Journal of Child Psychology and Psychiatry **:* (2014), pp **–** doi:10.1111/jcpp.

12303

Neonatal jaundice and increased risk of

attention-deficit hyperactivity disorder:
a population-based cohort study
Chang-Ching Wei,1,2,* Chun-Hung Chang,3,4,5,* Cheng-Li Lin,6 Shih-Ni Chang,7
Tsai-Chung Li,8,9 and Chia-Hung Kao10,11
1
Department of Pediatrics, China Medical University Hospital; 2College of Medicine, China Medical University;
3
Department of Psychiatry, China Medical University Hospital; 4China Medical University; 5Sunshine Psychiatric
Hospital; 6Management Office for Health Data, China Medical University Hospital; 7Department of Medical Research,
Taichung Veterans General Hospital; 8Department of Healthcare Administration, College of Health Science, Asia
University; 9Graduate Institute of Biostatistics, College of Management, China Medical University; 10Department of
Nuclear Medicine and PET Center, China Medical University Hospital; 11Graduate Institute of Clinical Medical
Science and School of Medicine, College of Medicine, China Medical University, Taichung, Taiwan

Background: Previous studies have posited conflicting results regarding the relationship between neonatal jaundice
and the subsequent risk of attention-deficit hyperactivity disorder (ADHD). We therefore performed a large
population study with a defined neonatal jaundice cohort to investigate the incidence and risk of physician-diag-
nosed ADHD in Taiwan. Methods: From 2000 to 2004, 24,950 neonatal jaundice cases and 69,964 matched
nonjaundice controls were identified. At the end of 2008, the incidence rate and hazard ratios (HRs) of
physician-diagnosed ADHD were calculated. Results: The incidence of ADHD was 2.48-fold greater in the jaundice
cohort than in the nonjaundice cohort (3.84 vs. 1.51 per 100,000 person-years) in the study period. The HR of ADHD
was substantially greater for male, preterm, and low-birth-weight infants with neonatal jaundice. The risk of
developing ADHD in the jaundice cohort was greater after a diagnosis of neonatal jaundice for more than 6 years (HR:
2.64; 95% confidence interval: 2.13–3.28). The risk of ADHD increased for neonates with higher serum bilirubin
levels requiring phototherapy and with longer admission days. Conclusion: Neonates with jaundice are at high risk
for developing physician-diagnosed ADHD during their growth period. A risk alert regarding neurologic consequences
is urgently required after a neonatal jaundice diagnosis. Additional studies should be conducted to clarify
the pathogenesis of these relationships. Keywords: Neonatal jaundice, attention-deficit hyperactivity disorder,
population-based cohort study.

Up to 85% of newborns develop physiologic jaun-

Introduction
dice, which is mostly caused by the reduced hepatic
Attention-deficit hyperactivity disorder (ADHD),
enzyme activity of glucuronyl transferase, which
which is one of the most common childhood neuro-
turns bilirubin into water-soluble bilirubin glucuro-
developmental disorders, is characterized by symp-
nides for elimination in the first few days of life
toms of impulsivity, inattention, and hyperactivity
(Maisels, 1982; Dennery, Seidman, & Stevenson,
(Froehlich et al., 2007). Although the causes of
2001). Although neonatal jaundice is generally
ADHD are not well understood, both genetic and
benign, severe hyperbilirubinemia can cause acute
environmental factors are thought to play roles
and chronic bilirubin encephalopathy (kernicterus),
(Swanson et al., 1998, 2007). Certain adverse events
which is characterized by choreoathetotic cerebral
during the perinatal period, such as preeclampsia,
palsy, hearing loss, and paralysis of the upward gaze
maternal cigarette and alcohol consumption, and
(Watchko, 2006 and Watchko & Tiribelli, 2013).
abdominal trauma during pregnancy, have been
Bilirubin is not neurotoxic per se, but it can damage
reported to be associated with symptoms of ADHD
neurons when its concentrations exceed the pro-
(Golmirzaei et al., 2013; Jangaard, Fell, Dodds, &
tein-binding capacities. This neuronal damage is
Allen, 2008; Swanson et al., 1998, 2007). Moreover,
followed by the aggregation and precipitation of
previous studies have examined the relationship
bilirubin crystals in neurons. For decades, inter-
between ADHD and certain postnatal events that
ventions, such as phototherapy or exchange
are associated with potential neurologic damage,
transfusion, have been widely used to prevent bili-
including neonatal jaundice, childhood head
rubin-induced brain damage. Despite efforts to
trauma, and epilepsy (Golmirzaei et al., 2013; Jang-
eliminate bilirubin encephalopathy, subtle neurode-
aard et al., 2008; Kuzniewicz, Escobar, & Newman,
velopmental disabilities, such as ADHD, have been
2009).
reported to be associated with neonatal jaundice,
even for lesser degrees of jaundice (Jangaard et al.,
*Contributed equally in this manuscript. 2008). However, this previous study had several
Conflict of interest statement: No conflicts declared.

© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA
2 Chang-Ching Wei et al.
limitations, including a small sample size and the
lack of a control for potential confounding factors in
the statistical analyses, and these should be noted
when interpreting these results. Besides the selective
basal ganglion damage in kernicterus, it is unknown
if neonatal jaundice contributes to the development
of ADHD and its associated impairments of the
frontostriatal network (Arnsten & Rubia, 2012; Duerr
& Ahdab-Barmada, 2000; Kernicterus, 2012; Scas-
sellati, Bonvicini, Faraone, & Gennarelli, 2012;
Watchko, 2006; Watchko & Tiribelli, 2013). In this
nationwide, population-based cohort study, we
investigated the subsequent risk of physician-diag-
nosed ADHD in children with neonatal jaundice and
its associated interventions for a better understand-
ing of the risk factors and pathogenesis of ADHD.
Methods
Data source
This study used available claims data from Taiwan’s National
Health Insurance, which is a government-run single-payer
modest cost-sharing program, which covers almost all citizens
and with which all hospitals are contracted (Cheng, 2009).
This study used the claims data from children who were
randomly selected from all of the insured population in
Taiwan. This study was exempt from full ethical review
because the database contains surrogate identification for
each person for patient privacy protection and data manage-
ment. The National Health Research Institutes encrypt
patients’ personal information for privacy protection and
provide researchers with anonymous identification numbers
that are associated with relevant claim information. The
International Classification of Disease, Ninth Revision
(ICD-9), was used to define the diagnostic disease codes.
Figure 1 Patient selection flowchart. NHIRD, National Health Insurance Research Database; ADHD, attention-deficit hyperactivity
disorder
© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
Study population
Our study cohort included neonates (<1 month) with a new
diagnosis of jaundice who were treated from 2000 to 2004 and
who were identified by the ICD-9, Clinical Modification
(ICD-9-CM), code 774. We excluded children with missing
information on sex or age. For each jaundice cohort, we
selected three matched controls. The controls were matched
with each case by sex, age, parental urbanization, parental
occupation, and the index date for entering the study. Finally,
24,950 patients with neonatal jaundice and 69,964 infants
without neonatal jaundice were included in this study.
Outcome measurement and comorbidities
We used the incident-user design with follow-up for each child
beginning on the date of first diagnosis in the neonatal
jaundice cohort (Figure 1). The follow-up for the nonneonatal
jaundice cohort began on the matched index date. Both
cohorts were followed until the date of physician-diagnosed
ADHD (ICD-9-CM 314), death, withdrawal from the National
Health Insurance program, or the end of 2008. In Taiwan, a
diagnosis of ADHD was made according to the Diagnostic and
Statistical Manual for Mental Disorders, Fourth edition
(DSM-IV), criteria by a pediatric psychiatrist or psychiatrist.
Only those having at least three consecutive corresponding
diagnoses were designated as having ADHD to improve the
diagnostic accuracy and avoid an overestimation of incidence.
We incorporated inpatient and outpatient diagnosis files to
ascertain the comorbidities during the perinatal period (listed by
ICD-9 code), including other fetal and newborn respiratory
conditions (ICD-9 code 770), infections (ICD-9 code 771), preterm
and low birth weight (ICD-9 codes 764 and 765), and other
birth conditions (ICD-9 codes 761–763 and 767–769) that might
be associated with the development of ADHD. We also incorpo-
rated a glucose-6-phosphate dehydrogenase (G6PD) deficiency
(ICD-9 code 282.2) as a comorbidity because of the increased
incidence and severity of neonatal jaundice in G6PD-deficient
infants, especially when they contact agents that are potentially
hemolytic for G6PD-deficient red cells (Valaes, 1994).
 Neonatal jaundice and ADHD 3

