Association of Peripheral Blood Levels of Brain-Derived Neurotrophic Factor With Autism Spectrum Disorder in Children: A Systematic Review and Meta-analysis | Autism Spectrum Disorders | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  PRISMA Flowchart of the Literature Search
PRISMA Flowchart of the Literature Search

BDNF indicates brain-derived neurotrophic factor.

Figure 2.  Forest Plot for Random-Effects Meta-analysis
Forest Plot for Random-Effects Meta-analysis

Differences in blood levels of brain-derived neurotrophic factor are shown between children with autism spectrum disorder (ASD) and healthy controls. The sizes of the squares are proportional to study weights. Diamond marker indicates pooled effect size.

Figure 3.  Forest Plot for Subgroup Analysis Stratified by Study Design
Forest Plot for Subgroup Analysis Stratified by Study Design

Pooled results compare blood levels of brain-derived neurotrophic factor between children with autism spectrum disorder (ASD) and healthy controls. Study design is stratified into neonate and nonneonate studies. The sizes of the squares are proportional to study weights. Diamond markers indicate pooled effect sizes.

Figure 4.  Forest Plot for Subgroup Analysis Stratified by Sample Source
Forest Plot for Subgroup Analysis Stratified by Sample Source

Pooled results compare blood levels of brain-derived neurotrophic factor between children with autism spectrum disorder (ASD) and healthy controls. Sample source is stratified into serum and plasma samples in the nonneonate group. The sizes of the squares are proportional to study weights. Diamond markers indicate pooled effect sizes.

Figure 5.  Funnel Plot of Precision by Hedges g Statistics
Funnel Plot of Precision by Hedges g Statistics

Publication bias in studies comparing brain-derived neurotrophic factor levels between children with autism spectrum disorder and healthy controls. The plots describe the effect size (Hedges g statisic) of studies against their precision (inverse of SE). Data markers indicate individual studies. Diamond marker indicates pooled effect.

