF)
Huang 2019185 Parallel groups EEG Theta-Beta Clinical sample of n = 45 NFB, n = 45 NFB: 56.40 (6.00), NFB: 27.0%, Not reported Not reported Parent Conners HI Executive Some concerns
RCT Neurofeedback preschoolers with PB, n = 45 NFB PB: 54.00 (2.40), PB: 24.0%, and Hyp function
(NFB) vs community +PB, n = 45 NT NFB+PB: 53.40 NFB+PB: 22.0%, (Attention)
Psychological clinical ADHD Dx (2.40), NT: 40.0%
Behavioral NT: 53.40 (4.80)
Intervention (PB)
vs Combined
27

(continued)
28

SHEPHARD et al.
TABLE 2 Continued
Neurocognitive
domains for
First author, Sample size (for Age at baseline, ADHD outcome outcome Quality rating
year, reference Trial design Intervention Population analysis) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) measure(s) (Risk of Bias)b
NFB+PB vs No-
treatment (NT)
Jarraya 2019186 Parallel groups Yoga vs Physical Community sample n = 15 Yoga, »60.0 Full sample: 62.2% “All children of high- Not reported Teacher ADHD Executive High
RCT Exercise (PE) vs of preschoolers n = 15 PE, average SES” Rating Scale function
www.jaacap.org

No-treatment n = 15 NT (DuPaul) (Attention)


(NT)
Jones 2007187 Parallel groups Incredible Y (IY) vs Community sample n = 50 IY, Full sample: 46.28 Full sample: 32.0% Not reported Not reported Parent Conners None High
RCT No-treatment of preschoolers n = 29 NT (6.16) Total ADHD
(NT) with elevated
ADHD symptoms
Kaaresen 2006188 Parallel groups Mother Infant Community sample n = 71 MIT, »0.00 (All infants) MIT: 48.0%, Maternal and paternal Not reported Parent Parenting None High
RCT Transaction of preterm infants n = 69 CAU CAU: 46.0% mean monthly income Stress Index
Program (MIT) vs (1,000 Norwegian Distractibility/
Care-as-Usual Krone). MIT: 15.8 (7.7) Hyp
(CAU) mothers, 21.1 (18.7)
fathers; CAU: 14.6 (6.7)
mothers, 19.9 (8.1)

ARTICLE IN PRESS
fathers
Landis 2019189 Parallel groups Summer Treatment Community sample n = 24 AWM, n = 25 AWM: 54.72 (7.56), AWM: 37.0%, Hollingshead Index. AWM: 63% Parent and Executive Some concerns
RCT Program pre- of preschoolers NWM NWM: 54.72 (7.92) NWM: 20.0% AWM mean: 40.13 Hispanic/Latino. Teacher function
Kindergarten with elevated (12.87); NWM mean: NWM: 88% DBDRS ADHD (Global EF,
(STP-preK) plus externalizing 39.50 (13.67) Hispanic/Latino WM)
Adaptive Working problems
Memory Training
(AWM) vs
STP_preK plus
Non-adaptive
Working Memory
Training (NWM)
Mendelsohn Factorial RCT Infant Toddler Community sample n = 143 VIP 0 3, VIP/CAU 0 3: 0.0 VIP 0 3: 57.0%, % Low SES. VIP 0 3: % Hispanic mothers. Parent BASC-2 Social functioning Some concerns
Journal of the American Academy of Child & Adolescent Psychiatry

2018190 Video Interaction of infants n = 132 CAU 0 3, −36.0 CAU 0 3: 51.0%, 93%; CAU 0 3: 91%; VIP 0 3: 94%; Attention and
Project (VIP 0-3) vs n = 123 VIP 3 5, VIP/CAU 3 5: 36.0 VIP 3 5: 52.0%, VIP 3 5: 92%; CAU CAU 0 3: 96%; Hyp
Infant Toddler n = 129 CAU 3 5 −54.0 CAU 3 5: 56.0% 3 5: 93% VIP 3 5: 92%;
Care-as-Usual CAU 3 5: 93%
(CAU 0 3) plus
Preschool VIP (VIP
3-5) vs preschool
CAU (CAU 3 5)
Nixon 2001191 Parallel groups Parent Child Community sample n = 17 PCIT, PCIT: 45.52 (6.83), PCT: 18.0%, Annual family income (1: Not reported DSM-IV Executive Some concerns
RCT Interaction of preschoolers n = 17 WLC, WLC: 46.76 (7.50), WLC: 30.0%, <$20k, 2: $20 40k, 3: Structured function
Therapy (PCIT) vs with behavioral n = 21 TDC TDC: 44.71 (5.82) TDC: 29.0% $40 70k, 4: Interview for (Flexibility)
Waitlist Control problems or $70 100k, 5: >$100k). Disruptive
(WLC) vs Typically typical PCIT: 3.0 (1.17); WLC: Behavior
Volume 00 / Number 00 / & 2021

Developing development 3.24 (1.20); TDC: 3.67 Disorders


Controls (TDC) (1.11). −ADHD scale
Papazian 2009192 Parallel groups Executive Function Community sample n = 13 EF, n = 12 NT »24.0−48.0 EF: 15.4%, Not reported Not reported Parent DSM-IV Executive High
RCT Training (EF) vs of preschoolers NT: 16.7% ADHD Rating function
No-training (NT) with elevated Scale (Flexibility)
ADHD symptoms
Re 2015193 Parallel groups Executive Function Community sample n = 13 EF, n = 13 NT EF: 63.42 (4.98), EF: 38.5%, Excluded children with 100% Caucasian Parent and Executive Some concerns
RCT Training (EF) vs of preschoolers NT: 63.03 (4.40) NT: 30.8% low SES Teacher function
No-training (NT) with elevated IPDDAI / (Inhibition,
ADHD symptoms IPDDAG IA and WM,
HI Impulsivity)
Solomon 2018194 Cluster parallel YMCA Playing to Community sample Cohort A: n = 49 PL, Cohort A: Cohort A: “SES diverse sample” “Ethnically diverse Parent and Executive Low
groups RCT Learn (PL) vs of preschoolers n = 68 TM. PL: 45.9 (36.5−62.3), PL: 46.9%, sample” Teacher SDQ function
Tools of the Mind Cohort B: n = 19 PL, TM: 54.1 (37.2 TM: 48.5%. (Inhibition,
(TM) n = 42 TM −55.34). Cohort B: Global EF)

(continued)
Volume 00 / Number 00 / & 2021
Journal of the American Academy of Child & Adolescent Psychiatry

TABLE 2 Continued
Neurocognitive
domains for
First author, Sample size (for Age at baseline, ADHD outcome outcome Quality rating
year, reference Trial design Intervention Population analysis) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) measure(s) (Risk of Bias)b
Cohort B: PL: 57.9%,
PL: 43.5 (37.0−57.1), TM: 40.5%
TM: 42.4 (37.0−50.0)
Sonuga-Barke Parallel groups Parent Training (PT) Community sample n = 30 PT, »36.0 Full sample: 38.5% Not reported Not reported PACS Executive Some concerns
2001195 RCT vs Parent of preschoolers n = 28 PCS, function
Counseling and with elevated n = 20 WLC (Attention)
Support (PCS) vs ADHD symptoms
Waitlist Control
(WLC)
Sonuga-Barke Parallel groups Parent Training (PT) Community sample n = 59 PT, »36.0 Not reported Whole sample: 15% Not reported PACS, Parent None Some concerns
2004196 RCT vs Waitlist Control of preschoolers n = 30 WLC Classes 1 and 2 Werry
(WLC) with elevated (professional), 49.4% Weiss Peters
ADHD symptoms classes 3 and 4 Activity Scale
(akilled), 35.6% classes and Behavior
5 and 6 Checklist
Sonuga-Barke Parallel groups New Forest Clinical sample of n = 133 NFPP, NFPP: 43.04 (7.01), NFPP: 24.0%, % of Parents who left Not reported Parent and Executive Low
2018197 RCT Parenting preschoolers with n = 131 IY, IY: 42.0 (6.49), IY: 29.0%, school without Teacher SNAP- function

ARTICLE IN PRESS
Programme ADHD Dx n = 42 TAU TAU: 42.3 (7.79) TAU: 40.0% qualifications. NFPP: IV IA and HI (Attention)
(NFPP) vs 17%; IY: 10%; TAU:
Incredible Y (IY) vs 10%
Treatment-as-
Usual (TAU)
Strayhorn 1989198 Parallel groups Training Exercises Community sample n = 50 TEPP-F, Full sample: 46.80 Full sample: 56.0% Full sample median Full sample: 64% Parent DSM-III-R Language ability Some concerns
RCT for Parents of of low-SES n = 46 TEPP-M (30.0−68.40) family income: $6,422 Black, 30% White, ADHD
Preschoolers preschoolers with 6% Other
−Full Intervention signs of
(TEPP-F) vs internalizing or
Minimal TEPP externalizing
(TEPP-M) problems
Tamm 2019199 Pilot parallel Generating Community sample n = 34 GAIM, n = 25 GAIM: 43.20 (5.88), GAIM: 30.6%, Annual family income. GAIM: 3% Hispanic, Parent and Language ability; Some concerns
groups RCT Attention, of preschoolers NT NT: 44.40 (5.52) NT: 25.0% GAIM: 6% <$25k, 8% 6% Black, 83% Teacher DBRS Emotional
Inhibition and with executive $25 49k, 19% White, 11% IA and HI functioning;
Memory function $50 75k, 25% Biracial. NT: 3% Executive
Intervention difficulties $75 100k, 42% > Asian, 22% Black, function
(GAIM) vs No- $100k. NT: 16% < 72% White, 3% (Inhibition,

EARLY PRECURSORS AND INTERVENTIONS IN ADHD


training (NT) $25k, 19% $25 49k, Biracial Flexibility,
13% $50 75k, 25% Planning/
$75 100k, 28% > Organisation,
$100k. WM)
Volckert 2015200 Parallel groups Inhibition Training Community sample n = 24 IT, IT: 60.13 (5.14), IT: 62.5% Monthly family income Not reported Parent and Emotional Some concerns
RCT (IT) vs Handicraft of preschoolers n = 23 HS HS: 60.52 (4.42) HS: 78.3% on 1 (low: 0 500 Teacher functioning;
Sessions (Control, euros) to 9 (high: Conners IA and Executive
HS) >4,000 euros) scale. HI function
www.jaacap.org

Full sample mean: (Attention,


6.00 (1.82) Inhibition,
Flexibility, WM)
Weisleder 2016201 Parallel groups Video Interaction Community sample n = 176 VIP, 0.0−36.0 VIP: 56.0%, % Low SES on % Hispanic. VIP: Parent BASC-2 Social Low
RCT Project (VIP) vs of low-income n = 111 BB, BB: 51.0%, Hollingshead Index. 94%; BB: 96%; Attention and functioning;
Building Blocks infants n = 176 TAU TAU: 48.0% VIP: 93%; BB: 92%; TAU: 92% Hyp Executive
(BB) vs TAU: 91% function
Treatment-as- (Attention)
Usual (TAU)
Halperin 2013202 Non-Rand. trial TEAMS Pre- vs Post- Community sample n = 28 62.28 (7.08) 34.5% Not reported 55.2% Caucasian, Parent and None Moderate
training of preschoolers 20.7% African Teacher ADHD
with research Dx American, 13.8% Rating Scale IV
of ADHD Asian, 10.3% (DuPaul)

(continued)
29
30

SHEPHARD et al.
TABLE 2 Continued
Neurocognitive
domains for
First author, Sample size (for Age at baseline, ADHD outcome outcome Quality rating
year, reference Trial design Intervention Population analysis) mo (SD) Sex (% female) SES Race/ ethnicitya measure(s) measure(s) (Risk of Bias)b
Multiracial, 34.5%
Hispanic.
Healey 2015203 Non-Rand. trial ENGAGE (ENG) Pre Community sample n = 25 ENG, ENG: 46.80 (7.44), ENG: 24.0%, Parental education (ENG ENG group only Parent BASC-2 IA Language and Serious
vs Post-training vs of preschoolers n = 44 NI NI: not reported NI: 25.0% group only reported): reported: 92% and Hyp motor abilities;
www.jaacap.org