Statistical analysis Table 1 Demographics between children with and without

neonatal jaundice
We compared the baseline characteristics between the neona-
tal jaundice cohort and the nonneonatal jaundice cohort with a Neonatal jaundice
chi-square test. The incidence densities of ADHD between the
two cohorts were calculated. Univariable and multivariable No Yes
Cox proportional-hazard regressions were used to examine the (N = 69,964) (N = 24,950)
effects of neonatal jaundice on the risk of ADHD, as shown by n % n % p-value
HRs with 95% confidence intervals (CIs). The multivariable
model was used to control for age, sex, urbanization, and the Age, years, mean 0.04 0.02 0.02 0.02 <.0001
comorbidities of other fetal and newborn respiratory condi- (SD)a
tions, infections, preterm birth, low birth weight, other birth Sex
conditions, and G6PD deficiency. We estimated the cohort-spe- Girl 29,583 42.3 9,965 39.9 <.0001
cific cumulative incidences by 1 (Kaplan–Meier survival) for Boy 40,381 57.7 14,985 60.1
unadjusted curves and 1 (direct adjusted survival function) Urbanizationb
considering age, sex urbanization, and comorbidities in the 1 (highest) 20,187 28.9 7,219 28.9 .03
Cox model under the assumption, while the differences in the 2 21,176 30.3 7,674 30.8
cumulative incidence curves between the two cohorts were 3 11,867 17.0 4,030 16.2
tested by log-rank tests and likelihood-ratio tests. A two-tailed 4 (lowest) 16,734 23.9 6,027 24.2
P-value less than 0.05 was considered statistically significant. Parental occupation
All statistical analyses were performed with SAS statistical White collar 43,118 61.6 15,423 61.8 .61
software (Version 9.3 for Windows; SAS Institute, Inc., Cary, Blue collar 16,707 23.9 5,976 24.0
NC). Othersc 10,139 14.5 3,551 14.2
Follow-up year, 6.66 1.40 6.58 1.35 <.0001
mean (SD)a
Results Comorbidity
Respiratory 1,163 1.66 1,558 6.24 <.0001
The records for 94,914 people were analyzed in this conditions
study (neonatal jaundice cohort, 24,950; nonneona- Infections 1,493 2.13 3,682 14.8 <.0001
tal jaundice cohort, 69,964). The mean age of the Preterm, low 1,245 1.78 2,300 9.22 <.0001
children with neonatal jaundice was 0.02 (standard birth weight
deviation, 0.02). In the neonatal jaundice group, Other birth 1,251 1.79 2,159 8.65 <.0001
conditions
there were more boys (60.1%) than girls and more G6PD deficiency 44 0.06 297 1.19 <.0001
parents were employed in white-collar positions
(61.8%). The mean follow-up periods in our study Chi-square test, and at-test comparing subjects with neonatal
were 6.58  1.35 years in the neonatal jaundice jaundice and nonneonatal jaundice.
b
The urbanization level was categorized by the population
cohort and 6.66  1.40 years in the nonneonatal
density of the residential area into four levels, with level 1 as
jaundice cohort. We observed that the neonatal the most urbanized and level 4 as the least urbanized.
jaundice cohort was significantly more likely to have c
Other occupations included primarily retired, unemployed, or
other fetal and newborn respiratory conditions (6.24 low-income populations.
vs. 1.66%, p < .0001), infections (14.8 vs. 2.13%, Glucose-6-phosphate dehydrogenase, (G6PD).
p < .0001), preterm birth and low birth weight (9.22
vs. 1.78%, p < .0001), other birth conditions (8.65
vs. 1.79%, p < .0001), and G6PD deficiency (1.19 vs. risk of ADHD (adjusted HR, 2.53; 95% CI, 2.23–2.88)
0.06%, p < .0001) compared with the nonneonatal after controlling for the significant variables in
jaundice controls (Table 1). Table 1. The related adjusted HR of ADHD was
Table 2 shows the ADHD incidence densities for 2.51 (95% CI, 1.82–3.47) and 2.54 (95% CI, 2.21–
both cohorts. Overall, the neonatal jaundice cohort 2.91) for girls and boys, respectively. An interaction
(3.98 per 1,000 person-years) exhibited a signifi- was not found between neonatal jaundice and sex
cantly higher risk of ADHD than did the nonneonatal (p = .89). Regardless of the participants’ clinical
jaundice cohort (1.59 per 1,000 person-years; characteristics, patients with neonatal jaundice
adjusted HR, 2.53; 95% CI, 2.23–2.88) (Table 2). had a higher adjusted HR of ADHD compared with
Figure 2 presents the 9-year cumulative incidence the nonneonatal jaundice cohort.
curves of ADHD by neonatal jaundice status without The multivariable analysis showed that photother-
any adjustment (A) and adjusted for age, sex, apy (adjusted HR, 1.25; 95% CI, 1.07–1.47) and
urbanization, and comorbidities (B). The difference frequent admission for neonatal jaundice (more than
in the cumulative incidence curves of ADHD between 7 times, adjusted HR, 1.81; 95% CI, 1.44–2.28) were
the patients with neonatal jaundice and those significantly associated with a higher risk of ADHD
without neonatal jaundice during the 9 years of (Table 3).
follow-up was pronounced in the unadjusted curves In Table 4, we stratified the follow-up time into
(p < .001) and adjusted curves (p < .001). To esti- ≤6 years or >6 years to analyze the risk of ADHD.
mate the risk of ADHD in the Cox proportional- Compared with the nonneonatal jaundice cohort, the
hazard regression model, the neonatal jaundice neonatal jaundice cohort was associated with an
cohort was associated with a significantly higher increased risk of ADHD, regardless of a follow-up

© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
4 Chang-Ching Wei et al.

Table 2 The risk of attention-deficit hyperactivity disorder in neonates with jaundice compared with those without jaundice
stratified by demographics in Cox proportional-hazard regression model

Nonneonatal jaundice Neonatal jaundice

Event Person-years IR Event Person-years IR Crude HRa (95% CI) Adjusted HRb (95% CI)

Allc 739 465,964 1.59 653 164,077 3.98 2.56 (2.30, 2.84)*** 2.53 (2.23, 2.88)***
Sexd
Girl 121 196,026 0.62 100 65,859 1.52 2.48 (1.90, 3.23)*** 2.51 (1.82, 3.47)***
Boy 618 269,938 2.29 553 98,218 5.63 2.53 (2.26, 2.84)*** 2.54 (2.21, 2.91)***
Comorbidity
Respiratory conditionse
No 721 458,425 1.57 610 153,919 3.96 2.57 (2.31, 2.86)*** 2.54 (2.23, 2.89)***
Yes 18 7,539 2.39 43 10,158 4.23 1.81 (1.05, 3.14)* 2.45 (1.14, 5.26)*
Infections
No 716 456,256 1.57 569 139,611 4.08 2.66 (2.38, 2.97)*** 2.53 (2.22, 2.88)***
Yes 23 9,708 2.37 84 24,466 3.43 1.44 (0.91, 2.29) 2.82 (1.50, 5.28)**
Preterm, low birth weight
No 718 458,007 1.57 569 148,873 3.82 2.49 (2.23, 2.78)** 2.51 (2.20, 2.86)***
Yes 21 7,958 2.64 84 15,204 5.52 2.13 (1.32, 3.44)** 2.83 (1.55, 5.19)***
Other birth conditions
No 720 457,843 1.57 585 149,654 3.91 2.54 (2.28, 2.84)*** 2.52 (2.21, 2.87)***
Yes 19 8,120 2.34 68 14,423 4.71 2.03 (1.22, 3.38)** 2.87 (1.48, 5.56)**
G6PD deficiency
No 739 465,691 1.59 642 162,098 3.96 2.55 (2.29, 2.83)*** 2.52 (2.22, 2.87)***
Yes 0 273 0.00 11 1,979 5.56 – –

IR, incidence rate, per 1,000 person-years.

Crude HRa, relative hazard ratio.
glucose-6-phosphate dehydrogenase, (G6PD).
Adjusted HRb, adjusted hazard ratio.
c
adjusted for age, sex, urbanization, and comorbidities.
d
adjusted for age, urbanization, and comorbidities.
e
adjusted for age, sex, urbanization, and other comorbidities.
*p < .05; **p < .01; ***p < .001.