1.
Caronna  EB, Milunsky  JM, Tager-Flusberg  H.  Autism spectrum disorders: clinical and research frontiers.  Arch Dis Child. 2008;93(6):518-523.PubMedGoogle ScholarCrossref
2.
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA: American Psychiatric Association; 2013.
3.
Ha  S, Sohn  IJ, Kim  N, Sim  HJ, Cheon  KA.  Characteristics of brains in autism spectrum disorder: structure, function and connectivity across the lifespan.  Exp Neurobiol. 2015;24(4):273-284.PubMedGoogle ScholarCrossref
4.
Giarelli  E, Wiggins  LD, Rice  CE,  et al.  Sex differences in the evaluation and diagnosis of autism spectrum disorders among children.  Disabil Health J. 2010;3(2):107-116.PubMedGoogle ScholarCrossref
5.
Global Burden of Disease Study 2013 Collaborators.  Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013.  Lancet. 2015;386(9995):743-800.PubMedGoogle ScholarCrossref
6.
Mizejewski  GJ.  Biomarker testing for suspected autism spectrum disorder in early childhood: is such testing now feasible?  Biomark Med. 2012;6(4):503-506.PubMedGoogle ScholarCrossref
7.
Fakhoury  M.  Autistic spectrum disorders: a review of clinical features, theories and diagnosis.  Int J Dev Neurosci. 2015;43:70-77.PubMedGoogle ScholarCrossref
8.
Garcia  KL, Yu  G, Nicolini  C,  et al.  Altered balance of proteolytic isoforms of pro–brain-derived neurotrophic factor in autism.  J Neuropathol Exp Neurol. 2012;71(4):289-297.PubMedGoogle ScholarCrossref
9.
Taurines  R, Segura  M, Schecklmann  M,  et al.  Altered peripheral BDNF mRNA expression and BDNF protein concentrations in blood of children and adolescents with autism spectrum disorder.  J Neural Transm (Vienna). 2014;121(9):1117-1128.PubMedGoogle ScholarCrossref
10.
Wang  M, Chen  H, Yu  T, Cui  G, Jiao  A, Liang  H.  Increased serum levels of brain-derived neurotrophic factor in autism spectrum disorder.  Neuroreport. 2015;26(11):638-641.PubMedGoogle ScholarCrossref
11.
Yoo  HJ, Yang  SY, Cho  IH, Park  M, Kim  SA.  Polymorphisms of BDNF gene and autism spectrum disorders: family based association study with Korean trios.  Psychiatry Investig. 2014;11(3):319-324.PubMedGoogle ScholarCrossref
12.
Bamji  SX, Rico  B, Kimes  N, Reichardt  LF.  BDNF mobilizes synaptic vesicles and enhances synapse formation by disrupting cadherin-β-catenin interactions.  J Cell Biol. 2006;174(2):289-299.PubMedGoogle ScholarCrossref
13.
Binder  DK, Scharfman  HE.  Brain-derived neurotrophic factor.  Growth Factors. 2004;22(3):123-131.PubMedGoogle ScholarCrossref
14.
Greenberg  ME, Xu  B, Lu  B, Hempstead  BL.  New insights in the biology of BDNF synthesis and release: implications in CNS function.  J Neurosci. 2009;29(41):12764-12767.PubMedGoogle ScholarCrossref
15.
Hempstead  BL.  Brain-derived neurotrophic factor: three ligands, many actions.  Trans Am Clin Climatol Assoc. 2015;126:9-19.PubMedGoogle Scholar
16.
Lu  B, Nagappan  G, Guan  X, Nathan  PJ, Wren  P.  BDNF-based synaptic repair as a disease-modifying strategy for neurodegenerative diseases.  Nat Rev Neurosci. 2013;14(6):401-416.PubMedGoogle ScholarCrossref
17.
Lu  Y, Christian  K, Lu  B.  BDNF: a key regulator for protein synthesis-dependent LTP and long-term memory?  Neurobiol Learn Mem. 2008;89(3):312-323.PubMedGoogle ScholarCrossref
18.
Molendijk  ML, Spinhoven  P, Polak  M, Bus  BA, Penninx  BW, Elzinga  BM.  Serum BDNF concentrations as peripheral manifestations of depression: evidence from a systematic review and meta-analyses on 179 associations (N=9484).  Mol Psychiatry. 2014;19(7):791-800.PubMedGoogle ScholarCrossref
19.
Fernandes  BS, Steiner  J, Berk  M,  et al.  Peripheral brain-derived neurotrophic factor in schizophrenia and the role of antipsychotics: meta-analysis and implications.  Mol Psychiatry. 2015;20(9):1108-1119.PubMedGoogle ScholarCrossref
20.
Song  JH, Yu  JT, Tan  L.  Brain-derived neurotrophic factor in Alzheimer’s disease: risk, mechanisms, and therapy.  Mol Neurobiol. 2015;52(3):1477-1493.PubMedGoogle ScholarCrossref
21.
Fernandes  BS, Berk  M, Turck  CW, Steiner  J, Gonçalves  CA.  Decreased peripheral brain-derived neurotrophic factor levels are a biomarker of disease activity in major psychiatric disorders: a comparative meta-analysis.  Mol Psychiatry. 2014;19(7):750-751.PubMedGoogle ScholarCrossref
22.
Pan  W, Banks  WA, Fasold  MB, Bluth  J, Kastin  AJ.  Transport of brain-derived neurotrophic factor across the blood-brain barrier.  Neuropharmacology. 1998;37(12):1553-1561.PubMedGoogle ScholarCrossref
23.
Bryn  V, Halvorsen  B, Ueland  T,  et al.  Brain derived neurotrophic factor (BDNF) and autism spectrum disorders (ASD) in childhood.  Eur J Paediatr Neurol. 2015;19(4):411-414.PubMedGoogle ScholarCrossref
24.
Connolly  AM, Chez  M, Streif  EM,  et al.  Brain-derived neurotrophic factor and autoantibodies to neural antigens in sera of children with autistic spectrum disorders, Landau-Kleffner syndrome, and epilepsy.  Biol Psychiatry. 2006;59(4):354-363.PubMedGoogle ScholarCrossref
25.
Nelson  KB, Grether  JK, Croen  LA,  et al.  Neuropeptides and neurotrophins in neonatal blood of children with autism or mental retardation.  Ann Neurol. 2001;49(5):597-606.PubMedGoogle ScholarCrossref
26.
Nelson  PG, Kuddo  T, Song  EY,  et al.  Selected neurotrophins, neuropeptides, and cytokines: developmental trajectory and concentrations in neonatal blood of children with autism or Down syndrome.  Int J Dev Neurosci. 2006;24(1):73-80.PubMedGoogle ScholarCrossref
27.
Rodrigues  DH, Rocha  NP, Sousa  LF, Barbosa  IG, Kummer  A, Teixeira  AL.  Circulating levels of neurotrophic factors in autism spectrum disorders.  Neuro Endocrinol Lett. 2014;35(5):380-384.PubMedGoogle Scholar
28.
Ray  B, Long  JM, Sokol  DK, Lahiri  DK.  Increased secreted amyloid precursor protein-α (sAPPα) in severe autism: proposal of a specific, anabolic pathway and putative biomarker.  PLoS One. 2011;6(6):e20405.PubMedGoogle ScholarCrossref
29.
Liberati  A, Altman  DG, Tetzlaff  J,  et al.  The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.  J Clin Epidemiol. 2009;62(10):e1-e34.PubMedGoogle ScholarCrossref
30.
Hedges  LV, Olkin  I.  Statistical Methods for Meta-analysis. New York, NY: Academic Press; 1985.
31.
Baumeister  D, Akhtar  R, Ciufolini  S, Pariante  CM, Mondelli  V.  Childhood trauma and adulthood inflammation: a meta-analysis of peripheral C-reactive protein, interleukin-6 and tumour necrosis factor-α.  Mol Psychiatry. 2015;21(5):642-649.PubMedGoogle ScholarCrossref
32.
Bowden  J, Tierney  JF, Copas  AJ, Burdett  S.  Quantifying, displaying and accounting for heterogeneity in the meta-analysis of RCTs using standard and generalised Q statistics.  BMC Med Res Methodol. 2011;11:41.PubMedGoogle ScholarCrossref
33.
Masi  A, Quintana  DS, Glozier  N, Lloyd  AR, Hickie  IB, Guastella  AJ.  Cytokine aberrations in autism spectrum disorder: a systematic review and meta-analysis.  Mol Psychiatry. 2015;20(4):440-446.PubMedGoogle ScholarCrossref
34.
Lau  J, Ioannidis  JP, Schmid  CH.  Quantitative synthesis in systematic reviews.  Ann Intern Med. 1997;127(9):820-826.PubMedGoogle ScholarCrossref
35.
Egger  M, Davey Smith  G, Schneider  M, Minder  C.  Bias in meta-analysis detected by a simple, graphical test.  BMJ. 1997;315(7109):629-634.PubMedGoogle ScholarCrossref