No-intervention with elevated 32% high school European Executive


controls at follow- ADHD symptoms incomplete, 32% descent, 8% function
up (NI) university degree, mixed European/ (Inhibition,
36% in between New Zealand WM)
Maori descent
Joekar 2017204 Non-Rand. trial Pay Attention Community sample n = 14 PAP, PAP: 68.84 (3.31), 0% (All- male Not reported Not reported Parent and Executive Serious
Program (PAP) vs of preschoolers n = 13 NT NT: 68.76 (3.40) sample) Teacher CSI-4 function
No-training (NT) with elevated IA and HI (Inhibition,
ADHD symptoms Attention)
van Rhijn 2017205 Non-Rand. trial Stay, Play and Talk Community sample n = 26 SPT, n = 13 Full sample: 39.17 Full sample: 46.2% Not reported Not reported Parent Social None Moderate
Program (SPT) vs of preschoolers WLC (24.0−47.0) Skills
Waitlist Control Improvement
(WLC) System Rating

ARTICLE IN PRESS
Scale (SSiS:RS)
Hyp/IA

Note: BASC = Behavior Assessment System for Children; CBCL = Child Behavior Checklist; Conners = Conners ADHD Rating Scale; CSI = Child Symptom Inventory; DBDRS = Disruptive
Behavior Disorders Rating Scale; DBRS = Disruptive Behavior Rating Scale; DISC = Diagnostic Interview Schedule for Children; DSM = Diagnostic and Statistical Manual of Mental Disorders;
HI = Hyperactive/Impulsive symptoms; Hyp = Hyperactivity; IA = Inattention symptoms; IO = Inattention/Overactivity; IPDDAI / IPDDAG = Scale for the Early Identification of ADHD Teacher/
Parent; K-SADS = Schedule for Affective Disorders and Schizophrenia for School-age Children−Present and Lifetime Version; Non-rand. = Non-randomized intervention study; PACS = Paren-
tal Account of Childhood Symptoms; RCT = randomized controlled trial; RI = Restless/Impulsive; SDQ = Strengths and Difficulties Questionnaire; SES = socio-economic status; SNAP = Swan-
son, Nolan and Pelham Rating Scale.
a
Racial and ethnic categories are reported as presented in the original studies.
b
Risk of Bias was assessed with the Cochrane Risk of Bias Tool for Randomised Controlled Trials (RoB-2)23 for randomized controlled trials and with the Cochrane ROBINS-I24 tool for non-ran-
Journal of the American Academy of Child & Adolescent Psychiatry

domized intervention studies.


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

participants from a range of SES backgrounds. Similarly, Neurocognition and Behavior


information on race or ethnicity was not reported (or Language Ability. Meta-analysis on 8 effect sizes from 4
insufficient information was provided) for many studies studies (Table S19, available online) showed no significant
(k = 16, 50%); a further 7 studies (21.9%) included intervention-related improvement in language ability
mainly White participants, and the remaining 9 studies (SMD = 0.08, p = .6448, Figure S25, available online,
(28.1%) included samples of diverse racial/ethnic back- Figure 2). Heterogeneity was significant [Q(7) = 19.22,
grounds. Risk of bias assessments indicated some level of p = .0075, I2 = 78.73%] but publication bias was not
concern for most studies (RCTs: k = 5 low, k = 16 some (Tau = 0.5000, p = .1087). One nonrandomized study
concerns, k = 7 high; nonrandomized studies: k = 2 mod- and 2 RCTs not included in meta-analysis also reported no
erate, k = 2 serious) (Table 2). All 32 studies were significant intervention-related improvement in language
included in narrative synthesis; 22 RCTs were included (Table S19, available online).
in meta-analyses. The interventions investigated are
detailed in Table S17, available online. Most studies
tested interventions that directly or indirectly targeted Social Functioning. Meta-analysis on 12 effect sizes from 5
self-regulation, and several targeted social and emotional studies (Table S20, available online) showed a marginal
functioning, language ability, and executive functions. intervention-related improvement in social functioning
Interventions were diverse and included parent-training (SMD = 0.40, p = .0522, Figure S26, available online,
programs, cognitive training, neurofeedback, and yoga, Figure 2), with significant heterogeneity [Q(11) = 47.38, p
among others (Table S17, available online). Outcome < .0001, I2 = 85.9%] but no publication bias
measures were assessed immediately postintervention and (Tau = 0.3939, p = .0863, Figure S26, available online).
included ADHD symptoms, language and motor abilities, Subgroup analyses were nonsignificant (Supplement 24,
social and emotional functioning, early regulatory prob- available online). One of 2 RCTs not included in meta-
lems, activity level, and executive functions (inhibition, analysis reported significant intervention-related improve-
attention, flexibility, working memory, planning/organi- ments in social functioning (Table S20, available online).
zation, reaction time and intraindividual variability,
impulsivity, and global executive function) (Table 2). Executive Function: Inhibition. Meta-analysis on 9 effect
Results of narrative synthesis and meta-analysis for each sizes from 7 studies (Table S21, available online) revealed
neurocognitive and behavioral outcome measure are no intervention-related improvement in inhibition
detailed in the following sections. Figure 2 presents a (SMD = 0.14, p = .1894, Figure S27, available online,
summary of the pooled effect sizes for each outcome mea- Figure 2), without significant heterogeneity [Q(8) = 9.22,
sure included in meta-analyses. p = .3239] or publication bias (Tau = 0.4444, p = .1194).
The remaining RCTs (k = 3) and nonrandomized trials
Intervention Effects (k = 2) also reported no intervention-related improvement
ADHD Symptoms. Meta-analysis on 29 effect sizes from 20 in inhibition (Table S21, available online).
RCTs (Table S18, available online) revealed significant
postintervention reductions in ADHD symptoms
(SMD = 0.43, p = .0003, Figure S23, available online), with Executive Function: Attention. Meta-analysis on 17 effect
significant heterogeneity [Q(28) = 87.70, p < .0001, sizes from 7 studies (Table S22, available online) showed no
I2 = 72.42%] and publication bias (Tau = 0.2956, significant intervention-related improvement in attention
p = .0246, Figures S23 and S24, available online). One (SMD = 0.68, p = .1541, Figure S28, available online). Het-
study186 was considered an outlier, with SMD lower CI erogeneity [Q(16) = 334.47, p < .0001, I2 = 97.57%] and
limits outside the pooled SMD upper CI limit. Because risk publication bias (Tau = 0.5735, p = .0009, Figures S28 and
of bias for this study was high, the analysis was repeated S29, available online) were significant. The large but non-
without the study; the pooled SMD was reduced but significant SMD reflected an outlier study185 with lower
remained significant (SMD = 0.36, p = .0010; Figure S24, SMD CIs outside the pooled SMD upper CI limit. Meta-
available online, Figure 2). Subgroup analysis were nonsig- analysis repeated without this study revealed a smaller and
nificant (Supplement 22, available online). Most RCTs not nonsignificant pooled effect size (SMD = 0.23, p = .2572,
included in meta-analysis (5 of 8) and all nonrandomized Figure S29, available online, Figure 2). However, 1 of 2
studies (4 of 4) also reported significant intervention-related remaining RCTs and 1 nonrandomized trial reported signif-
reductions in ADHD symptoms (Table S18, available icant intervention-related improvements in attention (Table
online). S22, available online).
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 31
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

Executive Function: Flexibility. Meta-analysis on 4 effect particular vulnerability. Findings from the second meta-
sizes from 4 studies (Table S23, available online) showed no analysis indicated that the early interventions tested in
intervention-related improvement in flexibility RCTs produced significant improvements in ADHD symp-
(SMD = 0.27, p = .1700, Figure S30, available online, toms and working memory, but not in other functions
Figure 2). Heterogeneity [Q(3) = 1.35, p = 0.7173, I2 = 0%] examined.
and publication bias (Tau = 0.6667, p = .3333, Figure S30, A surprisingly large number of studies have investigated
available online) were nonsignificant. One RCT not early-life neurocognitive and behavioral alterations in
included in meta-analyses also reported no intervention ADHD. Consistent with previous narrative reviews,11,19,20
effects on flexibility (Table S23, available online). meta-analyses showed that ADHD was significantly associ-
ated with poorer general cognitive, language and motor abil-
Executive Function: Working memory. Meta-analysis on ities, social and emotional difficulties, early regulatory and
10 effect sizes from 6 studies (Table S24, available online) sleep problems, sensory processing atypicalities, and elevated
showed significant intervention-related improvements in activity levels in the first 5 years of life. ADHD was also sig-
working memory (SMD = 0.37, p = .0200, Figure S31, nificantly associated with early-life executive function diffi-
available online, Figure 2), without significant heterogeneity culties, including poorer inhibition, attention, cognitive/
[Q(9) = 12.46; I2 = 35.54%] or publication bias behavioral flexibility, planning/organization, working mem-
(Tau = 0.3778, p = .1557, Figure S31, available online). ory, and impulse control, as well as slower reaction times,
Subgroup analysis showed that risk of bias was a significant greater intraindividual variability, and poorer global execu-
moderator, with larger effect sizes in studies with higher risk tive functioning. Narrative synthesis also indicated early
of bias (Figure S32, Supplement 28, available online). One alterations in brain structure and resting-state neurophysio-
nonrandomized study also reported significant intervention- logical activity, although further studies are needed to con-
related improvements in working memory (Table S24, firm these findings and to clarify the specific brain regions
available online). and oscillatory frequencies involved, as well as which
cognitive/behavioral functions are affected by such neural
atypicalities. With the exception of social functioning diffi-
Other Neurocognitive Outcome Measures. The remain-
culties, neurocognitive and behavioral alterations were pres-
ing neurocognitive outcome measures were investigated in
ent in the infancy/toddlerhood period (age 0 2 years) as
too few studies for meta-analysis. No significant interven-
well as in the preschool years (age 3 5 years), as indicated
tion-related improvements were reported for early regula-
by nonsignificant moderating effects of developmental stage
tory problems (2 of 2 RCTs), planning/organization (1 of 1
on associations with ADHD. These findings support the
RCT), intraindividual variability (1 of 1 RCT), or activity
widely held view that the developmental alterations that
level (1 of 1 RCT) (Table S25, available online). In contrast,
precede ADHD begin very early in life. However, it should
significant intervention-related improvements were reported
be noted that some of the precursors (eg, sensory processing,
for motor ability (2 of 3 RCTs; 1 of 1 nonrandomized
inhibition, flexibility, working memory) were investigated
study), emotional dysregulation (2 of 2 RCTs), reaction
mainly in preschoolers rather than infants/toddlers. Further-
time (1 of 1 RCT), impulsivity (2 of 2 RCTs), and global
more, publication bias was significant for language, sleep,
executive function (3 of 4 RCTs) (Table S25, available
inhibition, and working memory, and caution is therefore
online).
warranted when interpreting the associations between these
functions and ADHD.
DISCUSSION The pattern of neurocognitive and behavioral difficul-
This is the first systematic review and meta-analysis to inves- ties is partially consistent with current models of early
tigate which early-life neurocognitive and behavioral precur- developmental pathways to ADHD. For instance, Nigg
sors are associated with ADHD and whether these are et al.206,207 propose that extreme positive or negative emo-
targeted in early interventions. Findings from the first meta- tional reactivity in infancy leads to weakened executive
analysis indicated that multiple neurocognitive and behav- function in toddlerhood, elevated ADHD symptoms, and
ioral alterations are involved in the early development of irritability in the preschool years, and clinically significant
ADHD. The largest effect sizes (g > 0.50) were found for ADHD at school age. The current findings support the
sensory processing, activity level, and aspects of executive presence of excessive positive and negative emotional reac-
function (inhibition, flexibility, planning/organization, tivity in the first 2 years of life. However, executive function
intraindividual variability, impulsivity, and global executive difficulties were also present in the infancy/toddlerhood
function), suggesting that these may represent areas of period. Thus, weakened top-down control appears to co-
32 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