period less than 6 years (adjusted HR, 2.41; 95% CI,
2.06–2.83) or over 6 years (adjusted HR, 2.76; 95%
CI, 2.23–3.43).
Discussion
We conducted a large, population-based cohort
study to investigate the incidence and subsequent
risk of physician-diagnosed ADHD in children with
neonatal jaundice in an Asian population. The
results showed an increased risk of physician-
diagnosed ADHD among children with neonatal
jaundice, regardless of sex and the coexistence of
other perinatal insults, such as low birth weight,
prematurity, respiratory conditions, or neonatal
infection. Furthermore, the increased risk of
physician-diagnosed ADHD was associated with
neonates with a higher level of bilirubin who required
phototherapy and who had longer admission days.
Only three previous studies of ADHD with
conflicting results have found a causal link with
hyperbilirubinemia. Jangaard et al. (2008) have
investigated the neurologic outcomes of newborns
with neonatal jaundice in a birth cohort in Nova
Scotia, Canada. Although they found no cases of
kernicterus and no increased risk of cerebral palsy,
developmental delay, deafness, visual abnormali-
ties, or autism in infants with neonatal jaundice,
they first reported an increased risk of ADHD in
infants with total serum bilirubin levels ≥19 mg/dl
(adjusted HR, 1.9; 95% CI, 1.1–3.3). Kuzniewicz
et al. (2009) performed another birth cohort study
in Northern California, and they found no associa-
tion between total serum bilirubin levels ≥19 mg/dl
and ADHD diagnoses (adjusted HR, 1.07; 95% CI,
0.90–1.27). Another case–control study with limited
cases aimed to evaluate risk factors for ADHD.
ADHD symptoms were found to be associated with
certain maternal and paternal adverse events, such
as a history of trauma or accident during preg-
nancy, abortion procedures, unintended pregnancy,
or a history of head trauma, but neonatal jaundice
was not included (Golmirzaei et al., 2013). The
limitation of that study was that it was a cross-
sectional study with a limited patient number that
did not provide a better causal link. The findings of
our study, which used the largest population of all
of these studies, supported the findings of Jangaard
et al.
The strength of our study was that it adjusted for
perinatal comorbidities during the perinatal period,
including respiratory conditions, infections, pre-
term birth, and low birth weight, which may
confound the development of ADHD (Golmirzaei
et al., 2013; Silva, Colvin, Hagemann, & Bower,
2013; Singh, Kenney, Ghandour, Kogan, & Lu,
2013) and which have not been adjusted for in
previous studies. Moreover, the ethnicity of the

© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
 Neonatal jaundice and ADHD 5

(A)

(B)

Figure 2 The unadjusted (A) and adjusted (B) cumulative incidence curves of attention-deficit hyperactivity disorder between the cohorts
with (solid line) and without (dashed line) neonatal jaundice

Table 3 Incidence, crude and adjusted hazard ratio of attention-deficit hyperactivity disorder between neonatal jaundice patients
with and without specific treatment

Therapy for jaundice N Event Person-years IR Crude HRa (95% CI) Adjusted HRb (95% CI)

Phototherapy
No 12,008 272 79,144 3.44 1.00 (Reference) 1.00 (Reference)
Yes 12,942 381 84,932 4.49 1.32 (1.13, 1.54)** 1.25 (1.07, 1.47)*
Exchange transfusion
No 24,860 650 163,457 3.98 1.00 (Reference) 1.00 (Reference)
Yes 90 3 619 4.84 1.15 (0.37, 3.58) 1.18 (0.38, 3.66)
Admission days
≤3 18,260 434 119,964 3.62 1.00 (Reference) 1.00 (Reference)
4–6 3,440 91 22,761 4.00 1.10 (0.88, 1.38) 1.20 (0.95, 1.51)
≥7 3,250 128 21,351 5.99 1.65 (1.36, 2.01)** 1.81 (1.44, 2.28)**

IR, incidence rate, per 1,000 person-years.

Crude HRa, relative hazard ratio.
Adjusted HRb, adjusted for age, sex, urbanization, and comorbidities of other fetal and newborn respiratory conditions, infections,
preterm, low birth weight, other birth conditions, and Glucose-6-phosphate dehydrogenase (G6PD) deficiency.
*p < .01; **p < .001.

subjects in the current study was Chinese, which
differed from previous studies that have investi-
gated predominantly Caucasian populations (Jang-
aard et al., 2008; Kuzniewicz et al., 2009). Asian
populations are at a greater risk of neonatal
jaundice (Dennery et al., 2001). Genetic variations,
such as the UGT1A1 gene in the Gly71Arg muta-
tion, are common in Asian populations, thereby
increasing the incidence of severe neonatal hyper-
bilirubinemia (Akaba et al., 1998; Long, Zhang,

© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
6 Chang-Ching Wei et al.
Table 4 Trends of attention-deficit hyperactivity disorder event risk by stratified follow-up years
Nonneonatal jaundice
Follow-up time Event Person-years IR Event
ADHD
≤6 years 472 398,932 1.18 428
>6 years 267 67,031 3.98 225
IR, incidence rate, per 1,000 person-years.
Crude HRa, relative hazard ratio.
Adjusted HRb, adjusted for age, sex, urbanization, and comorbidities of other fetal and newborn respiratory conditions, infections,
preterm, low birth weight, other birth conditions, and Glucose-6-phosphate dehydrogenase (G6PD) deficiency.
*p < .001.
Fang, Luo, & Liu, 2011). Hence, our population
was better for an investigation of the long-term
neurologic outcomes of neonatal jaundice. Further-
more, we investigated the association between
interventions associated with neonatal jaundice
(Maisels, Watchko, & McDonagh, 2012), including
phototherapy and exchange transfusion, and the
subsequent risk of ADHD, which has not been
evaluated in previous studies. We observed that an
increased risk of ADHD was associated with chil-
dren with neonatal jaundice who were treated with
phototherapy and who had longer admission days.
These findings suggest a dose–effect phenomenon
of an increased risk of ADHD and the severity of
hyperbilirubinemia or that phototherapy may be
associated with increased risks of ADHD. Our
findings also showed that the subsequent risk of
ADHD increased with a longer follow-up period in
children with neonatal jaundice. In approximately
40 to 60% of children with ADHD, the symptoms
persist into adulthood (Volkow & Swanson, 2013).
Therefore, the long-term neurologic surveillance
and follow-up of these children are warranted.
To date, the causative pathophysiologic mecha-
nisms for ADHD have not yet been identified. Genes,
pre- and perinatal risks, psychosocial factors, and
environmental toxins have all been considered as
potential risk factors interacting during early devel-
opment to create a neurobiologic susceptibility to
ADHD. Most infants develop physiologic jaundice in
their early life. When serum bilirubin concentrations
exceed protein-binding capacities, bilirubin crys-
tals aggregate and precipitate in neurons. Biliru-
bin-induced neurologic damages have been reported
in areas of the globus pallidus, subthalamic nucleus,
brain-stem nuclei, hippocampal CA2 neurons, and
cerebellar Purkinje’s cells (Duerr & Ahdab-Barmada,
2000; Kernicterus, 2012; Watchko, 2006; Watchko
& Tiribelli, 2013). These neurologic damages might
cause dysfunction of the frontostriatal network
(Arnsten & Rubia, 2012) and dysregulation of the
frontal monoaminergic systems and hypothalamic–
pituitary–adrenal axis (Scassellati et al., 2012),
which were similar to the regions of functional
abnormalities during motor inhibition and Switch
tasks reported in fMRI studies of children with
© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
Neonatal jaundice
Person-years IR Crude HRa (95% CI) Adjusted HRb (95% CI)
142,185 3.01 2.55 (2.24, 2.91)* 2.41 (2.06, 2.83)*
21,892 10.30 2.58 (2.16, 3.08)* 2.76 (2.23, 3.43)*
ADHD (Cubillo et al., 2010). These results also
support our findings. Further trials may benefit from
genetic studies and functional imaging studies to
elucidate the possible mechanisms.
This study had several limitations. First, the
National Health Insurance Research Database did
not provide serum bilirubin levels or genetic and
environmental factors that might affect the risks of
ADHD. However, we used treatment modality and
admission days to represent advanced hyperbiliru-
binemia. Second, the subjects in this study were
Chinese, and, thus, the study results might not be
generalizable to other populations. The East Asian
population is at greater risk of neonatal jaundice
than other populations; therefore, our population
might reflect the risk of ADHD based on neonatal
jaundice. Third, this study might have underesti-
mated children with neonatal jaundice because only
those patients who received medical care were
enrolled. However, those infants might have lower
serum bilirubin levels, and our results showed that
the risk of ADHD increased with those requiring
phototherapy and those having longer admission
days. Last, there is no biomedical laboratory test
that is diagnostic for ADHD. The diagnosis is based
on the observation of a number of behavioral symp-
toms of inattention, impulsivity, and hyperactivity in
different settings and over a certain period. There is a
concern of the association of reliable and valid
measures of ADHD and the clinical diagnosis of
ADHD. Many factors influence clinicians in their
selection of diagnoses other than the behavior of the
child. Thus, different results might have been
obtained if the sample had been assessed by using
structured diagnostic interviews. In Taiwan, a diag-
nosis of ADHD is basically made according to
DSM-IV criteria by pediatric psychiatrists or psychi-
atrists, and, in this study, only those having at least
three consecutive corresponding diagnoses were
designated as having ADHD for better diagnostic
validity.
In conclusion, this study indicated a high subse-
quent risk of ADHD in children with neonatal jaun-
dice, and the risk increased for those requiring
phototherapy and with longer admission days.
Long-term neurologic surveillance and follow-up
 Neonatal jaundice and ADHD 7

