European Neuropsychopharmacology (2014) 24, 919–929

www.elsevier.com/locate/euroneuro

Blood serotonin levels in autism spectrum

disorder: A systematic review and
meta-analysis
Stefano Gabrielea,b, Roberto Saccoa,b, Antonio M. Persicoa,b,c,n

a
Unit of Child and Adolescent NeuroPsychiatry, Laboratory of Molecular Psychiatry and Neurogenetics,
University “Campus Bio-Medico”, Via Alvaro del Portillo 21, I-00128 Rome, Italy
b
Department of Experimental Neurosciences, I.R.C.C.S. “Fondazione Santa Lucia”, Rome, Italy
c
Mafalda Luce Center for Pervasive Developmental Disorders, Milan, Italy

Received 29 July 2013; received in revised form 9 January 2014; accepted 12 February 2014

KEYWORDS Abstract
5-HT; Elevated blood serotonin (5-HT) levels were the first biomarker identified in autism research.
Autism; Many studies have contrasted blood 5-HT levels in autistic patients and controls, but different
Biomarker; measurement protocols, technologies, and biomaterials have been used through the years. We
Endophenotype; performed a systematic review and meta-analysis to provide an overall estimate of effect size and
Meta-analysis
between-study heterogeneity, while verifying whether and to what extent different methodolo-
gical approaches influence the strength of this association. Our literature search strategy
identified 551 papers, from which 22 studies providing patient and control blood 5-HT values
were selected for meta-analysis. Significantly higher 5-HT levels in autistic patients compared to
controls were recorded both in whole blood (WB) [O.R.=4.6; (3.1–5.2); P=1.0  10 12], and in
platelet-rich plasma (PRP) [O.R.=2.6 (1.8–3.9); P=2.7  10 7]. Predictably, studies measuring
5-HT levels in platelet-poor plasma (PPP) yielded no significant group difference [O.R.=0.54 (0.2–
2–0); P=0.36]. Altogether, elevated 5-HT blood levels were recorded in 28.3% in WB and 22.5%
in PRP samples of autistic individuals, as reported in 15 and 4 studies, respectively. Studies
employing HPLC vs fluorometric assays yield similar cumulative effect sizes, but the former
display much lower variability. In summary, despite some limitations mainly due to small study
sample sizes, our results significantly reinforce the reliability of elevated 5-HT blood levels as a
biomarker in ASD, providing practical indications potentially useful for its inclusion in multi-
marker diagnostic panels for clinical use.
& 2014 Elsevier B.V. and ECNP. All rights reserved.

n
Corresponding author at: Unit of Child and Adolescent NeuroPsychiatry, Laboratory of Molecular Psychiatry and Neurogenetics, University
“Campus Bio-Medico”, Via Alvaro del Portillo 21, I-00128 Rome, Italy. Tel.: +39 06 2254 19155; fax: +39 06 5017 03333.
E-mail address: a.persico@unicampus.it (A.M. Persico).

http://dx.doi.org/10.1016/j.euroneuro.2014.02.004
0924-977X & 2014 Elsevier B.V. and ECNP. All rights reserved.
920
1. Introduction
The term “autism spectrum disorder” (ASD) defines a com-
plex and heterogeneous group of neurodevelopmental con-
ditions, characterized by impaired social and communication
skills, as well as stereotyped behaviors and restricted
patterns of interests (American Psychiatric Association,
2013). Despite many advances in our understanding of the
neurobiological and developmental processes underlying ASD
(Chugani, 2012; Freitas et al., 2012; State and Levitt, 2011),
our knowledge remains limited and its translational impact
into the clinics is still insufficient. Most importantly, beha-
vioral abnormalities are typically not evident until approxi-
mately 12–18 months of age (Mitchell et al., 2011; Wan et al.,
2013). Furthermore, ASD individuals vary enormously in
clinical presentation, severity, developmental trajectory,
and treatment response. This complexity is spurring an
intensive search to identify biological markers able to aid
clinicians in achieving earlier diagnoses and in predicting
clinical prognosis as well as treatment response (Walsh et al.,
2011). A biomarker can be defined as a biological variable
associated with the disease of interest across and within
individuals, measurable directly in a given patient or in his/
her biomaterials using sensitive and reliable quantitative
procedures. Elevated blood serotonin (5-hydroxytryptamine
or 5-HT) was the first biomarker identified in autism research
(Hanley et al., 1977; Schain and Freedman, 1961), and is still
one of the quantitative traits most consistently associated
with the disease in a sizable subgroup of ASD patients, as
reviewed elsewhere (Anderson et al., 1987; Cook and
Leventhal, 1996; Veenstra-VanderWeele and Blakely, 2012).
In particular, most studies have described significantly higher
mean 5-HT blood levels in approximately 30% autistic indivi-
duals on average, as compared to typically developing
controls. Elevated 5-HT blood levels are seemingly autism-
specific, as they are not present in cognitively impaired
individuals, and are equally distributed in different Pervasive
Developmental Disorder subtypes (i.e., Autistic Disorder vs
Pervasive developmental Disorder Not Otherwise Specified)
(Mulder et al., 2004).
Several studies have addressed the mechanisms under-
lying 5-HT blood elevation in ASD. Physiologically, periph-
eral 5-HT is produced by enterochromaffin cells in the
gut (Gershon, 2004). Approximately 99% of blood 5-HT is
sequestered inside platelets by the antidepressant-sensitive
5-HT transporter (SERT) and then inside vesicles by the
vesicular monoamine transporter (VMAT2) (Zalsman et al.,
2011), while only 1% of total 5-HT remains free in the
plasma and thus exposed to the action of catabolic enzymes
(Anderson et al., 1987). The regulation of SERT activity and
trafficking at the plasma membrane of platelet represents
an important and well investigated mechanism in the
regulation of 5-HT blood levels. The protein complex
involved in the SERT regulatory network includes αIIbβ3
integrin, protein phosphatase 1 (PP1), protein phosphatase
2A (PP2A), the 5-HT2A receptor and undergoes PKC-, PKG-,
p38 MAPK-mediated regulation (Veenstra-VanderWeele and
Blakely, 2012). An initial study by Katsui et al. (1986) found
increased 5-HT blood levels associated with increased SERT
density on the platelet membrane (Vmax), while the affinity
(Kd) of SERT for 5-HT was unchanged. This initial study was
S. Gabriele et al.
later replicated by Marazziti et al. (2000). Hence SERT
trafficking may be skewed in favor of SERT externalization in
individuals showing elevated 5-HT blood levels, although
the existence of additional factors cannot be excluded. The
molecular mechanisms favouring SERT externalization
involve different genes in males and females (Mei et al.,
2007; Weiss et al., 2005), including for example common
variants at ITGB3 (Napolioni et al., 2011; Weiss et al., 2004,
2005, 2006a, 2006b) and rare variants at SLC6A4 (Prasad
et al., 2005; Veenstra-VanderWeele et al., 2012). It is also
apparently modulated by sex hormones, since 5-HT blood
levels are especially elevated in some autistic children,
whereas after puberty this excess becomes less pronounced
(McBride et al., 1998).
Abnormal neurodevelopment underlies autistic behaviors
(DiCicco-Bloom et al., 2006). Interestingly, 5-HT plays a key
role not only as a neurotransmitter in the adult brain, by
regulating a variety of behavioral, autonomic and cognitive
functions (Murphy and Lesch, 2008), but also as a trophic
factor during prenatal neurodevelopment (Di Pino et al.,
2004; Persico, 2009). According to recent animal models,
during pregnancy embryonic 5-HT is initially produced by
the placenta using maternal tryptophan, whereas subse-
quently 5-HT synthesis is undertaken by serotonergic neu-
rons located in raphe nuclei and extending their projections
to the cortex, basal ganglia, amygdala, hippocampus and
hypothalamus (Bonnin et al., 2011). Abnormalities in brain
5-HT systems were also reported in ASD, including an
altered developmental trajectory of 5-HT turnover
(Chugani et al., 1999) and reduced binding of 5-HT recep-
tors and SERT (Murphy et al., 2006; Nakamura et al., 2010).
Remarkably, proteins regulating 5-HT homeostasis periph-
erally and in the central nervous system (CNS) are largely
conserved: the SLC6A4 and HTR2A genes encode the same
SERT and 5-HT2A receptor, respectively, both in platelets
and brain (Cook et al., 1994; Lesch et al., 1993). Several
genes involved in SERT trafficking, especially SLC6A4 and
ITGB3, were mapped to autism linkage regions, found to
host rare variants causing ASD, or were associated with
autism through common variants (Cantor et al., 2005;
Coutinho et al., 2004; Napolioni et al., 2011; Stone et al.,
2004; Weiss et al., 2004. 2006a). Therefore, the molecular
mechanisms involved in blood 5-HT elevation may not only
tag a pathogenetically homogeneous subgroup of autistic
patients, but also underlie at least in part the neurodeve-
lopmental abnormalities present in the CNS of these
patients.
The consistency of the association between elevated
5-HT blood levels and autism, as well as the relevance of
5-HT roles in neurodevelopment, collectively make 5-HT
blood levels a primary candidate for comprehensive ASD
biomarker panels currently under scrutiny. However, studies
evaluating peripheral 5-HT in autistic patients and controls
differ in sampling demographics (i.e., age range, sex,
ethnicity), patient characteristics, selection of controls,
5-HT measurement protocol, technologies and biomaterials.
We thus undertook a systematic review of all studies
assessing 5-HT blood levels in ASD patients and controls,
followed by a series of meta-analyses in order to provide an
overall estimate of the effect size and between-study
heterogeneity for the association between elevated 5-HT
blood levels and autism, while verifying whether and to
Blood serotonin levels in autism spectrum disorder: A systematic review and meta-analysis 921