occur with rather than result from emotional reactivity ADHD, relatively few studies have assessed these as inter-
problems, at least for some infants. Other developmental vention targets. Our systematic review revealed 28 RCTs
models of ADHD (see Sonuga-Barke208) highlight the and 4 nonrandomized trials testing a range of early interven-
importance of early alterations in neurobiological circuitry tions for ADHD that targeted neurocognitive/behavioral
involved in reward processing, which result in negative reac- precursors and/or measured these as outcomes. Meta-analy-
tions to imposed delays, impulsive delay-avoidant behaviors sis indicated that the early interventions tested in RCTs pro-
and hyperactivity, and difficulties engaging with and orga- duced significant improvements in ADHD symptoms,
nizing oneself within delay-rich environments. The current without evidence of publication bias, although heterogene-
findings of poorer organization/planning abilities, increased ity was significant and not accounted for by factors tested in
activity levels, and negative affect during the 0- to 5-year subgroup analyses. Heterogeneity may instead have
period may be consistent with this model, although it is reflected variability in the specific interventions tested, few
important to note that none of the studies included in this of which were evaluated across more than 1 RCT. Narrative
review specifically investigated reward alterations. synthesis of nonrandomized trials also indicated significant
These developmental models206−208 also emphasize improvements in ADHD symptoms in all interventions
that early causal pathways to ADHD are likely heteroge- tested. Meta-analyses of neurocognitive outcomes indicated
neous and differ across individuals. The significant hetero- that the interventions tested in RCTs produced significant
geneity in our meta-analyses, which was generally not improvements in working memory (without significant het-
accounted for by the subgroup factors that we analysed, erogeneity or publication bias), but not in other functions
would support this view. However, the large cross-study var- examined (language, inhibition, attention, flexibility, social
iability in neurocognitive/behavioral measures likely also functioning). Narrative synthesis further suggested interven-
contributed to heterogeneity in the meta-analyses. It is also tion-related improvements in motor ability, emotional dys-
crucial to highlight that the findings of the current review regulation, impulsivity, and global executive function across
are correlational and do not permit inferences concerning RCTs and nonrandomized trials, although study numbers
causal pathways to ADHD. Future longitudinal work with were low and there was some inconsistency in findings
repeated measurements of neurocognitive and behavioral across studies. Importantly, the majority of intervention
functions is needed to identify the specific developmental studies were conducted with preschool-aged children rather
trajectories that lead to ADHD. than with infants and toddlers; further research is therefore
The current findings also highlight the potential needed to assess whether early intervention can modify pre-
involvement of several functions not typically considered in cursors of ADHD earlier in development.
early developmental models of ADHD, including sleep, lan- There are several limitations to the current study. First,
guage, and sensory processing. Of the functions with the the findings must be interpreted within the context of sig-
largest effect sizes, overactivity and executive function diffi- nificant heterogeneity and publication bias. Further, the
culties have frequently been proposed as key to the develop- generalizability and representativeness of the current find-
ment of ADHD and as potential early intervention ings in regard to the global population is unclear, as most
targets,11,15,19,20,22,206,207 but atypical sensory processing studies either did not adequately report the socioeconomic
has more often been associated with other neurodevelop- and racial/ethnic backgrounds of their participant samples
mental conditions, particularly autism.10 Yet, a potential or included mainly middle-high SES and White partici-
role for sensory issues in both ADHD and autism was previ- pants. It will be crucial for future studies in this area to accu-
ously highlighted,19 and recent empirical work suggests that rately report this information and for greater efforts to be
sensory processing alterations contribute to inattention made in including diverse and representative samples. Sec-
symptoms in older children with ADHD.209 Although the ond, we could not obtain data to compute effect sizes from
current findings should be considered preliminary because all publications, particularly for those reporting nonsignifi-
the meta-analysis of sensory processing was based on only 7 cant findings. Small effect sizes from nonsignificant findings
studies, further research investigating sensory issues in the were therefore underrepresented in our meta-analyses, and
development of ADHD is warranted. In particular, it will the pooled effect sizes may overestimate the true associations
be important to know whether these are an early manifesta- between early neurocognitive precursors and ADHD and
tion of physiological arousal difficulties, which are proposed intervention effects. Third, we used multi-level models to
to represent a core biological mechanism in ADHD that adjust for nonindependence of data for multiple effect sizes
contributes to symptoms.210−212 reported for the same cohort in the same publication. How-
Compared to the volume of studies investigating early ever, some cohorts were analyzed in different publications;
neurocognitive and behavioral precursors associated with because it was not possible to identify whether the entire
Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 33
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

samples matched across these studies, we did not nest effect will develop the condition, as well as other related condi-
sizes across studies in the models, which may have affected tions. These are challenging studies in terms of ethical and
the pooled estimates. Fourth, we could not investigate methodological aspects but are necessary to inform clinical
effects of individual neurocognitive measures and interven- decisions.
tions because of large cross-study variability; resolving this In terms of clinical implications of the current work,
issue this will be an important step for future research. clinicians working with infants, toddlers, and preschoolers
Finally, it should be noted that several new studies on neu- should be aware that young children who later develop
rocognitive precursors to ADHD in infancy213−216 have ADHD are likely to experience many neurocognitive and
been published since the final search date of the current behavioral difficulties, in addition to symptoms. The cur-
review while we conducted the meta-analyses and prepared rent findings suggest that problems with executive function,
the article for publication. A “living” systematic review and overactivity, and sensory atypicalities may be most promi-
meta-analysis in which literature searches and syntheses of nent, but difficulties in other areas not typically associated
findings are regularly updated217 would have ensured that with ADHD, including general cognitive, language, and
the latest findings were included in the review. Although motor development, may also be present and may require
beyond the scope of the current work, we would be enthusi- additional learning support at preschool/school. Social and
astic to share the current data to begin a living systematic emotional difficulties are also likely, and thus a psychosocial
review of early-life precursors and treatment for ADHD. approach to treatment will be important. In terms of spe-
Results from the meta-analyses point to several direc- cific treatments, the evidence compiled in the meta-analysis
tions for future studies. For studies investigating precursors showed that short-term reductions of ADHD symptoms
of ADHD, it will be important to standardize neurocogni- and improvement in working memory with diverse techni-
tive/behavioral measures across cohorts and to investigate ques is possible, and probably behavioral parent-training
profiles of atypicality across children, instead of impairment programs are currently the most well-researched model of
in single functions, as has been suggested previously.11 The intervention. Based on the current findings, which are pro-
specificity of the identified precursors to ADHD over other visional, behavioral parent-training programs in which
common neurodevelopmental conditions (eg, autism) parents learn how to implement strategies to manage non-
should be established, as those conditions frequently co- compliance and to help their child develop self-regulatory
occur with ADHD and are also associated with atypicalities cognitive and emotional abilities, such as the Incredible
in many of the neurocognitive and behavioral functions Years187,197 and New Forest Parenting Programme174,197
revealed here.19,20 Longitudinal designs are needed to estab- parent-mediated interventions, could be recommended as a
lish the predictive value of early neurocognitive precursors treatment approach for these young children. Nevertheless,
and their interactive nature. It will be important to under- further research is needed to demonstrate the effects of the
stand, for example, which atypicalities are the earliest to interventions on early developmental trajectories of ADHD
emerge and whether there are key alterations that play a symptoms or the likelihood of developing ADHD or other
causal role in the development of other neurocognitive diffi- neurodevelopmental conditions.
culties and the emergence of ADHD symptoms. Such pre- In conclusion, the current findings provide initial sup-
cursors may be the most useful early intervention targets. port for the idea that neurocognitive and behavioral precur-
Future studies investigating early interventions for sors in the first years of life predict the development of
ADHD will benefit from a more robust knowledge base on ADHD and may be amenable to early intervention. Chil-
the precursors of ADHD. These studies will need to have a dren aged 0 to 5 years with current or later-emerging
solid theory of change supporting the models tested, specify- ADHD are likely to experience difficulties in a range of
ing how the techniques used are expected to modify out- neurocognitive and behavioral functions, particularly over-
comes. Future work should also investigate psychosocial activity, executive function difficulties, and sensory process-
and/or biological mediators of the behavioral effects of the ing atypicalities. Yet, results highlighted a clear gap in
interventions, which may inform the development of new translating knowledge about these precursors into preemp-
and more efficient interventions. The inclusion of neurosci- tive interventions. Early interventions, consisting mainly of
ence techniques such as electroencephalography and eye- behavioral parent-training programs, showed some effective-
tracking in the protocols of RCTs is an important strategy ness in reducing ADHD symptoms and improving some
in this direction. Crucially, studies need to specify as out- aspects of executive function, and may be the most appro-
comes the proximal targets for the interventions tested, but priate treatment model for infants and young children with
also long-term ADHD symptom severity and disability and or at elevated likelihood for developing ADHD. Neverthe-
the likelihood that vulnerable infants and young children less, future work is required to develop the most efficient
34 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry
Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

strategy to identify vulnerable infants and young children Sorgato, Cat~ao, Goodwin, Tye, Groom, Polanczyk
Methodology: Shephard, Zuccolo, Bolton, Groom, Polanczyk
and to improve their developmental trajectories. Project administration: Shephard, Zuccolo
Supervision: Polanczyk
Visualization: Shephard
Accepted March 19, 2021. Writing − original draft: Shephard, Zuccolo, Bolton, Tye, Groom, Polanczyk
Writing − review and editing: Shephard, Zuccolo, Idrees, Godoy, Salomone,
Drs. Shephard, Zuccolo, Polanczyk, Ms. Godoy, and Mr. Cat~ao are with Facul-
Ferrante, Sorgato, Cat~ao, Goodwin, Bolton, Tye, Groom, Polanczyk
dade de Medicina FMUSP, Universidade de S~ ao Paulo, Brazil.Drs. Shephard,
Goodwin, Tye, and Prof. Bolton are with Institute of Psychiatry, Psychology & ORCID
Neuroscience (IoPPN), King’s College London, United Kingdom.Prof. Bolton Elizabeth Shephard, PhD: https://orcid.org/0000-0002-5952-3565
is also with The Maudsley NIHR Biomedical Research Centre in Mental Health, Pedro F. Zuccolo, PhD: https://orcid.org/0000-0002-6340-6097
King’s College London and South London and Maudsley NHS Foundation Iman Idrees, MSc: https://orcid.org/0000-0003-1388-6085
Trust, London, United Kingdom. Ms. Idrees and Dr. Groom are with Institute Priscilla B.G. Godoy, MSc: https://orcid.org/0000-0001-8014-4482
of Mental Health, University of Nottingham, United Kingdom. Dr. Salomone Erica Salomone, PhD: https://orcid.org/0000-0002-8083-5942
and Mss. Ferrante and Sorgato are with the University of Milan-Bicocca, Italy. Luís F.C.C. Cat~ao, MD candidate: https://orcid.org/0000-0003-0269-2891
Amy Goodwin, PhD: https://orcid.org/0000-0003-2585-8452
Drs. Shepard and Zuccolo are co-first authors of this work.
Patrick F. Bolton, PhD, FRCPsych: https://orcid.org/0000-0002-5270-6262
E. Shephard is supported by a postdoctoral fellowship from the S~ ao Paulo Charlotte Tye, PhD: https://orcid.org/0000-0002-8567-9547
Research Foundation (FAPESP; ref: 18/22396-7). L.F.C. Cat~ao is supported by Madeleine J. Groom, PhD: https://orcid.org/0000-0002-5182-518X
a scientific initiation bursary from FAPESP (ref: 19/24819-5). C. Tye is sup- Guilherme V. Polanczyk, MD, PhD: https://orcid.org/0000-0003-2311-3289
ported by the Tuberous Sclerosis Association and the National Institute for
The authors are grateful to the authors of publications who responded to their
Health research (NIHR) Biomedical Research Centre at South London and
requests for summary data for the computation of effect sizes.
Maudsley NHS Foundation Trust and King’s College London. The views
expressed are those of the authors and not necessarily those of the NHS, the Disclosure: Prof. Polanczyk has been in the past 3 years a member of the advi-
NIHR, or the Department of Health and Social Care. G.V. Polanczyk is sup- sory boards of Shire/Takeda and Medice and a speaker for Shire/Takeda,
ported by FAPESP (grant 2016/22455-8) and the National Council for Scientific Novo Nordisk, and Ach e. He has received travel expenses for continuing edu-
and Technological Development (CNPq; grant 310582/2017-2). cation support from Shire/Takeda and royalties from Editora Manole. Drs.
Shephard, Zuccolo, Salomone, Goodwin, Prof. Bolton, Drs. Tye and Groom,
This article is part of a special series devoted to the subject of child and ado-
Mss. Idrees, Godoy, Ferrante, Sorgato, and Mr. Cat~ ao have reported no bio-
lescent attention-deficit/hyperactivity disorder (ADHD). The series covers a
medical financial interests or potential conflicts of interest.
range of topics in the area including genetics, neuroimaging, treatment, and
others. The series was edited by Guest Editor Jonathan Posner, MD, along Correspondence to Elizabeth Shephard, PhD, Instituto de Psiquiatria do Hos-
with Deputy Editor Samuele Cortese, MD, PhD. pital das Clínicas da FMUSP, Universidade de S~ ao Paulo, R. Dr. Ovídio Pires
de Campos, 785 - Cerqueira C esar, S~ao Paulo - SP, 05403-903, Brasil; e-mail:
This work has been prospectively registered: https://www.crd.york.ac.uk/pros
lizzieshephard@usp.br
pero/display_record.php?ID=CRD42020165286.
0890-8567/$36.00/© 2021 American Academy of Child & Adolescent Psychia-
Author Contributions
try. Published by Elsevier Inc. All rights reserved
Conceptualization: Shephard, Bolton, Tye, Groom, Polanczyk
Data curation: Shephard, Zuccolo https://doi.org/10.1016/j.jaac.2021.03.016
Formal analysis: Shephard, Zuccolo, Idrees, Godoy, Salomone, Ferrante,