are warranted for children with neonatal jaundice.
Substantial efforts should be exerted to clarify the
pathogenesis underlying these relationships.
Acknowledgements
This study was partially supported by the Bureau of
Health Promotion, Taiwan Ministry of Health Welfare,
R.O.C. (Taiwan) (DOH99-HP-1205), study project
grants (DMR-103-018 and DMR-103-020) from China
Medical University Hospital, Taiwan Ministry of
Health and Welfare Clinical Trial and Research Center
of Excellence (MOHW103-TDU-B-212-113002), Health
and welfare surcharge of tobacco products, China
Medical University Hospital Cancer Research Center
of Excellence (MOHW103-TD-B-111-03, Taiwan), and
International Research-Intensive Centers of Excellence
in Taiwan (I-RiCE) (NSC101-2911-I-002-303). The
funders had no role in study design, data collection
and analysis, decision to publish, or preparation of
the manuscript. Chang-Ching Wei and Chun-Hung
Chang conceptualized and designed the study, drafted
the initial manuscript, and approved the final sub-
mitted manuscript. Cheng-Li Lin, Chia-Hung Kao,
and Tsai-Chung Li conducted the initial analysis,
reviewed and revised the manuscript, and approved
the final submitted manuscript. Chia-Hung Kao coor-
dinated and supervised data collection, critically
reviewed the manuscript, and approved the final
submitted manuscript.
Correspondence
Chia-Hung Kao, Graduate Institute of Clinical Medical
Science and School of Medicine, College of Medicine,
China Medical University, No. 2, Yuh-Der Road,
Taichung 404, Taiwan; Email: d10040@mail.cmuh.
org.tw

Key points
• The incidence of ADHD was 2.48-fold greater in the jaundice cohort than in the nonjaundice cohort.
• The risk of developing ADHD was substantially greater for male, preterm, and low-birth-weight infants with
neonatal jaundice.
• The risk of ADHD in the jaundice cohort was greater after a diagnosis of neonatal jaundice for more than
6 years.
• The risk of ADHD increased for neonates with higher serum bilirubin levels requiring phototherapy and with
longer admission days.
• A risk alert regarding neurologic consequences is urgently required after a neonatal jaundice diagnosis.
Additional studies should be conducted to clarify the pathogenesis of these relationships.