what extent differences in biomaterials, experimental pro-

tocols and technologies influence results.

2. Experimental procedures

2.1. Literature search strategy

Studies eligible for this systematic review and meta-analysis

were identified through a strategy employing an initial a
priori search protocol, followed by reiterative modifications
aimed at progressively maximizing search efficiency by
yielding increased numbers of pertinent publications. Our
final search terms were as follows:
(Autism OR autistic OR ASD OR pervasive developmental
disorder OR Asperger) AND (serotonin OR 5-HT OR 5-hydro-
xytryptamine) AND (serotonergic OR levels OR peripheral OR
serum OR plasma OR urine OR saliva OR blood OR platelets
OR cerebrospinal fluid OR red blood cells).
This computer-aided search involved the PUBMED, Sco-
pus, Google Scholar, CINAHL, EmBase, and ERIC databases
and is updated to July 8th, 2013. Once original publications
were collected, bibliographies were manually searched for
additional eligible references.

2.2. Study selection criteria

A total of 526 studies were initially identified through the

procedure outlined above. The following exclusion criteria were
subsequently applied (Figure 1): (1) case reports, commentaries
and reviews; (2) studies not in English, German, French, Italian, Figure 1 Study flow chart. Abbreviations: ASD = Autism Spec-
or Spanish; (3) studies on animal and/or cellular models; trum Disorder; 5-HT =5-hydroxytryptamine
(4) reports on syndromic autism, Rett syndrome or specific
diagnoses other than idiopathic ASD; (5) publications lacking
using the χ2 goodness-of-fit test and the I2 statistic. Data
measures of 5-HT blood levels, including pharmacological,
were then analyzed using either a fixed effects model or a
genetic, brain imaging, and post-mortem studies; (6) studies
more conservative random effects model, depending on the
reporting the effects of pharmacological treatments or trypto-
absence or presence of significant between-study heteroge-
phan depletion on 5-HT blood levels; (7) studies providing direct
neity, respectively. The fixed effects model is based on the
5-HT measures of patients and first-degree relatives, while
assumption that a single true effect is shared by all studies
referring to previous literature for control levels; (8) studies
and that observed estimates vary only as a function of
showing 5-HT blood levels in dot plot and histogram format, but
chance. On the contrary, the random effects model assumes
not providing numerical results. In this case, studies were
that each study estimates different true effects and that the
excluded if corresponding authors were either unable to
various effects are normally distributed around a mean
provide the numerical data or did not respond; and (9) reporting
value. When heterogeneity is small, both models yield
of identical previously-published data sets.
essentially identical results. In either case, effect sizes were
At the end of this process, 22 studies were selected, each
calculated using mean and standard deviation of the case and
assessing patients with idiopathic autism (i.e., DSM-IV
control distributions, and combined using the inverse var-
diagnoses of either Autistic Disorder, Asperger Disorder or
iance method to generate a pooled effect size and 95%
Pervasive Developmental Disorder Not Otherwise Specified,
confidence intervals (CI). This strategy also accounts for
PDD-NOS), measuring 5-HT blood levels on autistic patients
variability in small sample sizes, such as those investigated
and typically developing controls, while providing results as
in several studies. Effect measures are reported as Odds
mean and either standard deviation or standard error of the
Ratios (O.R.). Meta-analytical procedures were performed
mean with known sample size. Data extracted from the
using the Comprehensive Meta Analysis Program (Biostat,
original publications are summarized in Table 1.
Version 2.0, 2005).
Publication bias was estimated by the method of Egger
2.3. Statistical analyses et al. (1997), which uses a linear regression approach to
measure funnel plot asymmetry on the natural logarithm
Meta-analyses were performed for each biomaterial and scale of odds ratios (O.R.). Statistical significance for the
analytical procedure, whenever at least three published intercept was determined applying the T test.
studies with usable data were available. For each meta- The percentage of autistic subjects with elevated 5-HT
analysis, between-study heterogeneity was first assessed blood levels was calculated for each separate study, as
 922
Table 1 Summary of 22 studies systematically reviewed and selected for meta-analysis of 5-HT blood levels in autistic patients vs controls.