REFERENCES 11. Sonuga-Barke EJ, Halperin JM. Developmental phenotypes and causal pathways
in attention deficit/hyperactivity disorder: potential targets for early intervention?
1. Posner J, Polanczyk GV, Sonuga-Barke E. Attention-deficit hyperactivity disorder. Lan-
J Child Psychol Psychiatry. 2010;51:368-389. https://doi.org/10.1111/j.1469-
cet. 2020;395:P450-P462. https://doi.org/10.1016/S0140-6736(19)33004-1.
7610.2009.02195.x.
2. Faraone SV, Asherson P, Banaschewski T, et al. Attention-deficit/hyperactivity disorder.
12. Wichstrøm L, Berg-Nielsen TS, Angold A, Egger HL, Solheim E, Sveen TH. Preva-
Nat Rev Dis Primers. 2015;1:15020. https://doi.org/10.1038/nrdp.2015.20.
lence of psychiatric disorders in preschoolers. J Child Psychol Psychiatry. 2012;53:695-
3. Cortese S, Adamo N, Del Giovane C, et al. Comparative efficacy and tolerability of
705. https://doi.org/10.1111/j.1469-7610.2011.02514.x.
medications for attention-deficit hyperactivity disorder in children, adolescents, and
13. Auerbach JG, Berger A, Atzaba-Poria N, et al. Temperament at 7, 12, and 25 months in
adults: a systematic review and network meta-analysis. Lancet Psychiatry. 2018;5:727-
children at familial risk for ADHD. Infant Child Dev. 2008;17:321-338. https://doi.
738. https://doi.org/10.1016/S2215-0366(18)30269-4.
org/10.1002/icd.579.
4. Chang Z, Ghirardi L, Quinn PD, Asherson P, D’Onofrio BM, Larsson H. Risks and
14. Einziger T, Levi L, Zilberman-Hayun Y, et al. Predicting ADHD symptoms in adoles-
benefits of attention-deficit/hyperactivity disorder medication on behavioral and neuro-
cence from early childhood temperament traits. J Abnorm Child Psychol.
psychiatric outcomes: a qualitative review of pharmacoepidemiology studies using
2018;46:265-276. https://doi.org/10.1007/s10802-017-0287-4.
linked prescription databases. Biol Psychiatry. 2019;86:335-343. https://doi.org/
15. Shephard E, Bedford R, Milosavljevic B, et al. Early developmental pathways to child-
10.1016/j.biopsych.2019.04.009.
hood symptoms of attention-deficit hyperactivity disorder, anxiety and autism spectrum
5. Charach A, Fernandez R. Enhancing ADHD medication adherence: challenges and
disorder. J Child Psychol Psychiatry. 2019;60:963-974. https://doi.org/10.1111/
opportunities. Curr Psychiatry Rep. 2013;15:371. https://doi.org/10.1007/s11920-
jcpp.12947. a.
013-0371-6.
16. Shephard E, Fatori D, Mauro LR, et al. Effects of maternal psychopathology and educa-
6. Greydanus DE, Nazeer A, Patel DR. Psychopharmacology of ADHD in pediatrics: cur-
tion level on neurocognitive development in infants of adolescent mothers living in pov-
rent advances and issues. Neuropsychiatr Dis Treat. 2009;5:171. https://doi.org/
erty in Brazil. Biol Psychiatry Cogn Neurosci Neuroimaging. 2019;4:925-934. https://
10.2147/ndt.s4075.
doi.org/10.1016/j.bpsc.2019.05.009. b.
7. Sonuga-Barke EJ, Brandeis D, Cortese S, et al. Nonpharmacological interventions for
17. Sullivan EL, Holton KF, Nousen EK, et al. Early identification of ADHD risk via infant
ADHD: systematic review and meta-analyses of randomized controlled trials of dietary
temperament and emotion regulation: a pilot study. J Child Psychol Psychiatry.
and psychological treatments. Am J Psychiatry. 2013;170:275-289. https://doi.org/
2015;56:949-957. https://doi.org/10.1111/jcpp.12426.
10.1176/appi.ajp.2012.12070991.
18. Willoughby MT, Gottfredson NC, Stifter CA. Observed temperament from ages 6 to
8. MTA Cooperative Group. Moderators and mediators of treatment response for children
36 months predicts parent-and teacher-reported attention-deficit/hyperactivity disorder
with attention-deficit/hyperactivity disorder: the Multimodal Treatment Study of chil-
symptoms in first grade. Dev Psychopathology. 2017;29:107-120. https://doi.org/
dren with Attention-deficit/hyperactivity disorder. Arch Gen Psychiatry.
10.1017/S0954579415001236.
1999;56:1088-1096. https://doi.org/10.1001/archpsyc.56.12.1088.
19. Johnson MH, Gliga T, Jones E, Charman T. Annual research review: infant develop-
9. Green J. Editorial Perspective: delivering autism intervention through development. J
ment, autism, and ADHD−early pathways to emerging disorders. J Child Psychol Psy-
Child Psychol Psychiatry. 2019;60:1353-1356. https://doi.org/10.1111/jcpp.13110.
chiatry. 2015;56:228-247. https://doi.org/10.1111/jcpp.12328.
10. Lord C, Brugha TS, Charman T, et al. Autism spectrum disorder. Nat Rev Dis Primers.
20. Visser JC, Rommelse NN, Greven CU, Buitelaar JK. Autism spectrum disorder and
2020;6:1-23. https://doi.org/10.1038/s41572-019-0138-4.
attention-deficit/hyperactivity disorder in early childhood: a review of unique and