36.
Soeken  KL, Sripusanapan  A.  Assessing publication bias in meta-analysis.  Nurs Res. 2003;52(1):57-60.PubMedGoogle ScholarCrossref
37.
Kazim  SF, Cardenas-Aguayo  MdelC, Arif  M,  et al.  Sera from children with autism induce autistic features which can be rescued with a CNTF small peptide mimetic in rats.  PLoS One. 2015;10(3):e0118627.PubMedGoogle ScholarCrossref
38.
Riikonen  R.  Neurotrophic factors in the pathogenesis of Rett syndrome.  J Child Neurol. 2003;18(10):693-697.PubMedGoogle ScholarCrossref
39.
Abdallah  MW, Pearce  BD, Larsen  N,  et al.  Amniotic fluid MMP-9 and neurotrophins in autism spectrum disorders: an exploratory study.  Autism Res. 2012;5(6):428-433.PubMedGoogle ScholarCrossref
40.
Meng  WD, Sun  SJ, Yang  J, Chu  RX, Tu  W, Liu  Q.  Elevated serum brain-derived neurotrophic factor (BDNF) but not BDNF gene Val66Met polymorphism is associated with autism spectrum disorders [published online January 28, 2016].  Mol Neurobiol. doi:10.1007/s12035-016-9721-9PubMedGoogle Scholar
41.
Hashimoto  K, Iwata  Y, Nakamura  K,  et al.  Reduced serum levels of brain-derived neurotrophic factor in adult male patients with autism.  Prog Neuropsychopharmacol Biol Psychiatry. 2006;30(8):1529-1531.PubMedGoogle ScholarCrossref
42.
Miyazaki  K, Narita  N, Sakuta  R,  et al.  Serum neurotrophin concentrations in autism and mental retardation: a pilot study.  Brain Dev. 2004;26(5):292-295.PubMedGoogle ScholarCrossref
43.
Nishimura  K, Nakamura  K, Anitha  A,  et al.  Genetic analyses of the brain-derived neurotrophic factor (BDNF) gene in autism.  Biochem Biophys Res Commun. 2007;356(1):200-206.PubMedGoogle ScholarCrossref
44.
Segura  M, Pedreño  C, Obiols  J,  et al.  Neurotrophin blood-based gene expression and social cognition analysis in patients with autism spectrum disorder.  Neurogenetics. 2015;16(2):123-131.PubMedGoogle ScholarCrossref
45.
Spratt  EG, Granholm  AC, Carpenter  LA,  et al.  Pilot study and review: physiological differences in BDNF, a potential biomarker in males and females with autistic disorder.  Int Neuropsychiatr Dis J. 2015;3(1):19-26.PubMedGoogle ScholarCrossref
46.
Abdallah  MW, Mortensen  EL, Greaves-Lord  K,  et al.  Neonatal levels of neurotrophic factors and risk of autism spectrum disorders.  Acta Psychiatr Scand. 2013;128(1):61-69.PubMedGoogle ScholarCrossref
47.
Al-Ayadhi  LY.  Relationship between Sonic hedgehog protein, brain-derived neurotrophic factor and oxidative stress in autism spectrum disorders.  Neurochem Res. 2012;37(2):394-400.PubMedGoogle ScholarCrossref
48.
Correia  CT, Coutinho  AM, Sequeira  AF,  et al.  Increased BDNF levels and NTRK2 gene association suggest a disruption of BDNF/TrkB signaling in autism.  Genes Brain Behav. 2010;9(7):841-848.PubMedGoogle ScholarCrossref
49.
Croen  LA, Goines  P, Braunschweig  D,  et al.  Brain-derived neurotrophic factor and autism: maternal and infant peripheral blood levels in the Early Markers for Autism (EMA) Study.  Autism Res. 2008;1(2):130-137.PubMedGoogle ScholarCrossref
50.
Enstrom  A, Onore  C, Tarver  A,  et al.  Peripheral blood leukocyte production of BDNF following mitogen stimulation in early onset and regressive autism.  Am J Biochem Biotechnol. 2008;4(2):121-129.Google ScholarCrossref
51.
Halepoto  DM, Bashir  S, Zeina  R, Al-Ayadhi  LY.  Correlation between Hedgehog (Hh) protein family and brain-derived neurotrophic factor (BDNF) in autism spectrum disorder (ASD).  J Coll Physicians Surg Pak. 2015;25(12):882-885.PubMedGoogle Scholar
52.
Kasarpalkar  NJ, Kothari  ST, Dave  UP.  Brain-derived neurotrophic factor in children with autism spectrum disorder.  Ann Neurosci. 2014;21(4):129-133.PubMedGoogle Scholar
53.
Katoh-Semba  R, Wakako  R, Komori  T,  et al.  Age-related changes in BDNF protein levels in human serum: differences between autism cases and normal controls.  Int J Dev Neurosci. 2007;25(6):367-372.PubMedGoogle ScholarCrossref
54.
Ricci  S, Businaro  R, Ippoliti  F,  et al.  Altered cytokine and BDNF levels in autism spectrum disorder.  Neurotox Res. 2013;24(4):491-501.PubMedGoogle ScholarCrossref
55.
Zhang  QB, Jiang  LF, Kong  LY, Lu  YJ.  Serum brain-derived neurotrophic factor levels in Chinese children with autism spectrum disorders: a pilot study.  Int J Dev Neurosci. 2014;37:65-68.PubMedGoogle ScholarCrossref
56.
Mansour  M, Mohamed  A, Azam  H, Henedy  M.  Brain derived neurotrophic factor in autism.  Curr Psychiatr. 2010;17(1):23-29.Google Scholar
57.
Liu  YH, Jiao  SS, Wang  YR,  et al.  Associations between ApoEε4 carrier status and serum BDNF levels: new insights into the molecular mechanism of ApoEε4 actions in Alzheimer’s disease.  Mol Neurobiol. 2015;51(3):1271-1277.PubMedGoogle ScholarCrossref
58.
Qin  X-Y, Cao  C, Cawley  NX,  et al.  Decreased peripheral brain-derived neurotrophic factor levels in Alzheimer’s disease: a meta-analysis study (N=7277) [published online April 26, 2016].  Mol Psychiatry. doi:10.1038/mp.2016.62PubMedGoogle Scholar
59.
Pandya  CD, Kutiyanawalla  A, Pillai  A.  BDNF-TrkB signaling and neuroprotection in schizophrenia.  Asian J Psychiatr. 2013;6(1):22-28.PubMedGoogle ScholarCrossref
60.
Weickert  CS, Hyde  TM, Lipska  BK, Herman  MM, Weinberger  DR, Kleinman  JE.  Reduced brain-derived neurotrophic factor in prefrontal cortex of patients with schizophrenia.  Mol Psychiatry. 2003;8(6):592-610.PubMedGoogle ScholarCrossref
61.
Chen  B, Dowlatshahi  D, MacQueen  GM, Wang  JF, Young  LT.  Increased hippocampal BDNF immunoreactivity in subjects treated with antidepressant medication.  Biol Psychiatry. 2001;50(4):260-265.PubMedGoogle ScholarCrossref
62.
Dunham  JS, Deakin  JF, Miyajima  F, Payton  A, Toro  CT.  Expression of hippocampal brain-derived neurotrophic factor and its receptors in Stanley consortium brains.  J Psychiatr Res. 2009;43(14):1175-1184.PubMedGoogle ScholarCrossref
63.
Chomiak  T, Hu  B.  Alterations of neocortical development and maturation in autism: insight from valproic acid exposure and animal models of autism.  Neurotoxicol Teratol. 2013;36:57-66.PubMedGoogle ScholarCrossref
64.
Tang  G, Gudsnuk  K, Kuo  SH,  et al.  Loss of mTOR-dependent macroautophagy causes autistic-like synaptic pruning deficits.  Neuron. 2014;83(5):1131-1143.PubMedGoogle ScholarCrossref
65.
Binder  DK, Croll  SD, Gall  CM, Scharfman  HE.  BDNF and epilepsy: too much of a good thing?  Trends Neurosci. 2001;24(1):47-53.PubMedGoogle ScholarCrossref
66.
Tuchman  R, Rapin  I.  Epilepsy in autism.  Lancet Neurol. 2002;1(6):352-358.PubMedGoogle ScholarCrossref
67.
Teng  HK, Teng  KK, Lee  R,  et al.  ProBDNF induces neuronal apoptosis via activation of a receptor complex of p75NTR and sortilin.  J Neurosci. 2005;25(22):5455-5463.PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    EXPAND ALL
    Two previous meta-analyses also demonstrate higher BDNF in autism compared to controls
    Dr. Bernard Crespi, FRSC | Department of Biological Sciences, Simon Fraser University, Canada
    Readers may wish to note that two recently-published papers, which were
    not cited in this article, have previously demonstrated higher BDNF in autism than in controls by meta-analyses.