Reference Ethnicity Sample size Mean age 7 SD Sex (M/F) Analytical Biomaterial Mean 5-HT 7 SD Unit of % ASD
(range) procedure measure hyper
Autistics Autistics controls Autistics Autistics Controls 5HT
controls controls

Anderson et al. (2012)

(a) American 18 10.175.5 (4–20) 18/0 HPLC WB 219777 180750 ng/ml 28.8
(b) (mixed) 24 14.278.5 (5-30) 20/4 PPP 0.1570.94 0.1570.94 ng/ml 2.3
Hranilovic et al. Croatian 9 26.176.6 (16–45) nr FL PRP 5347217 1997125 ng/10^9 PLT 26.8
(2007) 45 39.979.2 (20–55) 44/1
Mulder et al. Dutch 81 12.573.5 70/11 HPLC PRP 8257274 6307190 ng/10^9 PLT 25.1
(2004) 60 11.573.9 29/31
Coutinho et al. Portuguese 105 7.14 (2–14) nr HPLC PRP 3047207 2607137 ng/10^9 PLT 13.4
(2004) 52 7.27 (4 m–17 y) nr
Spivak et al. Israeli 10 24.374.5 6/4 HPLC PPP 3.271.9 7.573.7 ng/ml 0
(2004) 12 30.073.6 6/6
Croonenberghs et al. (2000)
(a) Belgian 13 (12–18) 13/0 HPLC WB 223739 206768 ng/ml o1
(b) 13 13/0 PRP 95746 76744 ng/ml 6.7
Leboyer et al. French 60 9.274.2 (3–23) 19/19 FL PPPn 1797135 1707123 ng/mlnn 4
(1999) 209 (2–16 y, N=91; 416, (2–16 y, sm; (whole) (o16 y)
N=118) 416, nr)
McBride et al. (1998)
(a) African- 18 7.072.8 17/1 HPLC WB 323793 2977105 ng/ml 2.4
American 10 8.272.6 6/4
(b) 7 16.772.5 nr HPLC WB 198739 184731 ng/ml 10.9
11 24.075.7 nr
(c) Hispanic- 13 6.272.7 11/2 HPLC WB 3257121 262789 ng/ml 17.1
American 7 7.772.9 5/2
(d) Caucasian- 27 6.773.1 25/2 HPLC WB 245778 187750 ng/ml 29.8
American 21 8.072.9 14/7
(e) 12 23.776.3 nr HPLC WB 187784 152750 ng/ml 22.1
16 24.374.7 nr
Tordjman et al. Caucasian- 30 5.5572.49 30/0 HPLC WB 2647105 196772 ng/ml 23.6
(1995) American 10 7.7072.45 10/0
Yuwiler et al. Caucasian- 16 10 (3–19) nr FL WB 219727 24279 ng/ml 6.6
(1992) American 4 33 (5–52) nr

S. Gabriele et al.
Piven et al. Caucasian- 28 21.5 (6–53) 24/4 HPLC PRP 2377128.4 97.7758.1 ng/ml 56.8
(1991) American 10 25.1 (4–47) 5/5
Minderaa et al. (1989)
(a) Caucasian- 17 19.474.9 14/3 HPLC WB 163781.7 116735.0 ng/ml 39
(b) American 20 22.077.5 15/5 6207303 4657162 ng/10^9 PLT 29.1
Minderaa et al. (1987)
 Blood serotonin levels in autism spectrum disorder: A systematic review and meta-analysis
(a) Caucasian- 14 20.674.6 nr HPLC WB 163786.3 113724.6 ng/ml 53.6
(b) American 17 20.376.9 nr 6307333 4437112 ng/10^9 PLT 45.6
Abramson et al. American 57 13.975 43/14 FL WB 3997210 184797 ng/ml 54
(1989 (mixed) 17 Age-matched 13/4
Geller et al. (1988)
(a) Caucasian- 19 (1.6–11.6) nr FL WB 243718 245712 ng/ml 7.5
(b) American 26 age-matched nr 11807130 1010770 ng/10^9 PLT 59.1
Anderson et al. (1987)
(a) Caucasian- 21 16.876.04 15/6 HPLC WB 20574.3 136750 ng/ml 33
(b) American 87 14.677.47 45/42 7767337 5227211 ng/10^9 PLT 31.2
Badcock et al. Australian 30 10.2 (4–19) nr HPLC WB 229785 190788 ng/mlnn 5.4
(1987) 106 9.7 (1–19) nr
de Villard et al. French 39 8.6 (2–18) 26/13 HPLC PRP 8637280 7367160 ng/10^9 PLT 24.8
(1986) 33 9.0 (20 m–18 y) 26/7
Hoshino et al. Japanese 37 4.7 (3–11) 34/3 FL WB 173762 124744 ng/ml 26.8
(1984) 12 10.5 6/6
Hanley et al. Caucasian- 27 (7–22) nr FL WB 134.5756.9 56.6748.8 ng/ml 36.7
(1977) American 6 age-matched nr
Yuwiler et al. Caucasian- 7 (33–107) 7/0 FL WB 272753 183728 ng/ml 73.2
(1971) American 4 (31–120) 3/1
Ritvo et al. Caucasian- 24 (3–8) nr FL WB 263763 216761 ng/ml 12
(1970) American 36 age-matched nr

Abbreviations: WB =Whole Blood; PRP =Platelet-Rich Plasma; PPP=Platelet-Poor Plasma; PLT=platelet; HPLC=high-pressure liquid chromatography; FL =fluorometric; sm=sex-matched;
nr=no reported.
n
Reported as PPP in Materials and Methods, and as WB in Results. Blood sample procedures appear most compatible with PPP.
nn
Original values expressed in molarity.