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 35


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

shared characteristics and developmental antecedents. Neurosci Biobehav Rev. 45. Berger A, Alyagon U, Hadaya H, Atzaba-Poria N, Auerbach JG. Response inhibition in
2016;65:229-263. https://doi.org/10.1016/j.neubiorev.2016.03.019. preschoolers at familial risk for attention deficit hyperactivity disorder: a behavioral and
21. Athanasiadou A, Buitelaar JK, Brovedani P, et al. Early motor signs of attention-deficit electrophysiological stop-signal study. Child Dev. 2013;84:1616-1632. https://doi.org/
hyperactivity disorder: a systematic review. Eur Child Adolesc Psychiatry. 2020;29:903- 10.1111/cdev.12072.
916. https://doi.org/10.1007/s00787-019-01298-5. 46. Bora S, Pritchard VE, Chen Z, Inder TE, Woodward LJ. Neonatal cerebral morphome-
22. Kostyrka-Allchorne K, Wass SV, Sonuga-Barke EJ. Research review: do parent ratings try and later risk of persistent inattention/hyperactivity in children born very preterm.
of infant negative emotionality and self-regulation predict psychopathology in child- J Child Psychol Psychiatry. 2014;55:828-838. https://doi.org/10.1111/jcpp.12200.
hood and adolescence? A systematic review and meta-analysis of prospective longitudi- 47. Breaux RP, Griffith SF, Harvey EA. Preschool neuropsychological measures as predic-
nal studies. J Child Psychol Psychiatry. 2020;61:401-416. https://doi.org/10.1111/ tors of later attention deficit hyperactivity disorder. J Abnorm Child Psychol.
jcpp.13144. 2016;44:1455-1471. https://doi.org/10.1007/s10802-016-0140-1.
23. Wells GA, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assess- 48. Bron TI, Van Rijen EH, Van Abeelen AM, Lambregtse-van den Berg MP. Develop-
ing the quality of nonrandomised studies in meta-analysis. 2004. Accessed January 30, ment of regulation disorders into specific psychopathology. Infant Ment Health J.
2020. http://www,ohri.ca/programs/clinical_epidemiology.oxford.htm. 2012;33:212-221. https://doi.org/10.1002/imhj.21325.
24. Sterne JAC, Savovic J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in 49. Bundgaard AKF, Asmussen J, Pedersen NS, Bilenberg N. Disturbed sleep and activity
randomised trials. BMJ. 2019;366:14898. https://doi.org/10.1136/bmj.l4898. in toddlers with early signs of attention deficit hyperactivity disorder (ADHD). J Sleep
25. Sterne JA, Hernan MA, Reeves BC, et al. ROBINS-I: a tool for assessing risk of bias in Res. 2018;27:e12686. https://doi.org/10.1111/jsr.12686.
non-randomised studies of interventions. BMJ. 2016;355:i4919. https://doi.org/ 50. Butcher PR, Van Braeckel K, Bouma A, Einspieler C, Stremmelaar EF, Bos AF. The
10.1136/bmj.i4919. quality of preterm infants’ spontaneous movements: an early indicator of intelligence
26. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, and behavior at school age. J Child Psychol Psychiatry. 2009;50:920-930. https://doi.
Austria: R Foundation for Statistical Computing; 2020. org/10.1111/j.1469-7610.2009.02066.x.
27. L€udecke D. esc: Effect Size Computation for Meta Analysis (Version 0.5.1); 2019. 51. Campbell SB, Ewing LJ, Breaux AM, Szumowski EK. Parent-referred problem three-
https://CRAN.R-project.org/package=esc. Accessed May 20, 2020. year-olds: follow-up at school entry. J Child Psychol Psychiatry. 1986;27:473-488.
28. Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw. https://doi.org/10.1111/j.1469-7610.1986.tb00635.x.
2010;36:1-48. https://doi.org/10.18637/jss.v036.i03. 52. Carlson EA, Jacobvitz D, Sroufe LA. A developmental investigation of inattentiveness
29. Langan D, Higgins JP, Jackson D, et al. A comparison of heterogeneity variance estima- and hyperactivity. Child Dev. 1995;66:37-54. https://doi.org/10.1111/j.1467-
tors in simulated random-effects meta-analyses. Res Synth Methods. 2019;10:83-98. 8624.1995.tb00854.x.
https://doi.org/10.1002/jrsm.1316. 53. Caspi A, Henry B, McGee RO, Moffitt TE, Silva PA. Temperamental origins of child
30. Harrer M, Cuijpers P, Furukawa T, Ebert DD. dmetar: Companion R Package For the and adolescent behavior problems: from age three to age fifteen. Child Dev.
Guide 'Doing Meta-Analysis in R'; 2019. R package version 0.0.9000; http://dmetar. 1995;66:55-68. https://doi.org/10.1111/j.1467-8624.1995.tb00855.x.
protectlab.org. Accessed February 1, 2020. 54. Cheung CHM, Bedford R, Johnson MH, Charman T, Gliga T. Visual search perfor-
31. Abulizi X, Pryor L, Michel G, Melchior M, Van Der Waerden J. EDEN Mother mance in infants associates with later ASD diagnosis. Dev Cogn Neurosci. 2018;29:4-
−Child Cohort Study Group. Temperament in infancy and behavioral and emotional 10. https://doi.org/10.1016/j.dcn.2016.09.003.
problems at age 5.5: the EDEN Mother-Child Cohort. PLoS One. 2017;12 e0171971. 55. Cunningham CE, Boyle MH. Preschoolers at risk for attention-deficit hyperactivity dis-
https://doi.org/10.1371/journal.pone.0171971. order and oppositional defiant disorder: family, parenting, and behavioral correlates. J
32. Agapitou P, Andreou G. Language deficits in ADHD preschoolers. Aust J Learn Diffic. Abnorm Child Psychol. 2002;30:555-569. https://doi.org/10.1023/a:1020855429085.
2008;13:39-49. https://doi.org/10.1080/19404150802093711. 56. de la Osa N, Granero R, Penelo E, Domenech JM, Ezpeleta L. The short and very short
33. Allely CS, Johnson PC, Marwick H, et al. Prediction of 7-year psychopathology from forms of the Children’s Behavior Questionnaire in a community sample of preschoolers.
mother-infant joint attention behaviors: a nested case−control study. BMC Pediatr. Assessment. 2014;21:463-476. https://doi.org/10.1177/1073191113508809.
2013;13:147. https://doi.org/10.1186/1471-2431-13-147. 57. Delobel-Ayoub M, Arnaud C, White-Koning M, et al. Behavioral problems and cogni-
34. Arnett AB, Pennington BF, Willcutt E, et al. A cross-lagged model of the development tive performance at 5 years of age after very preterm birth: the EPIPAGE Study. Pediat-
of ADHD inattention symptoms and rapid naming speed. J Abnorm Child Psychol. rics. 2009;123:1485-1492. https://doi.org/10.1542/peds.2008-1216.
2012;40:1313-1326. https://doi.org/10.1007/s10802-012-9644-5. 58. DeSantis A, Coster W, Bigsby R, Lester B. Colic and fussing in infancy, and sensory
35. Arnett AB, MacDonald B, Pennington BF. Cognitive and behavioral indicators of processing at 3 to 8 years of age. Infant Ment Health J. 2004;25:522-539. https://doi.
ADHD symptoms prior to school age. J Child Psychol Psychiatry. 2013;54:1284- org/10.1002/imhj.20025.
1294. https://doi.org/10.1111/jcpp.12104. 59. DeWolfe NA, Byrne JM, Bawden HN. Preschool inattention and impulsivity-hyperac-
36. Astbury J, Gill AO, Bajuk B. Relationship between two-year behavior and neurodeve- tivity: development of a clinic-based assessment protocol. J Atten Disord. 2000;4:80-
lopmental outcome at five years of very low-birthweight survivors. Dev Med Child 90. https://doi.org/10.1177/108705470000400202.
Neurol. 1987;29:370-379. https://doi.org/10.1111/j.1469-8749.1987.tb02491.x. 60. Doherty BR, Charman T, Johnson MH, et al. Visual search and autism symptoms:
37. Auerbach JG, Atzaba-Poria N, Berger A, Landau R. Emerging developmental pathways what young children search for and co-occurring ADHD matter. Dev Sci. 2018;21:
to ADHD: possible path markers in early infancy. Neural Plast. 2004;11 139852. e12661. https://doi.org/10.1111/desc.12661.
https://doi.org/10.1155/np.2004.29. 61. Dougherty LR, Bufferd SJ, Carlson GA, et al. Preschoolers' observed temperament and
38. Auerbach JG, Landau R, Berger A, Arbelle S, Faroy M, Karplus M. Neonatal behavior psychiatric disorders assessed with a parent diagnostic interview. J Clin Child Adolesc
of infants at familial risk for ADHD. Infant Behav Dev. 2005;28:220-224. https://doi. Psychol. 2011;40:295-306. https://doi.org/10.1080/15374416.2011.546046.
org/10.1016/j.infbeh.2004.12.002. 62. DuPaul GJ, McGoey KE, Eckert TL, VanBrakle J. Preschool children with attention-
39. Bates JE, Maslin CA, Frankel KA. Attachment security, mother-child interaction, and deficit/hyperactivity disorder: impairments in behavioral, social, and school functioning.
temperament as predictors of behavior-problem ratings at age three years. Monogr Soc J Am Acad Child Adolesc Psychiatry. 2001;40:508-515. https://doi.org/10.1097/
Res Child Dev 1985;167-193. 00004583-200105000-00009.
40. Becker K, Blomeyer D, El-Faddagh M, et al. From regulatory problems in infancy to 63. Elberling H, Linneberg A, Olsen EM, et al. Infancy predictors of hyperkinetic and per-
attention-deficit/hyperactivity disorder in childhood: a moderating role for the dopa- vasive developmental disorders at ages 5−7 years: results from the Copenhagen Child
mine D4 receptor gene? J Pediatr. 2010;156:798-803. https://doi.org/10.1016/j. Cohort CCC 2000. J Child Psychol Psychiatry. 2014;55:1328-1335. https://doi.org/
jpeds.2009.12.005. 10.1111/jcpp.12256.
41. Bedford R, Gliga T, Hendry A, et al. Infant regulatory function acts as a protective fac- 64. Ezpeleta L, Granero R. Executive functions in preschoolers with ADHD, ODD, and
tor for later traits of autism spectrum disorder and attention deficit/hyperactivity disor- comorbid ADHD-ODD: evidence from ecological and performance-based measures. J
der but not callous unemotional traits. J Neurodev Disord. 2019;11:14. https://doi. Neuropsychol. 2015;9:258-270. https://doi.org/10.1111/jnp.12049.
org/10.1186/s11689-019-9274-0. 65. Floyd RG, Kirby EA. Psychometric properties of measures of behavioral inhibition with
42. Begnoche JP, Brooker RJ, Vess M. EEG asymmetry and ERN: behavioral outcomes in pre- preschool-age children: implications for assessment of children at risk for ADHD. J
schoolers. PLoS One. 2016;11 e0155713. https://doi.org/10.1371/journal.pone.0155713. Atten Disord. 2001;5:79-91. https://doi.org/10.1177/108705470100500202.
43. Ben-Sasson A, Soto TW, Heberle AE, Carter AS, Briggs-Gowan MJ. Early and concur- 66. Foulon S, Pingault JB, Larroque B, Melchior M, Falissard B, C^ote SM. Developmental
rent features of ADHD and sensory over-responsivity symptom clusters. J Atten Disord. predictors of inattention-hyperactivity from pregnancy to early childhood. PLoS One.
2017;21:835-845. https://doi.org/10.1177/1087054714543495. 2015;10 e0125996. . https://doi.org/10.1371/journal.pone.0125996.
44. Berdan LE, Keane SP, Calkins SD. Temperament and externalizing behavior: social 67. Frenkel TI, Koss KJ, Donzella B, et al. ADHD symptoms in post-institutionalized chil-
preference and perceived acceptance as protective factors. Dev Psychol. 2008;44:957. dren are partially mediated by altered frontal EEG asymmetry. J Abnorm Child Psychol.
https://doi.org/10.1037/0012-1649.44.4.957. 2017;45:857-869. https://doi.org/10.1007/s10802-016-0208-y.