Duerr, R.H., & Ahdab-Barmada, M. (2000). Fetal and neonatal

References
Akaba, K., Kimura, T., Sasaki, A., Tanabe, S., Ikegami, T.,
Hashimoto, M., . . . & Hayasaka, K. (1998). Neonatal
hyperbilirubinemia and mutation of the bilirubin uridine
diphosphate-glucuronosyltransferase gene: A common
missense mutation among Japanese, Koreans and
Chinese. Biochemistry and Molecular Biology International,
46, 21–26.
Arnsten, A.F.T., & Rubia, K. (2012). Neurobiological Circuits
Regulating Attention, Cognitive Control, Motivation, and
Emotion: Disruptions in Neurodevelopmental Psychiatric
Disorders. Journal of the American Academy of Child and
Adolescent Psychiatry, 51, 356–367.
Cheng, T.M. (2009). Taiwan’s National Health Insurance
system: High value for the dollar. In K.G.H. Okma & L.
Crivelli (Eds.), Six countries, six reform models: The health
reform experience of Israel, the Netherlands, New Zealand,
Singapore, Switzerland and Taiwan (pp. 71–204). New
Jersey: World Scientific.
Cubillo, A., Halari, R., Ecker, C., Giampietro, V., Taylor, E., &
Rubia, K. (2010). Reduced activation and inter-regional
functional connectivity of fronto-striatal networks in adults
with childhood Attention-Deficit Hyperactivity Disorder
(ADHD) and persisting symptoms during tasks of motor
inhibition and cognitive switching. Journal of Psychiatric
Research, 44, 629–639.
Dennery, P.A., Seidman, D.S., & Stevenson, D.K. (2001). Drug
therapy: Neonatal hyperbilirubinemia. New England Journal
of Medicine, 344, 581–590.
physiology. European: Harwood Academic Publishers.
Froehlich, T.E., Lanphear, B.P., Epstein, J.N., Barbaresi, W.J.,
Katusic, S.K., & Kahn, R.S. (2007). Prevalence, recognition,
and treatment of attention-deficit/hyperactivity disorder in
a national sample of US children. Archives of Pediatrics &
Adolescent Medicine, 161, 857–864.
Golmirzaei, J., Namazi, S., Amiri, S., Zare, S., Rastikerdar, N.,
Hesam, A.A., . . . & Asadi, S. (2013). Evaluation of
attention-deficit hyperactivity disorder risk factors.
International Journal of Pediatrics, 2013, 953103.
Jangaard, K.A., Fell, D.B., Dodds, L., & Allen, A.C. (2008).
Outcomes in a population of healthy term and near-term
infants with serum bilirubin levels of >or=325 micromol/L
(>or=19 mg/dL) who were born in Nova Scotia, Canada,
between 1994 and 2000. Pediatrics, 122, 119–124.
Kernicterus, S.S. (2012). Care of the jaundiced neonate. New
York: McGraw-Hill.
Kuzniewicz, M., Escobar, G.J., & Newman, T.B. (2009). No
association between hyperbilirubinemia and attention-
deficit disorder. Pediatrics, 123, e367–e368.
Long, J., Zhang, S., Fang, X., Luo, Y., & Liu, J. (2011).
Neonatal hyperbilirubinemia and Gly71Arg mutation of
UGT1A1 gene: A Chinese case-control study followed by
systematic review of existing evidence. Acta Paediatrica,
100, 966–971.
Maisels, M.J. (1982). Jaundice in the newborn. Pediatrics in
Review, 3, 305–319.
Maisels, M.J.S.D., Watchko, J.F., & McDonagh, A.F. (2012).
Care of the jaundiced neonate. New York: McGraw-Hill.

© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
8 Chang-Ching Wei et al.
Scassellati, C., Bonvicini, C., Faraone, S.V., & Gennarelli, M.
(2012). Biomarkers and attention-deficit/hyperactivity
disorder: A systematic review and meta-analyses. Journal
of the American Academy of Child and Adolescent
Psychiatry, 51, e1020.
Silva, D., Colvin, L., Hagemann, E., & Bower, C. (2013).
Environmental Risk Factors by Gender Associated With
Attention-Deficit/Hyperactivity Disorder. Pediatrics, 133,
e14–e22.
Singh, G.K., Kenney, M.K., Ghandour, R.M., Kogan, M.D., &
Lu, M.C. (2013). Mental Health Outcomes in US Children
and Adolescents Born Prematurely or with Low Birthweight.
Depression Research and Treatment, 2013, 570743.
Swanson, J.M., Sergeant, J.A., Taylor, E., Sonuga-Barke, E.J.,
Jensen, P.S., & Cantwell, D.P. (1998). Attention-deficit
hyperactivity disorder and hyperkinetic disorder. Lancet,
351, 429–433.
Swanson, J.M., Kinsbourne, M., Nigg, J., Lanphear, B.,
Stefanatos, G.A., Volkow, N., . . . & Wadhwa, P.D., (2007).
Etiologic subtypes of attention-deficit/hyperactivity
© 2014 The Authors. Journal of Child Psychology and Psychiatry © 2014 Association for Child and Adolescent Mental Health.
disorder: brain imaging, molecular genetic and
environmental factors and the dopamine hypothesis.
Neuropsychology Review, 17, 39–59.
Valaes, T. (1994). Severe neonatal jaundice associated
with glucose-6-phosphate dehydrogenase deficiency:
Pathogenesis and global epidemiology. Acta Paediatrica,
83, 58–76.
Volkow, N.D., & Swanson, J.M. (2013). Clinical practice: Adult
attention deficit-hyperactivity disorder. New England
Journal of Medicine, 369, 1935–1944.
Watchko, J.F. (2006). Kernicterus and the molecular
mechanisms of bilirubin-induced CNS injury in newborns.
Neuromolecular Medicine, 8, 513–529.
Watchko, J.F., & Tiribelli, C. (2013). Bilirubin-induced
neurologic damage–mechanisms and management
approaches. New England Journal of Medicine, 369, 2021–
2030.
Accepted for publication: 19 June 2014