923
924
follows: (1) 5-HT blood levels in each control sample were
assumed to follow a normal distribution, as previously shown
in empirical data sets (Mulder et al., 2004; Hranilovic et al.,
2007); (2) the upper limit of the normal distribution was
defined in each control sample as +2 S.D. from mean control
values; (3) the mean value of 5-HT blood levels in autistic
subjects was then transformed into a z-score based on the
control sample distribution; (4) the proportion of autistic
subjects with 5-HT blood levels above the upper limit of the
control distribution was estimated, and (5) is expressed as
percentage of the total number of autistic patients in each
sample.
3. Results
3.1. Study characteristics
The Literature search yielded 551 records, including 469
records from PubMed, 76 records from Scopus, and 6 additional
records from Eric. Applying our exclusion criteria (Figure 1), 22
studies were selected for review and meta-analysis (Table 1),
whereas 529 were excluded (listed as Suppl. Refs.). These 22
papers were published between 1970 and 2012; thirteen
studies were conducted in United States, 6 were from Europe,
2 from Asia and 1 from Australia. Sample sizes varied widely,
ranging from 7 to 105 autistic individuals (Coutinho et al.,
2004; Yuwiler et al., 1971), and from 4 to 106 controls
(Badcock et al., 1987; Yuwiler et al., 1971). Similarly, mean
age of autistic and control individuals varied broadly, ranging
from pre-puberal to post-puberal. The systematic review
identified three different biomaterials used for 5-HT assays:
whole blood (WB), platelet-rich plasma (PRP), and platelet-
poor plasma (PPP). Blood 5-HT values were expressed as
ng/mL, or as ng/109 platelets (PLT) in some studies using
WB or PRP. Moreover, 13 studies performed 5-HT assessment
using HPLC as analytical procedure, whereas 9 adopted fluoro-
metric assays (Table 1).
3.2. Studies performing 5-HT assays on different
biomaterials
Fifteen publications describe 5-HT levels assessed in WB from
19 independent samples, reported either as 5-HT concentra-
tion (ng/ml) or normalized by platelet number (ng/109PLT).
One publication reports results from five independent sam-
ples (McBride et al., 1998), two use different biomaterials
(Anderson et al., 2012; Croonenberghs et al., 2000), while
four others report WB data with and without normalization
by platelet number (Table 1) (Anderson et al., 1987; Geller
et al., 1988; Minderaa et al., 1987, 1989). Collectively the
mean7SEM percentage of ASD individuals with elevated
5-HT blood levels from studies using WB was 28.374.06%
(Figure 2). Meta-analyses performed on this pool of publica-
tions (N=15) showed higher levels of 5-HT in ASD subjects
compared to controls [O.R.=4.6; (3.1–5.2); P=1.0  10 12],
as well as significant between-study heterogeneity (P=0.002;
I2 =53%). This heterogeneity is seemingly real and does not stem
from systematic differences in measurement methodology.
In particular, studies where WB 5-HT concentrations are
S. Gabriele et al.
Figure 2 Percentages of autistic patients with elevated 5-HT
blood levels from studies using different biomaterials and
expressing 5-HT content with or without normalization by
platelet count. Abbreviations: WB = whole blood; PRP =plate-
let-rich plasma; PPP=platelet-poor plasma; PLT =platelet.
reported as such (ng/ml), still display substantial
between-study heterogeneity (Po0.05; I2 = 73%) following
exclusion of four studies where WB 5-HT concentrations are
normalized by platelet count (Figure 2) (Anderson et al.,
1987; Geller et al., 1988; Minderaa et al., 1987, 1989).
Nonetheless, normalization by platelet count may enhance
effect size over plain 5-HT concentration in WB: mean7SEM
percentage of ASD individuals with elevated 5-HT blood
levels was 41.278.1% using the former approach, as
compared to 25.674.6% in the latter, with meta-analysis
yielding O.R. of 6.7 (3.7–12.2; P =2.8  10 7) and 3.6 (2.3–
5.4; P= 1.8  10 8), respectively (Suppl. Figures 1 and 2).
No publication bias was found for these meta-analyses
[intercept 0.46; 95%CI ( 2.12 to + 3.05); P = 0.70; Intercept
0.18; 95%CI ( 410.7 to +411.1); P= 0.99, respectively]
(Suppl. Figure S3).
Four publications reported 5-HT levels measured in PRP
and normalized by platelet count (ng/109 PLT) (Coutinho
et al., 2004; de Villard et al., 1986; Hranilovic et al., 2007;
Mulder et al., 2004). Significantly higher 5-HT levels were
found in ASD subjects compared to controls (O.R. = 2.6;
P = 2.7  10 7) (Figure 3B), and specifically in 22.573.6% of
autistic patients collectively assessed in this group of
studies (Figure 2). Between-studies heterogeneity was much
lower compared to studies using WB (P =0.11; I2 =49%), and
no publication bias was found [intercept 1.54; 95%CI ( 11.3
to + 14.4); P= 0.65] (Suppl. Figure S4). Only two studies
reported 5-HT concentrations (ng/ml) measured in PRP
without normalizing for platelet counts (Croonenberghs
et al., 2000; Piven et al., 1991). These studies are listed
in Table 1, but no meta-analysis was performed, as their
number is below the minimum threshold of N =3 set in our
criteria. Collectively, 25.675.2% of ASD subjects showed
elevated 5-HT blood levels in the six studies employing PRP.
Only three studies measured 5-HT levels in PPP of ASD
patients and controls (Anderson et al., 2012; Leboyer et al.,
1999; Spivak et al., 2004). In this set, no significant case-
control difference was observed [O.R. = 0.54 (0.2–2.0);
P = 0.36, n.s.], whereas significant between-studies hetero-
geneity was detected (P= 0.01; I2 = 77%) (Figure 3C). Only
Blood serotonin levels in autism spectrum disorder: A systematic review and meta-analysis 925

Figure 3 Odds Ratios (O.R.) from: (A) random effects meta-analysis of studies reporting whole blood (WB) 5-HT levels; (B) fixed
effects meta-analysis of studies reporting 5-HT levels measured in platelet-rich plasma (PRP); and (C) random effects meta-analysis
of studies measuring 5-HT levels in platelet-poor plasma (PPP). O.R. are displayed as a forest plot, represented by a point and
bounded by its confidence intervals.

271.4% of autistic patients showed PPP 5-HT levels two S.D.

above control values (Figure 2). Also in this set of studies,
no publication bias was found [intercept 3.58; 95%CI
( 32.6 to + 25.4); P= 0.36] (Suppl. Figure S5).
Several other studies contrasting ASD patients with first-
degree relatives, but not including unrelated controls, were
not be incorporated into our case-control meta-analytical
design (Cook et al., 1988a, 1990; Kuperman et al., 1985;
Leventhal et al., 1990; Perry et al., 1991; Cuccaro et al.,
1993; Persico et al., 2002).