36 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

68. Frick MA, Forslund T, Brocki KC. Can reactivity and regulation in infancy predict 89. Jacobson LA, Schneider H, Mahone EM. Preschool inhibitory control predicts ADHD
inattentive and hyperactive/impulsive behavior in 3-year-olds? Dev Psychopathol. group status and inhibitory weakness in school. Arch Clin Neuropsychol.
2019;31:619-629. https://doi.org/10.1017/S0954579418000160. a. 2018;33:1006-1014. https://doi.org/10.1093/arclin/acx124. b.
69. Frick MA, Bohlin G, Hedqvist M, Brocki KC. Temperament and cognitive regulation 90. Jacobvitz D, Sroufe LA. The early caregiver-child relationship and attention-deficit dis-
during the first 3 years of life as predictors of inattention and hyperactivity/impulsivity order with hyperactivity in kindergarten: a prospective study. Child Dev 19871496-
at 6 years. J Atten Disord. 2019;23:1291-1302. https://doi.org/10.1177/ 1504. https://doi.org/10.1111/j.1467-8624.1987.tb03862.x.
1087054718804342. b. 91. Jaspers M, de Winter AF, Buitelaar JK, Verhulst FC, Reijneveld SA, Hartman CA.
70. Friedman AH, Watamura SE, Robertson SS. Movement−attention coupling in infancy Early childhood assessments of community pediatric professionals predict autism spec-
and attention problems in childhood. Dev Med Child Neurol. 2005;47:660-665. trum and attention deficit hyperactivity problems. J Abnorm Child Psychol.
https://doi.org/10.1017/S0012162205001350. 2013;41:71-80. https://doi.org/10.1007/s10802-012-9653-4.
71. Gagne JR, Saudino KJ, Asherson P. The genetic etiology of inhibitory control and 92. Johnson P, Ahamat B, McConnachie A, et al. Motor activity at age one year does not
behavior problems at 24 months of age. J Child Psychol Psychiatry. 2011;52:1155- predict ADHD at seven years. Int J Methods Psychiatr Res. 2014;23:9-18. https://doi.
1163. https://doi.org/10.1111/j.1469-7610.2011.02420.x. org/10.1002/mpr.1436.
72. Gagne JR, Chang CN, Fang H, Spann C, Kwok OM. A multimethod study of inhibi- 93. Johnson S, Kochhar P, Hennessy E, Marlow N, Wolke D, Hollis C. Antecedents of
tory control and behavioral problems in preschoolers. Infant Child Dev. 2019;28: attention-deficit/hyperactivity disorder symptoms in children born extremely preterm. J
e2115. https://doi.org/10.1002/icd.2115. Dev Behav Pediatr. 2016;37:285. https://doi.org/10.1097/DBP.0000000000000298.
73. Gaspardo CM, Cassiano RG, Gracioli SM, Furini GC, Linhares MBM. Effects of neo- 94. Kidwell KM, Nelson TD, Nelson JM, Espy KA. A longitudinal study of maternal and
natal pain and temperament on attention problems in toddlers born preterm. J Pediatr child internalizing symptoms predicting early adolescent emotional eating. J Pediatr
Psychol. 2018;43:342-351. https://doi.org/10.1093/jpepsy/jsx140. Psychol. 2017;42:445-456. https://doi.org/10.1093/jpepsy/jsw085.
74. Girouard PC, Baillargeon RH, Tremblay RE, Glorieux J, LeFebvre F, Robaey P. Devel- 95. Landau R, Avital M, Berger A, et al. Parenting of 7-month-old infants at familial risk for
opmental pathways leading to externalizing behaviors in 5 year olds born before 29 attention deficit/hyperactivity disorder. Infant Ment Health J. 2010;31:141-158.
weeks of gestation. J Dev Behav Pediatr. 1998;19:244-253. https://doi.org/10.1097/ https://doi.org/10.1002/imhj.20249. a.
00004703-199808000-00002. 96. Landau R, Sadeh A, Vassoly P, Berger A, Atzaba-Poria N, Auerbach JG. Sleep patterns
75. Gliga T, Smith TJ, Likely N, Charman T, Johnson MH. Early visual foraging in rela- of 7-week-old infants at familial risk for attention deficit hyperactivity disorder. Infant
tionship to familial risk for autism and hyperactivity/inattention. J Atten Disord. Ment Health J. 2010;31:630-646. https://doi.org/10.1002/imhj.20275. b.
2018;22:839-847. https://doi.org/10.1177/1087054715616490. 97. Landis TD, Garcia AM, Hart KC, Graziano PA. Differentiating symptoms of ADHD
76. Gould JF, Hunt E, Roberts RM, Louise J, Collins CT, Makrides M. Can the Bayley in preschoolers: the role of emotion regulation and executive function. J Atten Disord
Scales of Infant Development at 18 months predict child behavior at 7 years? J Paediatr 2020 2020. https://doi.org/10.1177/1087054719896858.
Child Health. 2019;55:74-81. https://doi.org/10.1111/jpc.14163. 98. Lavigne JV, Gouze KR, Hopkins J, Bryant FB. Multi-domain predictors of attention
77. Guedeney A, Pingault J, Thorr A, Larroque B. The EDEN Mother-Child Cohort Study deficit/hyperactivity disorder symptoms in preschool children: cross-informant differen-
Group. Social withdrawal at 1 year is associated with emotional and behavioral problems ces. Child Psychiatry Hum Dev. 2016;47:841-856. https://doi.org/10.1007/s10578-
at 3 and 5 years: the Eden Mother-Child Cohort Study. Eur Child Adolesc Psychiatry. 015-0616-1.
2014;23:1181-1188. https://doi.org/10.1007/s00787-013-0513-8. 99. Lawson K, Ruff H. Early attention and negative emotionality predict later cognitive and
78. Gunn TE, Tavegia BD, Houskamp BM, et al. Relationship between sensory deficits behavioral function. Int J Behav Dev. 2004;28:157-165. https://doi.org/10.1080/
and externalizing behaviors in an urban, Latino preschool population. J Child Fam 01650250344000361. a.
Stud. 2009;18:653-661. https://doi.org/10.1007/s10826-009-9266-x. 100. Lawson KR, Ruff HA. Early focused attention predicts outcome for children born pre-
79. Gurevitz M, Geva R, Varon M, Leitner Y. Early markers in infants and toddlers for maturely. J Dev Behav Pediatr. 2004;25:399-406. https://doi.org/10.1097/00004703-
development of ADHD. J Atten Disord. 2014;18:14-22. https://doi.org/10.1177/ 200412000-00003. b.
1087054712447858. 101. Lemcke S, Parner ET, Bjerrum M, Thomsen PH, Lauritsen MB. Early development in
80. Gusdorf LM, Karreman A, van Aken MA, Dekovic M, van Tuijl C. The structure of children that are later diagnosed with disorders of attention and activity: a longitudinal
effortful control in preschoolers and its relation to externalizing problems. Br J Dev Psy- study in the Danish National Birth Cohort. Eur Child Adolesc Psychiatry.
chol. 2011;29:612-634. https://doi.org/10.1348/026151010X526542. 2016;25:1055-1066. https://doi.org/10.1007/s00787-016-0825-6.
81. Handen BL, Valdes L. Preschoolers with developmental disabilities: a comparison of an 102. Levine TA, Woodward LJ. Early inhibitory control and working memory abilities of
ADHD and a nonADHD group. J Dev Phys Disabil. 2007;19:579-592. https://doi. children prenatally exposed to methadone. Early Hum Dev. 2018;116:68-75. https://
org/10.1007/s10882-007-9071-7. doi.org/10.1016/j.earlhumdev.2017.11.010.
82. Hatch B, Healey DM, Halperin JM. Associations between birth weight and attention- 103. Lyons-Ruth K, Bureau JF, Riley CD, Atlas-Corbett AF. Socially indiscriminate attach-
deficit/hyperactivity disorder symptom severity: indirect effects via primary neuropsy- ment behavior in the strange situation: convergent and discriminant validity in relation
chological functions. J Child Psychol Psychiatry. 2014;55:384-392. https://doi.org/ to caregiving risk, later behavior problems, and attachment insecurity. Dev Psychopa-
10.1111/jcpp.12168. thol. 2009;21:355. https://doi.org/10.1017/S0954579409000376.
83. Healey DM, Marks DJ, Halperin JM. Examining the interplay among negative emo- 104. Mahone EM, Hoffman J. Behavior ratings of executive function among preschoolers
tionality, cognitive functioning, and attention deficit/hyperactivity disorder symptom with ADHD. Clin Neuropsychol. 2007;21:569-586. https://doi.org/10.1080/
severity. J Int Neuropsychol Soc. 2011;17:502-510. https://doi.org/10.1017/ 13854040600762724.
S1355617711000294. 105. Mahone EM, Crocetti D, Ranta ME, et al. A preliminary neuroimaging study of pre-
84. Houwen S, van der Veer G, Visser J, Cantell M. The relationship between motor per- school children with ADHD. Clin Neuropsychol. 2011;25:1009-1028. https://doi.org/
formance and parent-rated executive functioning in 3-to 5-year-old children: what is 10.1080/13854046.2011.580784.
the role of confounding variables? Hum Mov Sci. 2017;53:24-36. https://doi.org/ 106. Mariana M, Barkley RA. Neuropsychological and academic functioning in preschool
10.1016/j.humov.2016.12.009. children with attention deficit hyperactivity disorder. Dev Neuropsychol. 1997;13:111-
85. Hutchison AK, Hunter SK, Wagner BD, Calvin EA, Zerbe GO, Ross RG. Diminished 129. https://doi.org/10.1080/87565649709540671.
infant P50 sensory gating predicts increased 40-month-old attention, anxiety/depres- 107. Marks DJ, Berwid OG, Santra A, Kera EC, Cyrulnik SE, Halperin JM. Neuropsycho-
sion, and externalizing symptoms. J Atten Disord. 2017;21:209-218. https://doi.org/ logical correlates of ADHD symptoms in preschoolers. Neuropsychology.
10.1177/1087054713488824. 2005;19:446. https://doi.org/10.1037/0894-4105.19.4.446.
86. Hwang-Gu SL, Chen YC, Liang SHY, et al. Exploring the variability in reaction times 108. Mathiesen KS, Sanson A. Dimensions of early childhood behavior problems: stability
of preschoolers at risk of attention-deficit/hyperactivity disorder: an ex-Gaussian analy- and predictors of change from 18 to 30 months. J Abnorm Child Psychol. 2000;28:15-
sis. J Abnorm Child Psychol. 2019;47:1315-1326. https://doi.org/10.1007/s10802- 31. https://doi.org/10.1023/a:1005165916906.
018-00508-z. 109. McGee R, Partridge F, Williams S, Silva PA. A twelve-year follow-up of preschool
87. Ilott N, Saudino KJ, Wood A, Asherson P. A genetic study of ADHD and activity level hyperactive children. J Am Acad Child Adolesc Psychiatry. 1991;30:224-232. https://
in infancy. Genes Brain Behav. 2010;9:296-304. https://doi.org/10.1111/j.1601- doi.org/10.1097/00004583-199103000-00010.
183X.2009.00560.x. 110. McLaughlin KA, Fox NA, Zeanah CH, Sheridan MA, Marshall P, Nelson CA. Delayed
88. Jacobson LA, Crocetti D, Dirlikov B, et al. Anomalous brain development is evident in maturation in brain electrical activity partially explains the association between early
preschoolers with attention-deficit/hyperactivity disorder. J Int Neuropsychol Soc. environmental deprivation and symptoms of attention-deficit/hyperactivity disorder.
2018;24:531-539. https://doi.org/10.1017/S1355617718000103. a. Biol Psychiatry. 2010;68:329-336. https://doi.org/10.1016/j.biopsych.2010.04.005.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 37