    Peripheral brain-derived neurotrophic factor in autism spectrum
    disorder: a systematic review and meta-analysis.
    Zheng Z, Zhang L, Zhu T, Huang J, Qu Y, Mu D.
    Sci Rep. 2016 published Aug 10 6:31241.

    Meta-Analysis of BDNF Levels in Autism.
    Armeanu R, Mokkonen M, Crespi B.
    Cell Mol Neurobiol. 2016 published Aug 8.

    Future studies on this topic might usefully test for associations of BDNF levels
    with specific autism-related phenotypes, to help in understanding functional links
    of this
    neurotrophic factor with autism. It is also of interest that BDNF levels have been shown to be reduced in schizophrenia, bipolar disorder, and depression
    by meta-analyses (references in Armeanu et al. 2016), which suggests usefulness
    of this factor in differentiating autism spectrum from psychotic-affective spectrum disorders.
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Original Investigation
    November 2016

    Association of Peripheral Blood Levels of Brain-Derived Neurotrophic Factor With Autism Spectrum Disorder in Children: A Systematic Review and Meta-analysis

    Author Affiliations
    • 1Center on Translational Neuroscience, College of Life and Environmental Sciences, Minzu University of China, Beijing
    • 2Beijing Engineering Research Center of Food Environment and Health, Minzu University of China, Beijing
    • 3Section on Cellular Neurobiology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland
    JAMA Pediatr. 2016;170(11):1079-1086. doi:10.1001/jamapediatrics.2016.1626
    Key Points

    Question  Are peripheral blood levels of brain-derived neurotrophic factor (BDNF) altered in children with autism spectrum disorder (ASD)?