3.3. Studies performing 5-HT assays using

different analytical procedures

Measurements of 5-HT blood levels were initially per-

formed using fluorometric methods, accomplished by fluo-
rometer or spectrophotometer. Later, HPLC became the
more widespread approach. The effect size of differences
in WB 5-HT levels measured in autistic individuals and
controls applying these two different analytical procedures Figure 4 Odds Ratios (O.R.) from: (A) fixed effects meta-
is similar for HPLC [O.R. = 3.8 (2.7–5.3); P = 2,6  10 14] analysis of studies reporting 5-HT levels measured in WB using
(Figure 4A) (Anderson et al., 1987, 2012; Croonenberghs HPLC, and (B) random effects meta-analysis of studies reporting
et al., 2000; McBride et al., 1998; Minderaa et al., 1987, 5-HT levels measured in WB using fluorometric methods.
1989), and for fluorometric assays [O.R. = 3.3 (1.9–5.5);
P = 0.03] (Figure 4B) (Abramson et al., 1989; Geller et al.,
1988; Hanley et al., 1977; Hoshino et al., 1984; Ritvo to + 2.99); P = 0.71; intercept 0.54; 95%CI ( 6.04 to
et al., 1970; Yuwiler et al., 1971, 1992). However, sig- + 7.14), P-value = 0.83] (Suppl. Figure S6).
nificant between-study heterogeneity was found in the
latter set (P = 0.001; I2 = 73%) and not in HPLC studies,
which were much more homogenous (P = 0.17; I2 = 27%) 4. Discussion
(compare Figures 4A and 4B). This lower between-study
variance produced much greater statistical significance We conducted a systematic review and selection of studies
with HPLC studies, despite superimposable mean odds measuring 5-HT blood levels in autistic and control samples
ratios. Regression tests indicated no publication bias also using different biomaterials and methodologies. Data from
for these two meta-analyses [intercept 0.40; 95%CI ( 2.17 selected studies were then meta-analyzed, yielding results
926
expressed as overall percentage of ASD patients displaying
elevated 5-HT blood levels, and as global mean odds ratios.
The procedure employed here to systematically detect
published papers on 5-HT blood levels in autism was
broad-based and thorough. Our strict selection criteria
requiring that both case and control data be reported in
the same paper, no drug treatment applied, and population
controls be recruited rather than first-degree relatives,
reduced to a large extent the number of studies eligible
for meta-analysis, as is always the case when this approach
is employed to Literature surveys (Figure 1). Our results
reliably confirm the consistent association of elevated 5-HT
blood levels with autism, regardless of which technique and
biomaterial is used, with the predicted exception of PPP.
This conclusively demonstrates that peripheral 5-HT
deserves utmost attention in the definition of future
biomarker panels applicable to ASD.
Meta-analyses of studies employing WB and PRP reveal
higher 5-HT levels in ASD compared to control individuals in
all study sets, possibly with a larger effect size when 5-HT
amounts are measured in WB and normalized by platelet
count (Figures 2 and S2). Studies measuring 5-HT levels in
WB without normalizing by platelet count show a significant
degree of between-study heterogeneity, which negatively
impacts odds ratios. Heterogeneity may have been particu-
larly affected by two studies, Geller et al. (1988), and
Yuwiler et al. (1992), the former including only autistic
patients with WB 5-HT values similar to controls, as
reported by the authors, the latter for not matching ASD
cases and controls by age. This variable may be an
especially important confound, as 5-HT blood levels are
blunted by sex hormones and 5-HT elevations are more
prevalent in prepuberal than in postpuberal autistic indivi-
duals (McBride et al., 1998). Typically, most studies have not
applied strict age limitations, clumping together children
with adolescents, or adolescents with adults (Table 1).
Interestingly, many studies focussing on small children do
display greater 5-HT levels and broader differences
between ASD patients and controls. The set of studies using
WB without normalizing by platelet count may have suf-
fered to a greater extent from spurious differences intro-
duced by age and race/ethnicity, given the larger number of
studies in this data set. Instead, the difference between
platelet-normalized 5-HT amounts recorded in WB and PRP
(see Table 1 and Figure 2) does not reflect obvious differ-
ences in age range and methodology. Apparently platelet-
normalized PRP studies (Coutinho et al., 2004; de Villard
et al., 1986; Hranilovic et al., 2007; Mulder et al., 2004)
display lower 5-HT amounts especially among ASD cases and
larger between-study variability, as compared to platelet-
normalized WB studies (Anderson et al., 1987; Geller et al.,
1988; Minderaa et al., 1987, 1989). Hence methods
employed to obtain PRP may not all be equivalent in
preserving platelet 5-HT content for later measurement
and care should be placed into setting these preliminary
procedural steps, regardless of whether 5-HT amounts in
PRP will later be normalized by platelet count or simply
measured quantitatively.
As predictable, PPP represents an exception to the excess
of 5-HT present in WB and PRP of autistic individuals
compared to controls (Figure 3C). Serotonin levels in PPP
are so low that they often overlap with the sensitivity
S. Gabriele et al.
thresholds of technologies applied for 5-HT measurements,
making the estimation of mean 5-HT values in PPP less
reliable compared to WB and PRP (Brand and Anderson,
2011). Moreover, PPP collecting procedure do not exclude
completely contributions from platelet 5-HT. Probably due
to these methodological limitations, discrepant results were
reported in different studies (Anderson et al., 2012;
Leboyer et al., 1999; Spivak et al., 2004) [excluded from
our selection Vered et al. (2003); Connors et al. (2006)].
Nonetheless, collectively these results are in line with 5-HT
uptake and storage mechanisms in platelets as being
boosted in autistic patients with elevated 5-HT blood levels,
while neither peripheral 5-HT synthesis nor catabolism
appear significantly affected.
We also evaluated the weight that different analytical
procedures may have on peripheral 5-HT assessments in ASD.
Fluorometric methods were more often used in initial studies,
whereas HPLC later became the most frequently adopted
method (Table 1). Meta-analyses on HPLC and fluorometric
assays involved papers employing WB, since only this biomater-
ial was assessed in a minimum of three studies using either
technique. Mean odds ratios obtained for HPLC and fluorometric
assays were quite similar (3.8 vs 3.3, respectively; see Figures
4A and 4B). Despite this substantial equivalence, significant
between-study heterogeneity was found for fluorometric
assays. These discrepancies do not appear immediately due to
differences in experimental protocols, which were homoge-
neous and specific for discrimination between 5-HT and other 5-
hydroxyndoles (Bogdanski et al., 1956; Yuwiler et al., 1970).
Therefore heterogeneity in earlier studies should be primarily