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

111. Meeuwsen M, Perra O, van Goozen SH, Hay DF. Informants’ ratings of activity level externalizing behavior problems in young children. J Child Psychol Psychiatry.
in infancy predict ADHD symptoms and diagnoses in childhood. Dev Psychopathol. 2018;59:1044-1051. https://doi.org/10.1111/jcpp.12975.
2019;31:1255-1269. https://doi.org/10.1017/S0954579418000597. 133. Peterson ER, Dando E, D’Souza S, et al. Can infant temperament be used to predict
112. Melegari MG, Nanni V, Lucidi F, Russo PM, Donfrancesco R, Cloninger CR. Tempera- which toddlers are likely to have increased emotional and behavioral problems? Early
mental and character profiles of preschool children with ODD, ADHD, and anxiety disor- Educ Dev. 2018;29:435-449. https://doi.org/10.1080/10409289.2018.1457391.
ders. Compr Psychiatry. 2015;58:94-101. https://doi.org/10.1016/j.comppsych.2015.01.001. 134. Peyre H, Galera C, Waerden J, Van Der Hoertel N, Bernard JY. Relationship between
113. Mendez JL, Fogle LM. Parental reports of preschool children's social behavior: relations early language skills and the development of inattention/hyperactivity symptoms during
among peer play, language competence, and problem behavior. J Psychoeduc Assess. the preschool period: results of the EDEN Mother-Child Cohort. BMC Psychiatry.
2002;20:370-385. https://doi.org/10.1177/073428290202000405. 2016;16:380. https://doi.org/10.1186/s12888-016-1091-3.
114. Mikoteit T, Brand S, Perren S, et al. Visually detected NREM-S2 sleep spindle-density 135. Prior M, Leonard A, Wood G. A comparison study of preschool children diagnosed as
at age five predicted prosocial behavior positively and hyperactivity scores negatively at hyperactive. J Pediatr Psychol. 1983;8:191-207. https://doi.org/10.1093/jpepsy/
age nine. Sleep Med. 2018;48:101-106. https://doi.org/10.1016/j.sleep.2018.03.028. 8.2.191.
115. Miller CJ, Miller SR, Healey DM, Marshall K, Halperin JM. Are cognitive control and 136. Rabinovitz BB, O'Neill S, Rajendran K, Halperin JM. Temperament, executive control,
stimulus-driven processes differentially linked to inattention and hyperactivity in pre- and attention-deficit/hyperactivity disorder across early development. J Abnorm Psy-
schoolers? J Clin Child Adolesc Psychol. 2013;42:187-196. https://doi.org/10.1080/ chol. 2016;125:196. https://doi.org/10.1037/abn0000093.
15374416.2012.759116. 137. Rajendran K, Rindskopf D, O'Neill S, Marks DJ, Nomura Y, Halperin JM. Neuropsy-
116. Miller M, Iosif AM, Young GS, Hill MM, Ozonoff S. Early detection of ADHD: chological functioning and severity of ADHD in early childhood: a four-year cross-
insights from infant siblings of children with autism. J Clin Child Adolesc Psychol. lagged study. J Abnorm Psychol. 2013;122:1179. https://doi.org/10.1037/a0034237.
2018;47:737-744. https://doi.org/10.1080/15374416.2016.1220314. 138. Rajendran K, O'Neill S, Marks DJ, Halperin JM. Latent profile analysis of neuropsy-
117. Miller NV, Degnan KA, Hane AA, Fox NA. Chronis-Tuscano A. Infant temperament chological measures to determine preschoolers' risk for ADHD. J Child Psychol Psychi-
reactivity and early maternal caregiving: independent and interactive links to later child- atry. 2015;56:958-965. https://doi.org/10.1111/jcpp.12434.
hood attention-deficit/hyperactivity disorder symptoms. J Child Psychol Psychiatry. 139. Rints A, McAuley T, Nilsen ES. Social communication is predicted by inhibitory ability
2019;60:43-53. https://doi.org/10.1111/jcpp.12934. a. and ADHD traits in preschool-aged children: a mediation model. J Atten Disord.
118. Miller NV, Hane AA, Degnan KA, Fox NA, Chronis-Tuscano A. Investigation of a 2015;19:901-911. https://doi.org/10.1177/1087054714558873.
developmental pathway from infant anger reactivity to childhood inhibitory control and 140. Roberts JE, Crawford H, Will EA, et al. Infant social avoidance predicts autism but not
ADHD symptoms: interactive effects of early maternal caregiving. J Child Psychol Psy- anxiety in Fragile X syndrome. Front Psychiatry. 2019;10:199. https://doi.org/
chiatry. 2019;60:762-772. https://doi.org/10.1111/jcpp.13047. b. 10.3389/fpsyt.2019.00199.
119. Miller M, Iosif AM, Bell LJ, et al. Can familial risk for ADHD be detected in the first 141. Robson AL, Pederson DR. Predictors of individual differences in attention among low
two years of life? J Clin Child Adolesc Psychol 2020 2020. https://doi.org/10.1080/ birth weight children. J Dev Behav Pediatr. 1997;18:13-21. https://doi.org/10.1097/
15374416.2019.1709196. 00004703-199702000-00004.
120. Miyahara M, Healey DM, Halperin JM. One-week temporal stability of hyperactivity 142. Rogers CE, Anderson PJ, Thompson DK, et al. Regional cerebral development at term
in preschoolers with ADHD during psychometric assessment. Psychiatry Clin Neurosci. relates to school-age social−emotional development in very preterm children. J Am
2014;68:120-126. https://doi.org/10.1111/pcn.12096. Acad Child Adolesc Psychiatry. 2012;51:181-191. https://doi.org/10.1016/j.
121. Nesayan A, Asadi Gandomani R, Movallali G, Dunn W. The relationship between sen- jaac.2011.11.009.
sory processing patterns and behavioral patterns in children. J Occup Ther Sch Early 143. Rohrer-Baumgartner N, Zeiner P, Egeland J, et al. Does IQ influence associations
Interv. 2018;11:124-132. https://doi.org/10.1080/19411243.2018.1432447. between ADHD symptoms and other cognitive functions in young preschoolers? Behav
122. O'Callaghan FV, Al Mamun A, O'Callaghan M, et al. The link between sleep problems Brain Funct. 2014;10:16. https://doi.org/10.1186/1744-9081-10-16.
in infancy and early childhood and attention problems at 5 and 14 years: evidence from 144. Romano E, Tremblay RE, Farhat A, C^ote S. Development and prediction of hyperac-
a birth cohort study. Early Hum Dev. 2010;86:419-424. https://doi.org/10.1016/j.ear- tive symptoms from 2 to 7 years in a population-based sample. Pediatrics.
lhumdev.2010.05.020. 2006;117:2101-2110. https://doi.org/10.1542/peds.2005-0651.
123. Olson SL, Bates JE, Sandy JM, Lanthier R. Early developmental precursors of external- 145. Rosch KS, Crocetti D, Hirabayashi K, Denckla MB, Mostofsky SH, Mahone EM. Reduced
izing behavior in middle childhood and adolescence. J Abnorm Child Psychol. subcortical volumes among preschool-age girls and boys with ADHD. Psychiatry Res Neuro-
2000;28:119-133. https://doi.org/10.1023/a:1005166629744. imaging. 2018;271:67-74. https://doi.org/10.1016/j.pscychresns.2017.10.013.
124. Otterman DL, Koopman-Verhoeff ME, White TJ, Tiemeier H, Bolhuis K, Jansen PW. 146. Ruff HA, Lawson KR, Parrinello R, Weissberg R. Long-term stability of individual dif-
Executive functioning and neurodevelopmental disorders in early childhood: a prospec- ferences in sustained attention in the early years. Child Dev. 1990;61:60-75.
tive population-based study. Child Adolesc. 2019;13:38. https://doi.org/10.1186/ 147. Sanson A, Smart D, Prior M, Oberklaid F. Precursors of hyperactivity and aggression.
s13034-019-0299-7. J Am Acad Child Adolesc Psychiatry. 1993;32:1207-1216. https://doi.org/10.1097/
125. Overgaard KR, Aase H, Torgersen S, et al. Continuity in features of anxiety and atten- 00004583-199311000-00014.
tion deficit/hyperactivity disorder in young preschool children. Eur Child Adolesc Psy- 148. Espírito Santo JLD, Portuguez MW, Nunes ML. Cognitive and behavioral status of low
chiatry. 2014;23:743-752. https://doi.org/10.1007/s00787-014-0538-7. birth weight preterm children raised in a developing country at preschool age. Jornal de
126. Papageorgiou KA, Smith TJ, Wu R, Johnson MH, Kirkham NZ, Ronald A. Individual Pediatr. 2009;85:35-41. https://doi.org/10.2223/JPED.1859.
differences in infant fixation duration relate to attention and behavioral control in child- 149. Saudino KJ, Wang M, Flom M, Asherson P. Genetic and environmental links between
hood. Psychol Sci. 2014;25:1371-1379. https://doi.org/10.1177/0956797614531295. motor activity level and attention problems in early childhood. Dev Sci. 2018;21:
127. Pappa I, Mileva-Seitz VR, Szekely E, et al. DRD4 VNTRs, observed stranger fear in e12630. https://doi.org/10.1111/desc.12630.
preschoolers and later ADHD symptoms. Psychiatry Res. 2014;220:982-986. https:// 150. Schmid G, Wolke D. Preschool regulatory problems and attention-deficit/hyperactivity
doi.org/10.1016/j.psychres.2014.09.004. and cognitive deficits at school age in children born at risk: different phenotypes of dys-
128. Pauli-Pott U, Schloß S, Becker K. Maternal responsiveness as a predictor of self-regula- regulation? Early Hum Dev. 2014;90:399-405. https://doi.org/10.1016/j.earlhum-
tion development and attention-deficit/hyperactivity symptoms across preschool ages. dev.2014.05.001.
Child Psychiatry Hum Dev. 2018;49:42-52. https://doi.org/10.1007/s10578-017- 151. Schneider H, Ryan M, Mahone EM. Parent versus teacher ratings on the BRIEF-
0726-z. preschool version in children with and without ADHD. Child Neuropsychol.
129. Pauli-Pott U, Schloß S, Heinzel-Gutenbrunner M, Becker K. Multiple causal pathways 2020;26:113-128. https://doi.org/10.1080/09297049.2019.1617262.
in attention-deficit/hyperactivity disorder−do emerging executive and motivational 152. Scott N, Blair PS, Emond AM, et al. Sleep patterns in children with ADHD: a popula-
deviations precede symptom development? Child Neuropsychol. 2019;25:179-197. tion-based cohort study from birth to 11 years. J Sleep Res. 2013;22:121-128. https://
https://doi.org/10.1080/09297049.2017.1380177. doi.org/10.1111/j.1365-2869.2012.01054.x.
130. Perren S, Stadelmann S, Von Wyl A, Von Klitzing K. Pathways of behavioral and emo- 153. Seguin JR, Parent S, Tremblay RE, Zelazo PD. Different neurocognitive functions reg-
tional symptoms in kindergarten children: what is the role of pro-social behavior? Eur ulating physical aggression and hyperactivity in early childhood. J Child Psychol Psychi-
Child Adolesc Psychiatry. 2007;16:209-214. https://doi.org/10.1007/s00787-006- atry. 2009;50:679-687. https://doi.org/10.1111/j.1469-7610.2008.02030.x.
0588-6. 154. Silberg JL, Gillespie N, Moore AA, et al. Shared genetic and environmental influences
131. Perrin HT, Heller NA, Loe IM. School readiness in preschoolers with symptoms of on early temperament and preschool psychiatric disorders in Hispanic twins. Twin Res
attention-deficit/hyperactivity disorder. Pediatrics. 2019;144 e20190038. . https://doi. Hum Genet. 2015;18:171-178. https://doi.org/10.1017/thg.2014.88.
org/10.1542/peds.2019-0038. 155. Silverman IW, Ragusa DM. A short-term longitudinal study of the early development
132. Petersen IT, Hoyniak CP, Bates JE, Staples AD, Molfese DL. A longitudinal, within- of self-regulation. J Abnorm Child Psychol. 1992;20:415-435. https://doi.org/
person investigation of the association between the P3 ERP component and 10.1007/BF00918985.