    Findings  In this meta-analysis of 19 studies with 2896 unique participants, children with ASD demonstrated significantly increased blood levels of BDNF compared with healthy control children.

    Meaning  This study provides a novel perspective into the etiology of ASD, and future investigations into BDNF levels for early diagnosis of the disease may be warranted.

    Abstract

    Importance  Accumulating evidence suggests that brain-derived neurotrophic factor (BDNF) may be implicated in the developmental outcomes of children with autism spectrum disorder (ASD).

    Objective  To use meta-analysis to determine whether children with ASD have altered peripheral blood levels of BDNF.

    Data Source  A systematic search of PubMed, PsycINFO, and Web of Science was performed for English-language literature through February 7, 2016. The search terms included brain-derived neurotrophic factor or BDNF in combination with autism, without year restriction. Two additional records were retrieved after a review of the reference lists of selected articles.

    Study Selection  Studies were included if they provided data on peripheral blood levels of BDNF in children with ASD and healthy control children. Studies that included adults or with overlapping samples were excluded.

    Data Extraction and Synthesis  Data were extracted by 2 independent observers from 19 included studies. Data were pooled using a random-effects model with Comprehensive Meta-analysis software.

    Main Outcomes and Measures  Blood levels of BDNF in children with ASD compared with healthy controls. Altered levels of BDNF were hypothesized to be related to ASD.

    Results  This meta-analysis included 19 studies with 2896 unique participants. Random-effects meta-analysis of all 19 studies showed that children with ASD had significantly increased peripheral blood levels of BDNF compared with healthy controls (Hedges g, 0.490; 95% CI, 0.185-0.794; P = .002). Subgroup analyses in 4 studies revealed that neonates diagnosed with ASD later in life had no association with blood levels of BDNF (Hedges g, 0.384; 95% CI, −0.244 to 1.011; P = .23), whereas children in the nonneonate ASD group (15 studies) demonstrated significantly increased BDNF levels compared with healthy controls (Hedges g, 0.524; 95% CI, 0.206 to 0.842; P = .001). Further analysis showed that children in the nonneonate ASD group had increased BDNF levels in serum (10 studies) (Hedges g, 0.564; 95% CI, 0.168 to 0.960; P = .005) but not in plasma (5 studies) (Hedges g, 0.436; 95% CI, −0.176 to 1.048; P = .16). Meta-regression analyses revealed that sample size had a moderating effect on the outcome of the meta-analysis in the nonneonate group. In addition, no publication bias was found in the meta-analysis.

    Conclusions and Relevance  Children with ASD have increased peripheral blood levels of BDNF, strengthening the clinical evidence of an abnormal neurotrophic factor profile in this population.

    Introduction

    Autism spectrum disorder (ASD) consists of a group of complex neurodevelopmental disorders characterized by impairments in social interaction, verbal and nonverbal communication, and repetitive behaviors.1 According to the DSM-5,2 ASD includes autistic disorder, childhood disintegrative disorder, pervasive developmental disorder not otherwise specified, and Asperger syndrome. According to the Centers for Disease Control and Prevention,3 about 1 in 68 children in the United States have been diagnosed with ASD as of 2014, and the disease is 4 to 5 times more common in male than in female children.4 The prevalence of ASD has increased dramatically during the past several decades, and it affected 21.7 million people globally as of 2013.5 No explanation has been established for this continued increase, although improved diagnosis and environmental influences are considered to be 2 reasons.6

    Despite the fact that the research into ASD has greatly increased during the past several decades, the exact etiology of ASD remains poorly understood. However, the complexity of the disease is considered to arise from the interactions between genetic mutations and environmental influences.7 In addition, findings from analyses of postmortem and peripheral tissue and molecular genetic studies led to the hypothesis that neurotrophins—as crucial moderators of neurodevelopment and neuroplasticity—affect the pathophysiologic features of ASD.8-11

    A prominent and widely expressed neurotrophin family member is brain-derived neurotrophic factor (BDNF), which plays a key role in neuronal survival and growth, cell differentiation, synapse formation, synaptic plasticity, and cognitive functions.12-17 In addition, skewed BDNF expression has been found to be associated with several neuropsychiatric diseases, including depression,18 schizophrenia,19 and Alzheimer disease.20 In ASD, a substantial number of studies measured peripheral blood levels of BDNF in children with ASD as a window to understand what occurs in the brain because peripheral blood levels of BDNF are known to be highly correlated with BDNF levels in the brain.21,22 However, studies of the association between blood BDNF levels and ASD yielded inconsistent findings. Some studies10,23-25 demonstrated that children with ASD had significantly increased peripheral blood levels of BDNF compared with healthy control individuals, whereas other studies26,27 found no association between blood levels of BDNF and ASD. Furthermore, some groups reported decreased blood levels of BDNF in children with ASD.9,28 Given the inconsistent findings, a meta-analysis on this subject is warranted. In this study, we undertook what is, to our knowledge, the first systematic review with meta-analysis of studies measuring peripheral blood levels of BDNF in children with ASD compared with healthy controls and performed subgroup and meta-regression analyses to adjust for potential confounders affecting the outcome of the meta-analysis.

    Methods
    Search Strategy and Study Selection

    Two independent investigators (X.-Y.Q. and C.C.) performed a systematic search of peer-reviewed English-language articles using PubMed, PsycINFO, and Web of Science through February 7, 2016. The search terms used for the database search included brain-derived neurotrophic factor or BDNF in combination with autism, without year restriction. Original human studies that reported data on peripheral blood levels of BDNF in patients with ASD and healthy controls were included. Exclusion criteria were (1) adult study participants and (2) samples that overlapped with another study. This meta-analysis adhered to the guidelines recommended by Preferred Reporting Items for Systematic Reviews and Meta-analysis (the PRISMA statement).29

    Data Extraction

    Both investigators independently extracted the data from the included studies. Sample sizes, mean BDNF concentrations, SDs, and P values were extracted as primary outcomes to generate effect sizes. Data on sex distribution, mean patient age, publication year, sample source (serum or plasma sample or dried blood spot), and assay type were also extracted. Demographic and clinical characteristics of the included studies are summarized in the eTable in the Supplement.