due to non-technical factors, such as demographic variables,
more than to differences in assay specificity and sensitivity.
This systematic review and meta-analysis displays some
limitations, which should be duly acknowledged. While the
number of studies using WB, HPLC and fluorometric assays is
sufficient to provide reliable estimations of effect size and
between-study heterogeneity (Figures 3A, 4A, and 4B),
studies normalizing 5-HT amounts by platelet number and
employing PRP were relatively few and their meta-analyses
should be viewed with some caution (Figure 3B and S2).
Secondly, the influence of age and race/ethnicity on 5-HT
blood levels has been recognized only since the study by
McBride et al. (1998). Prior to this study, and also subse-
quently, the vast majority of published papers combines
prepuberal to postpuberal individuals into single patient
and control samples. We acknowledge the importance of
these variables by listing in Table 1 mean age and S.D.
(range) for all ASD and control samples, as well as ethnic
groups as pertaining to each original study. Unfortunately it
was not possible to control for race and especially for age in
our meta-analyses. Moreover, large difference in sample
size may have affected between-study heterogeneity.
Random-effects model used for meta-analyses showing
greater heterogeneity, accounted for this factor yielding
more reliable results (Walker et al., 2008). Finally, the
number of studies included in our systematic review was
much smaller than the number initially selected using the
Literature search strategy described above. However, this
procedure largely increases the reliability of our analyses
and enhances confidence in our conclusions: (a) studies on
syndromic autism and other disorders (n = 41) were
excluded, because genetic/genomic abnormalities typically
Blood serotonin levels in autism spectrum disorder: A systematic review and meta-analysis
resulting in syndromic autism, can themselves impact 5-HT
blood levels differently from idiopathic autism; (b) studies
with pharmacologically treated samples (n =78) were
excluded because psychoactive drugs, especially selective
serotonin uptake inhibitors (SSRIs), block 5-HT uptake into
platelets, spuriously reducing 5-HT blood levels to a very
large extent (Persico et al., 2002); (c) the strict inclusion of
studies assessing in parallel both patients and controls using
the same technology, while removing studies not assessing
population controls (n= 11), increases the reliability of our
meta-analyses. Importantly, several among these eleven
studies contrasted ASD patients with their first-degree
relatives (Cook et al., 1990; Cook et al., 1988; Kuperman
et al., 1985; Leventhal et al., 1990; Perry et al., 1991;
Cuccaro et al., 1993; Persico et al., 2002). These studies
provide strong evidence of familiality, supporting a genetic
basis for blood 5-HT elevation and demonstrating its value
not only as a biomarker, but also as a familial endopheno-
type (Ruggeri et al., 2013; Sacco et al., 2010). However, the
present review and meta-analysis is focused on potential
uses of 5-HT blood levels as a biomarker and thus requires
that patient values be contrasted exclusively with popula-
tion control values.
The aim of this systematic review and meta-analysis was
to assess peripheral blood 5-HT as a potential diagnostic
biomarker for ASD. Indeed elevated 5-HT blood levels are
confirmed the best-replicated biomarker in ASD, regardless
of which biomaterial is assessed (WB, PRP), whether 5-HT
amounts are normalized by platelet counts or not, and
which technique is used to measure 5-HT (fluorometric or
HPLC). Within this framework, our study also provides
elements for the development of a more consistent proto-
col, which could be shared by investigators in order to
perform more comparable studies and to improve diagnostic
tools as well as therapeutic strategies. In particular, our
results encourage the use of protocols employing WB,
normalizing by platelet number and measuring 5-HT by
HPLC, as those promising greater sensitivity and reliability.
More homogeneous assessments will further reinforce the
use of 5-HT blood levels as a biomarker of autism, and the
conclusive definition of its potential in fostering earlier
diagnoses, in estimating spontaneous developmental trajec-
tories and in predicting treatment response.
Role of the funding source
This work was supported by the Italian Ministry for University,
Scientific Research and Technology (PRIN n.2006058195 and
n.2008BACT54_002), the Italian Ministry of Health (RFPS-2007-5-
640174 and RF-2011-02350537), the Fondazione Gaetano e Mafalda
Luce (Milan, Italy), Autism Aid ONLUS (Naples, Italy), Autism Speaks
(Princeton, NJ), the Autism Research Institute (San Diego, CA), and
the Innovative Medicines Initiative Joint Undertaking (EU-AIMS, n.
115300). The authors declare that these funding sources had no role
in study design, in the collection, analysis and interpretation of the
data, in the writing of this report and in the decision to submit this
paper for publication.
Contributors
SG performed the systematic review, extracted data from papers
and contributed to meta-analyses. RS performed and supervised
927
statistical analyses. AMP designed the study. SG and AMP wrote
the manuscript. All authors approved the final version of the
manuscript.
Conflict of interest
The authors declare no conflict of interest.
Acknowledgments
We wish to thank all the investigators who kindly provided addi-
tional information regarding their papers for appropriate data
management.
Appendix A. Supporting information
Supplementary data associated with this article can be
found in the online version at http://dx.doi.org/10.1016/
j.euroneuro.2014.02.004.
References
Abramson, R.K., Wright, H.H., Carpenter, R., Brennan, W., Lumpuy, O.,
Cole, E., Young, S.R., 1989. Elevated blood serotonin in autistic
probands and their first-degree relatives. J. Autism Dev. Disord. 19,
397–407.
American Psychiatric Association, 2013. Diagnostic and Statistical
Manual of Mental Disorders (5th ed.) American Psychiatric
Publishing, Arlington, VA.
Anderson, G.M., Freedman, D.X., Cohen, D.J., Volkmar, F.R., Hoder,
E.L., McPhedran, P., Minderaa, R.B., Hansen, C.R., Young, J.G.,
1987. Whole blood serotonin in autistic and normal subjects. J.
Child. Psychol. Psychiatry 28, 885–900.
Anderson, G.M., Hertzig, M.E., McBride, P.A., 2012. Brief report:
platelet-poor plasma serotonin in autism. J. Autism Dev. Disord.
42, 1510–1514.
Badcock, N.R., Spence, J.G., Stern, L.M., 1987. Blood serotonin
levels in adults, autistic and non-autistic children with a
comparison of different methodologies. Ann. Clin. Biochem.
24, 625–634.
Bogdanski, D.F., Pletscher, A., Brodie, B.B., Undenfriend, S., 1956.
Identification and assay of serotonin in brain. J. Pharmacol. Exp.
Ther. 117, 82–88.
Bonnin, A., Goeden, N., Chen, K., Wilson, M.L., King, J., Shih, J.C.,
Blakely, R.D., Deneris, E.S., Levitt, P., 2011. A transient
placental source of serotonin for the fetal forebrain. Nature