38 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
EARLY PRECURSORS AND INTERVENTIONS IN ADHD

156. Sj€owall D, Bohlin G, Rydell AM, Thorell LB. Neuropsychological deficits in pre- 177. Capodieci A, Gola ML, Cornoldi C, Re AM. Effects of a working memory training
school as predictors of ADHD symptoms and academic achievement in late ado- program in preschoolers with symptoms of attention-deficit/hyperactivity disorder. J Clin
lescence. Child Neuropsychol. 2017;23:111-128. https://doi.org/10.1080/ Exp Neuropsychol. 2018;40:17-29. https://doi.org/10.1080/13803395.2017.1307946.
09297049.2015.1063595. 178. Christakis DA, Zimmerman FJ, Garrison MM. Effect of block play on language acqui-
157. Skogan AH, Zeiner P, Egeland J, et al. Inhibition and working memory in young pre- sition and attention in toddlers: a pilot randomized controlled trial. Arch Pediatr Ado-
school children with symptoms of ADHD and/or oppositional-defiant disorder. Child lesc Med. 2007;161:9670-9971. https://doi.org/10.1001/archpedi.161.10.967.
Neuropsychol. 2014;20:607-624. https://doi.org/10.1080/09297049.2013.838213. 179. Cohen SCL, Harvey DJ, Shields RH, et al. Effects of yoga on attention, impulsivity,
158. Skogan AH, Zeiner P, Egeland J, Urnes AG, Reichborn-Kjennerud T, Aase H. Parent and hyperactivity in preschool-aged children with attention-deficit hyperactivity disor-
ratings of executive function in young preschool children with symptoms of attention- der symptoms. J Dev Behav Pediatr. 2018;39:200-209. https://doi.org/10.1097/
deficit/-hyperactivity disorder. Behav Brain Funct. 2015;11:1-11. https://doi.org/ DBP.0000000000000552.
10.1080/09297049.2013.838213. 180. DuPaul GJ, Kern L, Belk G, et al. Face-to-face versus online behavioral parent training
159. Slopen N, McLaughlin KA, Fox NA, Zeanah CH, Nelson CA. Alterations in neural for young children at risk for ADHD: treatment engagement and outcomes. J Clin
processing and psychopathology in children raised in institutions. Arch Gen Psychiatry. Child Adolesc Psychol. 2018;47(Suppl 1):S369-S383. https://doi.org/10.1080/
2012;69:1022-1030. https://doi.org/10.1001/archgenpsychiatry.2012.444. 15374416.2017.1342544.
160. Smith E, Meyer BJ, Koerting J, et al. Preschool hyperactivity specifically elevates long- 181. Feil EG, Small JW, Seeley JR, et al. Early intervention for preschoolers at risk for atten-
term mental health risks more strongly in males than females: a prospective longitudinal tion-deficit/hyperactivity disorder: preschool first step to success. Behav Disord.
study through to young adulthood. Eur Child Adolesc Psychiatry. 2017;26:123-136. 2016;41:95-106. https://doi.org/10.17988/0198-7429-41.2.95.
https://doi.org/10.1007/s00787-016-0876-8. 182. Franke N, Keown LJ, Sanders MR. An RCT of an online parenting program for parents
161. Tamm L, Brenner SB, Bamberger ME, Becker SP. Are sluggish cognitive tempo symp- of preschool-aged children with ADHD symptoms. J Atten Disord. 2020;24:1716-
toms associated with executive functioning in preschoolers? Child Neuropsychol. 1726. https://doi.org/10.1177/1087054716667598.
2018;24:82-105. https://doi.org/10.1080/09297049.2016.1225707. 183. Halperin JM, Marks DJ, Chacko A, et al. Training Executive, Attention, and Motor Skills
162. Towe-Goodman NR, Stifter CA, Coccia MA, Cox MJ. Family Life Project Key Investi- (TEAMS): a preliminary randomized clinical trial of preschool youth with ADHD. J Abnorm
gators. Interparental aggression, attention skills, and early childhood behavior problems. Child Psychol. 2020;48:375-389. https://doi.org/10.1007/s10802-019-00610-w.
Dev Psychopathol. 2011;23:563. https://doi.org/10.1017/S0954579411000216. 184. Healey D, Healey M. Randomized controlled trial comparing the effectiveness of structured-
163. Treyvaud K, Doyle LW, Lee KJ, et al. Family functioning, burden and parenting stress play (ENGAGE) and behavior management (TRIPLE P) in reducing problem behaviors in
2 years after very preterm birth. Early Hum Dev. 2011;87:427-431. https://doi.org/ preschoolers. Sci Rep. 2019;9:3497. https://doi.org/10.1038/s41598-019-40234-0.
10.1016/j.earlhumdev.2011.03.008. 185. Huang X-X, Ou P, Qian Q-F, et al. [Clinical effect of psychological and behavioral
164. von Stauffenberg C, Campbell SB. Predicting the early developmental course of symp- intervention combined with biofeedback in the treatment of preschool children with
toms of attention deficit hyperactivity disorder. J Appl Dev Psychol. 2007;28:536-552. attention deficit hyperactivity disorder]. Zhongguo dang dai er ke za zhi = Chinese J
https://doi.org/10.1016/j.appdev.2007.06.011. Contemp Pediatr. 2019;21:229-233.
165. Walcott CM, Scheemaker A, Bielski K. A longitudinal investigation of inattention and 186. Jarraya S, Wagner M, Jarraya M, Engel FA. 12 Weeks of kindergarten-based yoga prac-
preliteracy development. J Atten Disord. 2010;14:79-85. https://doi.org/10.1177/ tice increases visual attention, visual-motor precision and decreases behavior of inatten-
1087054709333330. tion and hyperactivity in 5-year-old children. Front Psychol. 2019;10:796. https://doi.
166. Waller R, Hyde LW, Grabell A, Alves M, Olson SL. Differential associations of early org/10.3389/fpsyg.2019.00796.
callous-unemotional, ODD, and ADHD behaviors: multiple pathways to conduct 187. Jones K, Daley D, Hutchings J, Bywater T, Eames C. Efficacy of the Incredible Years
problems. J Child Psychol Psychiatry. 2014;56:657-666. https://doi.org/10.1111/ Basic parent training programme as an early intervention for children with conduct
jcpp.12326. problems and ADHD. Child Care Health Dev. 2007;33:749−756; doi: 10.1111/
167. Wang MV, Aarø LE, Ystrom E. Language delay and externalizing problems in preschool j.1365-2214.2007.00747.x.
age: a prospective cohort study. J Abnorm Child Psychol. 2018;46:923-933. https:// 188. Kaaresen PI, Rønning JA, Ulvund SE, Dahl LB. A randomized, controlled trial of the
doi.org/10.1007/s10802-017-0391-5. effectiveness of an early-intervention program in reducing parenting stress after preterm
168. Wichstrøm L, Penelo E, Rensvik Viddal K, de la Osa N, Ezpeleta L. Explaining the rela- birth. Pediatrics. 2006;118:e9-e19. https://doi.org/10.1542/peds.2005-1491.
tionship between temperament and symptoms of psychiatric disorders from preschool 189. Landis TD, Hart KC, Graziano PA. Targeting self-regulation and academic functioning
to middle childhood: hybrid fixed and random effects models of Norwegian and Span- among preschoolers with behavior problems: are there incremental benefits to including
ish children. J Child Psychol Psychiatry. 2018;59:285-295. https://doi.org/10.1111/ cognitive training as part of a classroom curriculum? Child Neuropsychol.
jcpp.12772. 2019;25:688-704. https://doi.org/10.1080/09297049.2018.1526271.
169. Williams KE, Sciberras E. Sleep and self-regulation from birth to 7 years: A retrospective 190. Mendelsohn AL, Cates CB, Weisleder A, et al. Reading aloud, play, and social-emotional
study of children with and without attention-deficit hyperactivity disorder at 8 to 9 development. Pediatrics. 2018;141 e20173393. https://doi.org/10.1542/peds.2017-3393.
years. J Dev Behav Pediatr. 2016;37:385-394. https://doi.org/10.1097/DBP. 191. Nixon RDV. Changes in hyperactivity and temperament in behaviorally disturbed pre-
0000000000000281. schoolers after parent−child interaction therapy (PCIT). Behav Change. 2001;18:168-
170. Winsler A, Diaz RM, McCarthy EM, Atencio DJ, Chabay LA. Mother-child interac- 176. https://doi.org/10.1375/bech.18.3.168.
tion, private speech, and task performance in preschool children with behavior prob- 192. Papazian O, Alfonso I, Luzondo RJ, Araguez N. [Training of executive function in pre-
lems. J Child Psychol Psychiatry. 1999;40:891-904. school children with combined attention deficit hyperactivity disorder: a prospective,
171. Woodward LJ, Lu Z, Morris AR, Healey DM. Preschool self regulation predicts controlled and randomized trial]. Rev Neurol. 2009;48(Suppl 2):S119-S122.
later mental health and educational achievement in very preterm and typically develop- 193. Re AM, Capodieci A, Cornoldi C. Effect of training focused on executive functions
ing children. Clinical Neuropsychol. 2017;31:404-422. https://doi.org/10.1080/ (attention, inhibition, and working memory) in preschoolers exhibiting ADHD symp-
13854046.2016.1251614. toms. Front Psychol. 2015;6:1161. https://doi.org/10.3389/fpsyg.2015.01161.
172. Yew SGK, O’Kearney R. Language difficulty at school entry and the trajectories of 194. Solomon T, Plamondon A, O’Hara A, et al. A cluster randomized-controlled trial of the
hyperactivity-inattention problems from ages 4 to 11: evidence from a population-rep- impact of the Tools of the Mind curriculum on self-regulation in Canadian pre-
resentative cohort study. J Abnorm Child Psychol. 2017;45:1105-1118. https://doi. schoolers. Front Psychol. 2018;8:2366. https://doi.org/10.3389/fpsyg.2017.02366.
org/10.1007/s10802-016-0241-x. 195. Sonuga-Barke EJ, Daley D, Thompson M, Laver-Bradbury C, Weeks A. Parent-based
173. Youngwirth SD, Harvey EA, Gates EC, Hashim RL, Friedman-Weieneth JL. Neuro- therapies for preschool attention-deficit/hyperactivity disorder: a randomized, controlled
psychological abilities of preschool-aged children who display hyperactivity and/or trial with a community sample. J Am Acad Child Adolesc Psychiatry. 2001;40:402-408.
oppositional-defiant behavior problems. Child Neuropsychol. 2007;13:422-443. https://doi.org/10.1097/00004583-200104000-00008.
https://doi.org/10.1080/13825580601025890. 196. Sonuga-Barke EJS, Thompson M, Daley D, Laver-Bradbury C. Parent training for
174. Abikoff HB, Thompson M, Laver-Bradbury C, et al. Parent training for preschool attention deficit/hyperactivity disorder: is it as effective when delivered as routine rather
ADHD: a randomized controlled trial of specialized and generic programs. J Child Psy- than as specialist care? Br J Clin Psychol. 2004;43:449-457. https://doi.org/10.1348/
chol Psychiatry. 2015;56:618-631. https://doi.org/10.1111/jcpp.12346. 0144665042388973.
175. Barkley RA, Shelton TL, Crosswait C, et al. Multi-method psycho-educational inter- 197. Sonuga-Barke EJ, Barton J, Daley D, et al. A comparison of the clinical effectiveness
vention for preschool children with disruptive behavior: preliminary results at post- and cost of specialised individually delivered parent training for preschool attention-def-
treatment. J Child Psychol Psychiatry. 2000;41:319-332. icit/hyperactivity disorder and a generic, group-based programme: a multi-centre, rand-
176. Bierman KL, Nix RL, Heinrichs BS, et al. Effects of Head Start REDI on children’s omised controlled trial of the New Forest Parenting Programme versus Incredible Years.
outcomes 1 year later in different kindergarten contexts. Child Dev. 2014;85:140-159. Eur Child Adolesc Psychiatry. 2018;27:797-809. https://doi.org/10.1007/s00787-017-
https://doi.org/10.1111/cdev.12117. 1054-3.

Journal of the American Academy of Child & Adolescent Psychiatry www.jaacap.org 39


Volume 00 / Number 00 / & 2021
ARTICLE IN PRESS
SHEPHARD et al.

198. Strayhorn JM, Weidman CS. Reduction of attention deficit and internalizing symptoms 208. Sonuga-Barke EJ. Causal models of attention-deficit/hyperactivity disorder: from com-
in preschoolers through parent-child interaction training. J Am Acad Child Adolesc Psy- mon simple deficits to multiple developmental pathways. Biol Psychiatry.
chiatry. 1989;28:888-896. https://doi.org/10.1097/00004583-198911000-00013. 2005;57:1231-1238. https://doi.org/10.1016/j.biopsych.2004.09.008.
199. Tamm L, Epstein JN, Loren RE, et al. Generating attention, inhibition, and memory: A 209. Dellapiazza F, Michelon C, Vernhet C, et al. Sensory processing related to attention in
pilot randomized trial for preschoolers with executive functioning deficits. J Clin Child children with ASD, ADHD, or typical development: results from the ELENA cohort.
Adolesc Psychol. 2019;48(Suppl 1):S131-S145. https://doi.org/10.1080/15374416. Eur Child Adolesc Psychiatry 2020 2020. https://doi.org/10.1007/s00787-020-01516-
2016.1266645. 5.
200. Volckaert AMS, No€el MP. Training executive function in preschoolers reduce external- 210. Sergeant J. The cognitive-energetic model: an empirical approach to attention-deficit
izing behaviors. Trends Neurosci Educ. 2015;4:37-47. https://doi.org/10.1016/j. hyperactivity disorder. Neurosci Biobehav Rev. 2000;24:7-12. https://doi.org/10.1016/
tine.2015.02.001. s0149-7634(99)00060-3.
201. Weisleder A, Cates CB, Dreyer BP, et al. Promotion of positive parenting and preven- 211. Halperin JM, Schulz KP. Revisiting the role of the prefrontal cortex in the pathophysi-
tion of socioemotional disparities. Pediatrics. 2016;137 e20153239. https://doi.org/ ology of attention-deficit/hyperactivity disorder. Psychol Bull. 2006;132:560. https://
10.1542/peds.2015-3239. doi.org/10.1037/0033-2909.132.4.560.
202. Halperin JM, Marks DJ, Bedard ACV, et al. Training executive, attention, and motor 212. Bellato A, Arora I, Hollis C, Groom MJ. Is autonomic nervous system function atypical
skills: a proof-of-concept study in preschool children with ADHD. J Atten Disord. in attention deficit hyperactivity disorder (ADHD)? A systematic review of the evidence.
2013;17(8):711-721. https://doi.org/10.1177/1087054711435681. Neurosci Biobehav Rev. 2020;108:182-206. https://doi.org/10.1016/j.neubiorev.2019.
203. Healey DM, Halperin JM. Enhancing Neurobehavioral Gains with the Aid of Games 11.001.
and Exercise (ENGAGE): initial open trial of a novel early intervention fostering the 213. Gui A, Mason L, Gliga T, et al. Look duration at the face as a developmental endophe-
development of preschoolers self-regulation. Child Neuropsychol. 2015;21:465-480. notype: elucidating pathways to autism and ADHD. Dev Psychopathol. 2020;32:
https://doi.org/10.1080/09297049.2014.906567. 1303-1322. https://doi.org/10.1017/S0954579420000930.
204. Joekar S, Amiri S, Joekar S, Birashk B, Aghebati A. Effectiveness of a visual attention 214. Hatch B, Iosif AM, Chuang A, de la Paz L, Ozonoff S, Miller M. Longitudinal differen-
training program on the reduction of ADHD symptoms in preschool children at risk ces in response to name among infants developing ASD and risk for ADHD. J Autism
for ADHD in Isfahan: a pilot study. Iran J Psychiatry Behav Sci. 2017;11:e7862. Dev Disord 2020 2020. https://doi.org/10.1007/s10803-020-04369-8.
https://doi.org/10.5812/ijpbs.7862. 215. Miller M, Austin S, Iosif AM, et al. Shared and distinct developmental pathways to
205. van Rhijn T, Osborne C, Ranby S, et al. Peer play in inclusive child care settings: assess- ASD and ADHD phenotypes among infants at familial risk. Dev Psychopathol.
ing the impact of stay, play, and talk, a peer-mediated social skills program. Child Care 2020;32:1323-1334. https://doi.org/10.1017/S0954579420000735.
Pract 2019 2019. https://doi.org/10.1080/13575279.2019.1588707. 216. Stephens RL, Elsayed HE, Reznick JS, Crais ER, Watson LR.Infant attentional behav-
206. Nigg JT, Goldsmith HH, Sachek J. Temperament and attention deficit hyperactivity iors are associated with ADHD symptomatology and executive function in early child-
disorder: the development of a multiple pathway model. J Clin Child Adolesc Psychol. hood [published online ahead of print August 4, 2020]. J Atten Disord. doi:10.1177/
2004;33:42-53. https://doi.org/10.1207/S15374424JCCP3301_5. 1087054720945019.
207. Nigg JT, Sibley MH, Thapar A, Karalunas SL. Development of ADHD: etiology, het- 217. Elliott JH, Synnot A, Turner T, et al. Living systematic review: 1. Introduction—the
erogeneity, and early life course. Annu Rev Psychol. 2020;2:559-583. https://doi.org/ why, what, when, and how. J Clin Epidemiol. 2017;91:23-30. https://doi.org/
10.1146/annurev-devpsych-060320-093413. 10.1016/j.jclinepi.2017.08.010.

40 www.jaacap.org Journal of the American Academy of Child & Adolescent Psychiatry


Volume 00 / Number 00 / & 2021

You might also like