    Statistical Analysis

    All statistical analyses were performed using Comprehensive Meta-analysis software (version 2; Biostat Inc). Most of the effect sizes were generated by sample sizes, mean BDNF concentrations, and SDs; the rest were generated by sample sizes and P values. One study reported P values as an inequality (P < .0001) rather than an exact value, and the P value was rounded down to the nearest .00009 to allow compatibility with the meta-analysis software. Effect sizes were calculated as standardized mean differences in BDNF levels between children with ASD and healthy controls and converted to the Hedges g statistic, which provides an unbiased effect size adjusted for sample size.30 The statistical difference of the pooled effect size was assessed using 95% CIs. We used a random-effects model for the meta-analysis because within-study and between-study confounders were hypothesized to result in a difference of the true effect size.31 We performed sensitivity analysis by removing 1 study at a time to assess whether a single study influenced the outcomes of the meta-analysis.

    We used the Cochran Q test32 to assess the statistical difference of heterogeneity among studies, and P < .10 was considered to be statistically significant. The impact of heterogeneity that reflects the inconsistent levels among studies was determined by the I2 index. An I2 index of 0.25, 0.50, and 0.75 would indicate small, moderate, and high levels of heterogeneity, respectively.33 We performed unrestricted maximum-likelihood random-effects meta-regressions of effect sizes34 to evaluate whether the theoretically relevant covariates, including sample size, sex distribution (proportion of male), mean age of the children with ASD, and publication year of each study served as confounders to influence the effect size. Under the random-effects model, study weight is the inverse of the total variance (within-study variance plus between-study variance, τ2 statistic).

    Publication Bias

    Publication bias was first visually inspected by funnel plots generated by plotting the effect sizes against the precision (inverse of SE) for each study. The Egger test35 was assessed to determine the statistical significance of publication bias, which indicates the degree of funnel plot asymmetry. We further investigated publication bias using the classic fail-safe N method,36 which computes the number of missing studies (with a mean effect of zero) that would need to be added to the analysis to yield a statistically nonsignificant overall effect. P < .05 was considered statistically significantly different in all analyses of this study except where noted.

    Results

    We first performed a systematic search, which produced 166 records from PubMed, 105 records from PsycINFO, 338 records from Web of Science, and 2 additional records identified from the reference lists of retrieved articles. After reading the titles and abstracts, 28 appropriate articles were identified for full-text analysis. We scrutinized the 28 articles and excluded 9 articles for lack of necessary data on BDNF levels,37,38 patient samples that overlapped with another study,39,40 adult study participants,41,42 data reporting BDNF messenger RNA levels,43,44 and no data reported on healthy controls.45 Thus, 19 studies with 2896 unique participants9,10,23-28,46-56 met criteria for inclusion in the present meta-analysis (Figure 1).

    Main Association of ASD in Children With BDNF Level

    We performed a random-effects meta-analysis on the extracted 19 studies encompassing 1411 children with ASD and 1485 healthy controls. The results showed that peripheral blood levels of BDNF were significantly increased in children with ASD when compared with levels in healthy controls (Hedges g = 0.490; 95% CI, 0.185-0.794; P = .002) (Figure 2). Sensitivity analysis indicated that no individual study significantly influenced the significant difference on blood BDNF levels between children with ASD and healthy controls (eFigure 1 in the Supplement). However, significant heterogeneity among studies was found in our meta-analysis (Q18 = 211.853; I2 = 91.504; P < .001).

    Investigation of Heterogeneity

    To investigate the potential sources that explained the high levels of heterogeneity, we performed subgroup analyses. Four of 19 studies25,26,46,49 in the meta-analysis analyzed dried blood spot samples from neonates diagnosed with ASD later in life; the other 15 studies9,10,23,27,28,47,48,50-56 analyzed blood samples (serum or plasma) from children with ASD after they were diagnosed (nonneonate ASD group). Therefore, we performed a meta-analysis based on whether blood samples were obtained from neonates. As shown in Figure 3, the nonneonate ASD group had significantly increased blood BDNF levels compared with healthy controls (Hedges g = 0.524; 95% CI, 0.206-0.842; P = .001). In contrast, neonates diagnosed with ASD later in life did not show a difference in blood BDNF levels compared with healthy controls (Hedges g = 0.384; 95% CI, −0.244 to 1.011; P = .23). However, high levels of heterogeneity among studies were still found in the neonate (Q3 = 57.526; I2 = 94.785; P < .001) and nonneonate (Q14 = 94.338; I2 = 85.160; P < .001) studies, although the heterogeneity was slightly reduced in the nonneonate group. Further analysis of the nonneonate studies showed that children with ASD had significantly increased BDNF levels in serum samples (Figure 4) (10 studies9,10,24,47,51-56) (Hedges g = 0.564; 95% CI, 0.168-0.960; P = .005) but not in plasma samples (5 studies23,27,28,48,50) (Hedges g = 0.436; 95% CI, −0.176 to 1.048; P = .16) compared with healthy controls. Again, we found high levels of heterogeneity among studies measuring serum (Q9 = 58.236; I2 = 84.546; P < .001) or plasma (Q4 = 36.049; I2 = 88.904; P < .001) BDNF levels.

    We next performed meta-regression analyses to test whether the continuous variables, including age and sex of the patients, sample size, and publication year of each study, could explain the high levels of heterogeneity among studies. These tested variables did not show moderating effects on the outcome of the meta-analysis (eFigure 2 in the Supplement) in the included 19 studies.9,10,23-28,46-56 Meta-regression in the nonneonate studies showed that sample size positively correlated with the effect sizes (eFigure 3A in the Supplement), indicating that sample size had moderating effects on the outcome of the meta-analysis in the nonneonate studies. In addition, we found a nonsignificant negative correlation between age and effect size in the nonneonate studies (eFigure 3B in the Supplement), suggesting that age might be a confounding factor for studies analyzing BDNF levels in children with ASD. Sex (eFigure 3C in the Supplement) and publication year (eFigure 3D in the Supplement) did not have moderating effects in the nonneonate studies.

    Publication Bias

    Visual inspection of funnel plots (Figure 5) showed no publication bias in this meta-analysis, which was confirmed by results of the Egger test (t17 = 1.64; P = .12). In addition, the classic fail-safe N method indicated that 466 missing studies would be required to make P > .05 for the meta-analysis, which further confirmed that the positive results of our meta-analysis are unlikely to be caused by publication bias.

    Discussion

    The strength of this work is that the meta-analysis included a large body of studies and demonstrated that peripheral blood levels of BDNF are significantly increased in children with ASD compared with healthy controls. Through sensitivity analysis, we conclude that our results were not unduly influenced by a particular study. In addition, we found no indication that the outcome of the present meta-analysis was caused by publication bias. To the best of our knowledge, this meta-analysis is the first to be performed on this subject. Because the role of circulating BDNF levels in children with ASD has been controversial for more than a decade, and therefore their significance with respect to the etiology of ASD remains elusive, this study provides clinical evidence that children with ASD have increased peripheral blood levels of BDNF. This finding offers a novel perspective into a potential molecular pathway that contributes to the developmental outcomes of ASD.