472, 347–350.
Brand, T., Anderson, G.M., 2011. The measurement of platelet-poor
plasma serotonin: a systematic review of prior reports and
recommendations for improved analysis. Clin. Chem. 57,
1376–1386.
Cantor, R.M., Kono, N., Duvall, J.A., Alvarez-Retuerto, A., Stone, J.L.,
Alarcón, M., Nelson, S.F., Geschwind, D.H., 2005. Replication of
autism linkage: fine-mapping peak at 17q21. Am. J. Hum. Genet.
76, 1050–1056.
Chugani, D.C., Muzik, O., Behen, M., Rothermel, R., Janisse, J.J.,
Lee, J., Chugani, H.T., 1999. Developmental changes in brain
serotonin synthesis capacity in autistic and nonautistic children.
Ann. Neurol. 45, 287–295.
Chugani, D.C., 2012. Neuroimaging and neurochemistry of autism.
Pediatr. Clin. North Am. 59, 63–73.
928
Connors, S.L., Matteson, K.J., Sega, G.A., Lozzio, C.B., Carroll, R.C.,
Zimmerman, A.W., 2006. Plasma serotonin in autism. Pediatr.
Neurol. 35, 182–186.
Cook Jr., E.H., Leventhal, B.L., Freedman, D.X., 1988. Free
serotonin in plasma: autistic children and their first-degree
relatives. Biol. Psychiatry 24, 488–491.
Cook Jr., E.H., Leventhal, B.L., Heller, W., Metz, J., Wainwright,
M., Freedman, D.X., 1990. Autistic children and their
first-degree relatives: relationships between serotonin and
norepinephrine levels and intelligence. J. Neuropsychiatry Clin.
Neurosci. 2, 268–274.
Cook Jr., E.H., Fletcher, K.E., Wainwright, M., Marks, N., Yan, S.Y.,
Leventhal, B.L., 1994. Primary structure of the human platelet
serotonin 5-HT2A receptor: identify with frontal cortex seroto-
nin 5-HT2A receptor. J. Neurochem. 63, 465–469.
Cook, E.H., Leventhal, B.L., 1996. The serotonin system in autism.
Curr. Opin. Pediatr. 8, 348–354.
Coutinho, A.M., Oliveira, G., Morgadinho, T., Fesel, C., Macedo, T.R.,
Bento, C., Marques, C., Ataide, A., Miguel, T., Borges, L., Vicente,
A.M., 2004. Variants of the serotonin transporter gene (SLC6A4)
significantly contribute to hyperserotonemia in autism. Mol. Psy-
chiatry 9, 264–271.
Croonenberghs, J., Delmeire, L., Verkerk, R., Lin, A.H., Meskal, A.,
Neels, H., Van der Planken, M., Scharpe, S., Deboutte, D., Pison, G.,
Maes, M., 2000. Peripheral markers of serotonergic and noradrener-
gic function in post-pubertal, caucasian males with autistic disorder.
Neuropsychopharmacology 22, 275–283.
Cuccaro, M.L., Wright, H.H., Abramson, R.K., Marsteller, F.A.,
Valentine, J., 1993. Whole-blood serotonin and cognitive func-
tioning in autistic individuals and their first-degree relatives. J.
Neuropsychiatry Clin. Neurosci. 5, 94–101.
de Villard, R., Flachaire, E., Thoulon, J.M., Dalery, J., Maillet, J.,
Chauvin, C., Quincy, C., Renaud, B., 1986. Platelet serotonin
concentrations in autistic children and members of their
families. Encephale 12, 139–142.
Di Pino, G., Moessner, R., Lesch, K.P., Lauder, J.M., Persico, A.M.,
2004. Roles for serotonin in neurodevelopment: more than just
neural transmission. Curr. Neuropharmacol. 2, 403–418.
DiCicco-Bloom, E., Lord, C., Zwaigenbaum, L., Courchesne, E.,
Dager, S.R., Schmitz, C., Schultz, R.T., Crawley, J., Young, L.J.,
2006. The developmental neurobiology of autism spectrum
disorder. J. Neurosci. 26, 6897–6906.
Egger, M., Smith, G.D., Phillips, A.N., 1997. Meta-analysis: princi-
ples and procedures. Brit. Med. J. 315, 1533–1537.
Freitas B.C., Trujillo C.A., Carromeu C., Yusupova M., Herai R.H.,
Muotri A.R., Stem cells and modeling of autism spectrum
disorders, Exp Neurol. Oct 2. 2012 (Epub ahead of print).
Geller, E., Yuwiler, A., Freeman, B.J., Ritvo, E., 1988. Platelet
size, number, and serotonin content in blood of autistic, child-
hood schizophrenic, and normal children. J. Autism Dev. Disord.
18, 119–126.
Gershon, M.D., 2004. Review article: serotonin receptors and
transporters: roles in normal and abnormal gastrointestinal
motility. Aliment. Pharmacol. Ther. 20, 3–14.
Hanley, H.G., Stahl, S.M., Freedman, D.X., 1977. Hyperserotonemia
and amine metabolites in autistic and retarded children. Arch.
Gen. Psychiatry 34, 521–531.
Hoshino, Y., Yamamoto, T., Kaneko, M., Tachibana, R., Watanabe,
M., Ono, Y., Kumashiro, H., 1984. Blood serotonin and free
tryptophan concentration in autistic children. Neuropsychobiol-
ogy 11, 22–27.
Hranilovic, D., Bujas-Petkovic, Z., Vragovic, R., Vuk, T., Hock, K.,
Jernej, B., 2007. Hyperserotonemia in adults with autistic
disorder. J. Autism Dev. Disord. 37, 1934–1940.
Katsui, T., Okuda, M., Usuda, S., Koizumi, T., 1986. Kinetics of
3
H-serotonin uptake by platelets in infantile autism and devel-
opmental language disorder (including five pairs of twins).
J. Autism Dev. Disord. 16, 69–76.
S. Gabriele et al.
Kuperman, S., Beeghly, J.H., Burns, T.L., Tsai, L.Y., 1985. Serotonin
relationships of autistic probands and their first-degree rela-
tives. J. Am. Acad. Child Psychiatry 24, 186–190.
Leboyer, M., Philippe, A., Bouvard, M., Guilloud-Bataille, M.,
Bondoux, D., Tabuteau, F., 1999. Whole blood serotonin and
plasma beta-endorphin in autistic probands and their first-
degree relatives. Biol. Psychiatry 45, 158–163.
Lesch, K.P., Wolozin, B.L., Murphy, D.L., Riederer, P., 1993. Primary
structure of the human platelet serotonin uptake site: identity
with the brain serotonin transporter. J. Neurochem. 60,
2319–2322.
Leventhal, B.L., Cook Jr., E.H., Morford, M., Ravitz, A., Freedman,
D.X., 1990. Relationships of whole blood serotonin and plasma
norepinephrine within families. J. Autism Dev. Disord. 20,
499–511.
Marazziti, D., Muratori, F., Cesari, A., Masala, I., Baroni, S., Giannac-
cini, G., Dell'Osso, L., Cosenza, A., Pfanner, P., Cassano, G.B.,
2000. Increased density of the platelet serotonin transporter in
autism. Pharmacopsychiatry 33, 165–168.
McBride, P.A., Anderson, G.M., Hertzig, M.E., Snow, M.E., Thomp-
son, S.M., Khait, V.D., Shapiro, T., Cohen, D.J., 1998. Effects of
diagnosis, race, and puberty on platelet serotonin levels in
autism and mental retardation. J. Am. Acad. Child Adolesc.
Psychiatry 37, 767–776.
Mei, H., Cuccaro, M.L., Martin, E.R., 2007. Multifactor dimension-
ality reduction-phenomics: a novel method to capture genetic
heterogeneity with use of phenotypic variables. Am. J. Hum.
Genet. 81, 1251–1261.
Minderaa, R.B., Anderson, G.M., Volkmar, F.R., Akkerhuis, G.W.,
Cohen, D.J., 1987. Urinary 5-hydroxyindoleacetic acid and
whole blood serotonin and tryptophan in autistic and normal
subjects. Biol. Psychiatry 22, 933–940.
Minderaa, R.B., Anderson, G.M., Volkmar, F.R., Harcherick, D.,
Akkerhuis, G.W., Cohen, D.J., 1989. Whole blood serotonin and
tryptophan in autism: temporal stability and the effects of
medication. J. Autism Dev. Disord. 19, 129–136.
Mitchell, S., Cardy, J.O., Zwaigenbaum, L., 2011. Differentiating
autism spectrum disorder from other developmental delays in
the first two years of life. Dev. Disabil. Res. Rev. 17, 130–140.
Mulder, E.J., Anderson, G.M., Kema, I.P., de Bildt, A., van Lang, N.D.,
den Boer, J.A., Minderaa, R.B., 2004. Platelet serotonin levels in
pervasive developmental disorders and mental retardation: diag-
nostic group differences, within-group distribution, and behavioral
correlates. J. Am. Acad. Child Adolesc. Psychiatry 43, 491–499.
Murphy, D.G., Daly, E., Schmitz, N., Toal, F., Murphy, K., Curran, S.,
Erlandsson, K., Eersels, J., Kerwin, R., Ell, P., Travis, M., 2006.
Cortical serotonin 5-HT2A receptor binding and social commu-
nication in adults with Asperger's syndrome: an in vivo SPECT
study. Am. J. Psychiatry 163, 934–936.
Murphy, D.L., Lesch, K.P., 2008. Targeting the murine serotonin
transporter: insights into human neurobiology. Nat. Rev. Neu-
rosci. 9, 85–96.
Nakamura, K., Sekine, Y., Ouchi, Y., Tsujii, M., Yoshikawa, E.,
Futatsubashi, M., Tsuchiya, K.J., Sugihara, G., Iwata, Y., Suzuki,
K., Matsuzaki, H., Suda, S., Sugiyama, T., Takei, N., Mori, N.,
2010. Brain serotonin and dopamine transporter bindings in
adults with high functioning autism. Arch. Gen. Psychiatry 67,
59–68.
Napolioni, V., Lombardi, F., Sacco, R., Curatolo, P., Manzi, B.,
Alessandrelli, R., Militerni, R., Bravaccio, C., Lenti, C., Saccani,
M., Schneider, C., Melmed, R., Pascucci, T., Puglisi-Allegra, S.,
Reichelt, K.L., Rousseau, F., Lewin, P., Persico, A.M., 2011.
Family-based association study of ITGB3 in autism spectrum
disorder and its endophenotypes. Eur. J. Hum. Genet. 19,
353–359.
Perry, B.D., Cook Jr., E.H., Leventhal, B.L., Wainwright, M.S.,
Freedman, D.X., 1991. Platelet 5-HT2 serotonin receptor
Blood serotonin levels in autism spectrum disorder: A systematic review and meta-analysis
binding sites in autistic children and their first-degree relatives.
Biol. Psychiatry 30, 121–130.
Persico, A.M., Pascucci, T., Puglisi-Allegra, S., Militerni, R.,
Bravaccio, C., Schneider, C., 2002. Serotonin transporter gene
promoter variants do not explain the hyperserotoninemia in
autistic children. Mol. Psychiatry 7, 795–800.
Persico, A.M., 2009. Developmental roles of the serotonin trans-
porter. In: Kalueff, A.V. (Ed.), Experimental Models in Serotonin
Transporter Research. Cambridge University Press, UK,
pp. 78–104.
Piven, J., Tsai, G.C., Nehme, E., Coyle, J.T., Chase, G.A., Folstein,
S.E., 1991. Platelet serotonin, a possible marker for familial
autism. J. Autism Dev. Disord. 21, 51–59.
Prasad, H.C., Zhu, C.B., McCauley, J.L., Samuvel, D.J., Rama-
moorthy, S., Shelton, R.C., Hewlett, W.A., Sutcliffe, J.S.,
Blakely, R.D., 2005. Human serotonin transporter variants dis-
play altered sensitivity to protein kinase G and p38 mitogen-
activated protein kinase. Proc. Natl. Acad. Sci. USA 102,
11545–11550.
Ritvo, E.R., Yuwiler, A., Geller, E., Ornitz, E.M., Saeger, K.,
Plotkin, S., 1970. Increased blood serotonin and platelets in
early infantile autism. Arch. Gen. Psychiatry 23, 566–572.
Ruggeri, B., Sarkans, U., Schumann, G., Persico, A.M., 2013.
Biomarkers in Autism Spectrum Disorder: The Old and The
New. Psychopharmacology (Berl.), 2013 Oct 6. [Epub ahead of
print] PubMed PMID: 24096533.
Sacco, R., Curatolo, P., Manzi, B., Militerni, R., Bravaccio, C.,
Frolli, A., Lenti, C., Saccani, M., Elia, M., Reichelt, K.L.,
Pascucci, T., Puglisi-Allegra, S., Persico, A.M., 2010. Principal
pathogenetic components and biological endophenotypes in
autism spectrum disorders. Autism Res. 3, 237–252.
Schain, R.J., Freedman, D.X., 1961. Studies on 5-hydroxyindole
metabolism in autistic and other mentally retarded children.
J. Pediatr. 58, 315–320.
Spivak, B., Golubchik, P., Mozes, T., Vered, Y., Nechmad, A.,
Weizman, A., Strous, R.D., 2004. Low platelet-poor plasma
levels of serotonin in adult autistic patients. Neuropsychobiology
50, 157–160.
State, M.W., Levitt, P., 2011. The conundrums of understanding
genetic risks for autism spectrum disorders. Nat. Neurosci. 14,
1499–1506.
Stone, J.L., Merriman, B., Cantor, R.M., Yonan, A.L., Gilliam, T.C.,
Geschwind, D.H., Nelson, S.F., 2004. Evidence for sex-specific
risk alleles in autism spectrum disorder. Am. J. Hum. Genet. 75,
1117–1123.
Tordjman, S., Anderson, G.M., McBride, P.A., Hertzig, M.E., Snow,
M.E., Hall, L.M., Ferrari, P., Cohen, D.J., 1995. Plasma andro-
gens in autism. J. Autism. Dev. Disord. 25, 295–304.
Veenstra-VanderWeele, J., Blakely, R.D., 2012. Networking in autism:
leveraging genetic, biomarker and model system findings in the
search for new treatments. Neuropsychopharmacology 37,
196–212.
929
Veenstra-VanderWeele, J., Muller, C.L., Iwamoto, H., Sauer, J.E.,
Owens, W.A., Shah, C.R., Cohen, J., Mannangatti, P., Jessen, T.,
Thompson, B.J., Ye, R., Kerr, T.M., Carneiro, A.M., Crawley, J.
N., Sanders-Bush, E., McMahon, D.G., Ramamoorthy, S., Daws,
L.C., Sutcliffe, J.S., Blakely, R.D., 2012. Autism gene variant
causes hyperserotonemia, serotonin receptor hypersensitivity,
social impairment and repetitive behavior. Proc. Natl. Acad. Sci.
USA 109, 5469–5474.
Vered, Y., Golubchik, P., Mozes, T., Strous, R., Nechmad, A., Mester, R.,
Weizman, A., Spivak, B., 2003. The platelet-poor plasma 5-HT
response to carbohydrate rich meal administration in adult autistic
patients compared with normal controls. Hum. Psychopharmacol.
18, 395–399.
Walker, E., Hernandez, A.V., Kattan, M.W., 2008. Meta-analysis: its
strengths and limitations. Cleve Clin. J. Med. 75, 431–439.
Walsh, P., Elsabbagh, M., Bolton, P., Singh, I., 2011. In search of
biomarkers for autism: scientific, social and ethical challenges.
Nat. Rev. Neurosci. 12, 603–612.
Wan, M.W., Green, J., Elsabbagh, M., Johnson, M., Charman, T.,
Plummer, F., 2013. Quality of interaction between at-risk infants
and caregiver at 12–15 months is associated with 3-year autism
outcome. J. Child. Psychol. Psychiatry 54, 763–771.
Weiss, L.A., Veenstra-Vanderweele, J., Newman, D.L., Kim, S.J.,
Dytch, H., McPeek, M.S., Cheng, S., Ober, C., Cook Jr., E.H.,
Abney, M., 2004. Genome-wide association study identifies ITGB3 as
a QTL for whole blood serotonin. Eur. J. Hum. Genet. 12, 949–954.
Weiss, L.A., Abney, M., Cook Jr., E.H., Ober, C., 2005. Sex-specific
genetic architecture of whole blood serotonin levels. Am. J.
Hum. Genet. 76, 33–41.
Weiss, L.A., Kosova, G., Delahanty, R.J., Jiang, L., Cook, E.H.,
Ober, C., 2006a. Variation in ITGB3 is associated with whole-
blood serotonin level and autism susceptibility. Eur. J. Hum.
Genet. 14, 923–931.
Weiss, L.A., Ober, C., Cook, E.H., 2006b. ITGB3 shows genetic and
expression interaction with SLC6A4. Hum. Genet. 120, 93–100.
Yuwiler, A., Plotkin, S., Geller, E., Ritvo, E.R., 1970. A rapid
accurate procedure for the determination of serotonin in whole
human blood. Biochem. Med. 3, 426–431.
Yuwiler, A., Ritvo, E.R., Bald, D., Kipper, D., Koper, A., 1971.
Examination of circadian rhythmicity of blood serotonin and
platelets in autistic and non-autistic children. J. Autism Child
Schizophr. 1, 421–435.
Yuwiler, A., Shih, J.C., Chen, C.H., Ritvo, E.R., Hanna, G.,
Ellison, G.W., King, B.H., 1992. Hyperserotoninemia and antiser-
otonin antibodies in autism and other disorders. J. Autism. Dev.
Disord. 22, 33–45.
Zalsman, G., Aslanov-Farbstein, D., Rehavi, M., Roz, N., Vermeiren,
R., Laor, N., Weizman, A., Toren, P., 2011. Platelet vesicular
monoamine transporter 2 density in the disruptive behavior
disorders. J. Child Adolesc. Psychopharmacol. 21, 341–344.