    Although high levels of between-study heterogeneity were found in the meta-analysis, the strength of this work is that we used subgroup and meta-regression analyses to adjust for potential moderators. Subgroup analyses showed that neonates diagnosed with ASD later in life did not have altered blood BDNF levels, whereas children in the nonneonate ASD group manifested increased peripheral blood levels of BDNF. Although the levels of between-study heterogeneity were only slightly reduced in the nonneonate studies, meta-regression analysis indicated that age might have a moderating effect on the outcome of the meta-analysis, thus raising awareness that future work analyzing BDNF in children with ASD would need to consider this potential confounding factor. Meta-regression analysis also revealed that sample size was positively correlated with the effect size in the nonneonate group, which is reasonable because larger sampling of cases and controls would be a stronger approach for clinical studies.

    Most of the studies included in the meta-analysis analyzed serum BDNF levels, which provides us a sufficient number of studies to conclude that children with ASD are accompanied by the increased serum BDNF levels. Evidence suggests that the changes in serum BDNF levels may reflect what occurs in the brain. In particular, reduced serum BDNF concentrations in several neuropsychiatric disorders, such as schizophrenia19 and Alzheimer disease,57,58 were consistently observed by biochemical analyses of BDNF levels in the brains.20,59,60 In addition, a recent meta-analysis demonstrated that serum BDNF levels are low in untreated patients with depression and restored by antidepressant treatments,18 which is also consistent with what occurs in the brain.61,62 Therefore, the increased serum BDNF levels in children with ASD in our meta-analysis very likely reflects what occurs in the brains of children with ASD. In support of this theory, 1 study8 showed that BDNF levels were significantly increased in the postmortem brains from patients with ASD compared with controls.

    Notwithstanding its significant strengths, this meta-analysis has some inherent limitations. First, the meta-analysis of peripheral blood levels of BDNF in children with ASD compared with healthy controls provides us a pooled result originating from cross-sectional studies. Therefore, whether an increase in BDNF levels is a cause for ASD development or an epiphenomenon, such as a counterbalance consequence of ASD development, remains unclear. However, the hypothesis that the hyperactivity of BDNF contributes to the development of ASD is plausible in view of the early postnatal neocortical overgrowth that has been proposed as a core mechanism of ASD,63 and BDNF is a well-known potent stimulator of neuronal growth.13 The idea that increased BDNF levels affect the neuropathologic features of ASD is further supported by the fact that children with ASD have excess synapses64 because a major function of BDNF is to enhance synapse formation.12 In addition, in vitro and in vivo studies suggest that BDNF contributes to epileptogenesis.65 Thus, the hyperactivity of BDNF may provide a common molecular pathway for the long-standing association between ASD and epilepsy.66

    Second, the numbers of studies that analyzed dried blood spot levels of BDNF from neonates and plasma BDNF levels from nonneonates were small. However, only 1 study25 demonstrated that neonates diagnosed with ASD later in life had increased peripheral blood BDNF levels compared with healthy controls, whereas the other 3 studies26,46,49 did not find any differences. Therefore, we can reasonably conclude that neonates diagnosed with ASD later in life did not have the altered blood levels of BDNF. Although the random-effects meta-analysis did not show a significant association between plasma BDNF levels and children with ASD, after excluding an outlier28 (which had a small sample size) from the 5 studies measuring plasma BDNF levels, the meta-analysis showed that children with ASD had significantly increased plasma levels of BDNF compared with healthy controls (Hedges g = 0.803; 95% CI, 0.398-1.209; P < .001). Therefore, children with ASD likely have increased plasma levels of BDNF, but further studies on plasma BDNF levels in this population are necessary to substantiate this idea.

    Third, cross-sectional studies do not indicate the developmental stage at which the manifestation of increased BDNF in children with ASD starts, and longitudinal studies on BDNF levels may be necessary to address this question. Finally, most commercially available kits do not differentiate between pro-BDNF and mature BDNF, as in the studies included herein. One exception is a study that showed that total BDNF levels, but not pro-BDNF levels, were significantly increased in children with ASD.23 Given the proposed opposing effects of pro-BDNF and mature BDNF (a major function of pro-BDNF is generally considered to be induction of apoptosis67), a study on whether the ratio of pro-BDNF to mature BDNF levels is altered in children with ASD would be interesting.

    Conclusions

    Our meta-analysis demonstrated increased peripheral blood levels of BDNF as a manifestation of ASD in children, which strengthens the clinical evidence of an abnormal neurotrophic factor profile in children with ASD. Thus, further investigations into BDNF levels as a potential key to early diagnosis and therapeutic target of ASD are warranted.

    Back to top
    Article Information

    Corresponding Author: Yong Cheng, PhD, Section on Cellular Neurobiology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, 49 Convent Dr, Bldg 49, Room 6C80, Bethesda, MD 20892 (chengy4@mail.nih.gov).

    Accepted for Publication: May 19, 2016.

    Published Online: September 19, 2016. doi:10.1001/jamapediatrics.2016.1626

    Author Contributions: Drs Feng and Qin contributed equally to this work. Dr Cheng had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Study concept and design: Qin, Feng, Cheng.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Feng, Cheng.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Qin, Cheng.

    Obtained funding: Qin, Feng, Cheng.

    Administrative, technical, or material support: Qin, Cheng.

    Study supervision: Cheng.

    Conflict of Interest Disclosures: None reported.

    Funding/Support: This study was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health; grants YLDX01013 and 2015MDTD13C from the Minzu University 985 Academic Team-building Fund; and grant B08044 from the 111 Project of China.

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    References
    1.
    Caronna  EB, Milunsky  JM, Tager-Flusberg  H.  Autism spectrum disorders: clinical and research frontiers.  Arch Dis Child. 2008;93(6):518-523.PubMedGoogle ScholarCrossref
    2.
    American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA: American Psychiatric Association; 2013.
    3.
    Ha  S, Sohn  IJ, Kim  N, Sim  HJ, Cheon  KA.  Characteristics of brains in autism spectrum disorder: structure, function and connectivity across the lifespan.  Exp Neurobiol. 2015;24(4):273-284.PubMedGoogle ScholarCrossref
    4.
    Giarelli  E, Wiggins  LD, Rice  CE,  et al.  Sex differences in the evaluation and diagnosis of autism spectrum disorders among children.  Disabil Health J. 2010;3(2):107-116.PubMedGoogle ScholarCrossref
    5.
    Global Burden of Disease Study 2013 Collaborators.  Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013.  Lancet. 2015;386(9995):743-800.PubMedGoogle ScholarCrossref
    6.
    Mizejewski  GJ.  Biomarker testing for suspected autism spectrum disorder in early childhood: is such testing now feasible?  Biomark Med. 2012;6(4):503-506.PubMedGoogle ScholarCrossref
    7.
    Fakhoury  M.  Autistic spectrum disorders: a review of clinical features, theories and diagnosis.  Int J Dev Neurosci. 2015;43:70-77.PubMedGoogle ScholarCrossref
    8.
    Garcia  KL, Yu  G, Nicolini  C,  et al.  Altered balance of proteolytic isoforms of pro–brain-derived neurotrophic factor in autism.  J Neuropathol Exp Neurol. 2012;71(4):289-297.PubMedGoogle ScholarCrossref
    9.
    Taurines  R, Segura  M, Schecklmann  M,  et al.  Altered peripheral BDNF mRNA expression and BDNF protein concentrations in blood of children and adolescents with autism spectrum disorder.  J Neural Transm (Vienna). 2014;121(9):1117-1128.PubMedGoogle ScholarCrossref
    10.
    Wang  M, Chen  H, Yu  T, Cui  G, Jiao  A, Liang  H.  Increased serum levels of brain-derived neurotrophic factor in autism spectrum disorder.  Neuroreport. 2015;26(11):638-641.PubMedGoogle ScholarCrossref
    11.
    Yoo  HJ, Yang  SY, Cho  IH, Park  M, Kim  SA.  Polymorphisms of BDNF gene and autism spectrum disorders: family based association study with Korean trios.  Psychiatry Investig. 2014;11(3):319-324.PubMedGoogle ScholarCrossref
    12.
    Bamji  SX, Rico  B, Kimes  N, Reichardt  LF.  BDNF mobilizes synaptic vesicles and enhances synapse formation by disrupting cadherin-β-catenin interactions.  J Cell Biol. 2006;174(2):289-299.PubMedGoogle ScholarCrossref
    13.
    Binder  DK, Scharfman  HE.  Brain-derived neurotrophic factor.  Growth Factors. 2004;22(3):123-131.PubMedGoogle ScholarCrossref
    14.
    Greenberg  ME, Xu  B, Lu  B, Hempstead  BL.  New insights in the biology of BDNF synthesis and release: implications in CNS function.  J Neurosci. 2009;29(41):12764-12767.PubMedGoogle ScholarCrossref
    15.
    Hempstead  BL.  Brain-derived neurotrophic factor: three ligands, many actions.  Trans Am Clin Climatol Assoc. 2015;126:9-19.PubMedGoogle Scholar
    16.
    Lu  B, Nagappan  G, Guan  X, Nathan  PJ, Wren  P.  BDNF-based synaptic repair as a disease-modifying strategy for neurodegenerative diseases.  Nat Rev Neurosci. 2013;14(6):401-416.PubMedGoogle ScholarCrossref
    17.
    Lu  Y, Christian  K, Lu  B.  BDNF: a key regulator for protein synthesis-dependent LTP and long-term memory?  Neurobiol Learn Mem. 2008;89(3):312-323.PubMedGoogle ScholarCrossref
    18.
    Molendijk  ML, Spinhoven  P, Polak  M, Bus  BA, Penninx  BW, Elzinga  BM.  Serum BDNF concentrations as peripheral manifestations of depression: evidence from a systematic review and meta-analyses on 179 associations (N=9484).  Mol Psychiatry. 2014;19(7):791-800.PubMedGoogle ScholarCrossref
    19.
    Fernandes  BS, Steiner  J, Berk  M,  et al.  Peripheral brain-derived neurotrophic factor in schizophrenia and the role of antipsychotics: meta-analysis and implications.  Mol Psychiatry. 2015;20(9):1108-1119.PubMedGoogle ScholarCrossref
    20.
    Song  JH, Yu  JT, Tan  L.  Brain-derived neurotrophic factor in Alzheimer’s disease: risk, mechanisms, and therapy.  Mol Neurobiol. 2015;52(3):1477-1493.PubMedGoogle ScholarCrossref
    21.
    Fernandes  BS, Berk  M, Turck  CW, Steiner  J, Gonçalves  CA.  Decreased peripheral brain-derived neurotrophic factor levels are a biomarker of disease activity in major psychiatric disorders: a comparative meta-analysis.  Mol Psychiatry. 2014;19(7):750-751.PubMedGoogle ScholarCrossref
    22.
    Pan  W, Banks  WA, Fasold  MB, Bluth  J, Kastin  AJ.  Transport of brain-derived neurotrophic factor across the blood-brain barrier.  Neuropharmacology. 1998;37(12):1553-1561.PubMedGoogle ScholarCrossref
    23.
    Bryn  V, Halvorsen  B, Ueland  T,  et al.  Brain derived neurotrophic factor (BDNF) and autism spectrum disorders (ASD) in childhood.  Eur J Paediatr Neurol. 2015;19(4):411-414.PubMedGoogle ScholarCrossref
    24.
    Connolly  AM, Chez  M, Streif  EM,  et al.  Brain-derived neurotrophic factor and autoantibodies to neural antigens in sera of children with autistic spectrum disorders, Landau-Kleffner syndrome, and epilepsy.  Biol Psychiatry. 2006;59(4):354-363.PubMedGoogle ScholarCrossref
    25.
    Nelson  KB, Grether  JK, Croen  LA,  et al.  Neuropeptides and neurotrophins in neonatal blood of children with autism or mental retardation.  Ann Neurol. 2001;49(5):597-606.PubMedGoogle ScholarCrossref
    26.
    Nelson  PG, Kuddo  T, Song  EY,  et al.  Selected neurotrophins, neuropeptides, and cytokines: developmental trajectory and concentrations in neonatal blood of children with autism or Down syndrome.  Int J Dev Neurosci. 2006;24(1):73-80.PubMedGoogle ScholarCrossref
    27.
    Rodrigues  DH, Rocha  NP, Sousa  LF, Barbosa  IG, Kummer  A, Teixeira  AL.  Circulating levels of neurotrophic factors in autism spectrum disorders.  Neuro Endocrinol Lett. 2014;35(5):380-384.PubMedGoogle Scholar
    28.
    Ray  B, Long  JM, Sokol  DK, Lahiri  DK.  Increased secreted amyloid precursor protein-α (sAPPα) in severe autism: proposal of a specific, anabolic pathway and putative biomarker.  PLoS One. 2011;6(6):e20405.PubMedGoogle ScholarCrossref
    29.
    Liberati  A, Altman  DG, Tetzlaff  J,  et al.  The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.  J Clin Epidemiol. 2009;62(10):e1-e34.PubMedGoogle ScholarCrossref
    30.
    Hedges  LV, Olkin  I.  Statistical Methods for Meta-analysis. New York, NY: Academic Press; 1985.
    31.
    Baumeister  D, Akhtar  R, Ciufolini  S, Pariante  CM, Mondelli  V.  Childhood trauma and adulthood inflammation: a meta-analysis of peripheral C-reactive protein, interleukin-6 and tumour necrosis factor-α.  Mol Psychiatry. 2015;21(5):642-649.PubMedGoogle ScholarCrossref
    32.
    Bowden  J, Tierney  JF, Copas  AJ, Burdett  S.  Quantifying, displaying and accounting for heterogeneity in the meta-analysis of RCTs using standard and generalised Q statistics.  BMC Med Res Methodol. 2011;11:41.PubMedGoogle ScholarCrossref
    33.
    Masi  A, Quintana  DS, Glozier  N, Lloyd  AR, Hickie  IB, Guastella  AJ.  Cytokine aberrations in autism spectrum disorder: a systematic review and meta-analysis.  Mol Psychiatry. 2015;20(4):440-446.PubMedGoogle ScholarCrossref
    34.
    Lau  J, Ioannidis  JP, Schmid  CH.  Quantitative synthesis in systematic reviews.  Ann Intern Med. 1997;127(9):820-826.PubMedGoogle ScholarCrossref
    35.
    Egger  M, Davey Smith  G, Schneider  M, Minder  C.  Bias in meta-analysis detected by a simple, graphical test.  BMJ. 1997;315(7109):629-634.PubMedGoogle ScholarCrossref
    36.
    Soeken  KL, Sripusanapan  A.  Assessing publication bias in meta-analysis.  Nurs Res. 2003;52(1):57-60.PubMedGoogle ScholarCrossref
    37.
    Kazim  SF, Cardenas-Aguayo  MdelC, Arif  M,  et al.  Sera from children with autism induce autistic features which can be rescued with a CNTF small peptide mimetic in rats.  PLoS One. 2015;10(3):e0118627.PubMedGoogle ScholarCrossref
    38.
    Riikonen  R.  Neurotrophic factors in the pathogenesis of Rett syndrome.  J Child Neurol. 2003;18(10):693-697.PubMedGoogle ScholarCrossref
    39.
    Abdallah  MW, Pearce  BD, Larsen  N,  et al.  Amniotic fluid MMP-9 and neurotrophins in autism spectrum disorders: an exploratory study.  Autism Res. 2012;5(6):428-433.PubMedGoogle ScholarCrossref
    40.
    Meng  WD, Sun  SJ, Yang  J, Chu  RX, Tu  W, Liu  Q.  Elevated serum brain-derived neurotrophic factor (BDNF) but not BDNF gene Val66Met polymorphism is associated with autism spectrum disorders [published online January 28, 2016].  Mol Neurobiol. doi:10.1007/s12035-016-9721-9PubMedGoogle Scholar
    41.
    Hashimoto  K, Iwata  Y, Nakamura  K,  et al.  Reduced serum levels of brain-derived neurotrophic factor in adult male patients with autism.  Prog Neuropsychopharmacol Biol Psychiatry. 2006;30(8):1529-1531.PubMedGoogle ScholarCrossref
    42.
    Miyazaki  K, Narita  N, Sakuta  R,  et al.  Serum neurotrophin concentrations in autism and mental retardation: a pilot study.  Brain Dev. 2004;26(5):292-295.PubMedGoogle ScholarCrossref
    43.
    Nishimura  K, Nakamura  K, Anitha  A,  et al.  Genetic analyses of the brain-derived neurotrophic factor (BDNF) gene in autism.  Biochem Biophys Res Commun. 2007;356(1):200-206.PubMedGoogle ScholarCrossref
    44.
    Segura  M, Pedreño  C, Obiols  J,  et al.  Neurotrophin blood-based gene expression and social cognition analysis in patients with autism spectrum disorder.  Neurogenetics. 2015;16(2):123-131.PubMedGoogle ScholarCrossref
    45.
    Spratt  EG, Granholm  AC, Carpenter  LA,  et al.  Pilot study and review: physiological differences in BDNF, a potential biomarker in males and females with autistic disorder.  Int Neuropsychiatr Dis J. 2015;3(1):19-26.PubMedGoogle ScholarCrossref
    46.
    Abdallah  MW, Mortensen  EL, Greaves-Lord  K,  et al.  Neonatal levels of neurotrophic factors and risk of autism spectrum disorders.  Acta Psychiatr Scand. 2013;128(1):61-69.PubMedGoogle ScholarCrossref
    47.
    Al-Ayadhi  LY.  Relationship between Sonic hedgehog protein, brain-derived neurotrophic factor and oxidative stress in autism spectrum disorders.  Neurochem Res. 2012;37(2):394-400.PubMedGoogle ScholarCrossref
    48.
    Correia  CT, Coutinho  AM, Sequeira  AF,  et al.  Increased BDNF levels and NTRK2 gene association suggest a disruption of BDNF/TrkB signaling in autism.  Genes Brain Behav. 2010;9(7):841-848.PubMedGoogle ScholarCrossref
    49.
    Croen  LA, Goines  P, Braunschweig  D,  et al.  Brain-derived neurotrophic factor and autism: maternal and infant peripheral blood levels in the Early Markers for Autism (EMA) Study.  Autism Res. 2008;1(2):130-137.PubMedGoogle ScholarCrossref
    50.
    Enstrom  A, Onore  C, Tarver  A,  et al.  Peripheral blood leukocyte production of BDNF following mitogen stimulation in early onset and regressive autism.  Am J Biochem Biotechnol. 2008;4(2):121-129.Google ScholarCrossref
    51.
    Halepoto  DM, Bashir  S, Zeina  R, Al-Ayadhi  LY.  Correlation between Hedgehog (Hh) protein family and brain-derived neurotrophic factor (BDNF) in autism spectrum disorder (ASD).  J Coll Physicians Surg Pak. 2015;25(12):882-885.PubMedGoogle Scholar
    52.
    Kasarpalkar  NJ, Kothari  ST, Dave  UP.  Brain-derived neurotrophic factor in children with autism spectrum disorder.  Ann Neurosci. 2014;21(4):129-133.PubMedGoogle Scholar
    53.
    Katoh-Semba  R, Wakako  R, Komori  T,  et al.  Age-related changes in BDNF protein levels in human serum: differences between autism cases and normal controls.  Int J Dev Neurosci. 2007;25(6):367-372.PubMedGoogle ScholarCrossref
    54.
    Ricci  S, Businaro  R, Ippoliti  F,  et al.  Altered cytokine and BDNF levels in autism spectrum disorder.  Neurotox Res. 2013;24(4):491-501.PubMedGoogle ScholarCrossref
    55.
    Zhang  QB, Jiang  LF, Kong  LY, Lu  YJ.  Serum brain-derived neurotrophic factor levels in Chinese children with autism spectrum disorders: a pilot study.  Int J Dev Neurosci. 2014;37:65-68.PubMedGoogle ScholarCrossref
    56.
    Mansour  M, Mohamed  A, Azam  H, Henedy  M.  Brain derived neurotrophic factor in autism.  Curr Psychiatr. 2010;17(1):23-29.Google Scholar
    57.
    Liu  YH, Jiao  SS, Wang  YR,  et al.  Associations between ApoEε4 carrier status and serum BDNF levels: new insights into the molecular mechanism of ApoEε4 actions in Alzheimer’s disease.  Mol Neurobiol. 2015;51(3):1271-1277.PubMedGoogle ScholarCrossref
    58.
    Qin  X-Y, Cao  C, Cawley  NX,  et al.  Decreased peripheral brain-derived neurotrophic factor levels in Alzheimer’s disease: a meta-analysis study (N=7277) [published online April 26, 2016].  Mol Psychiatry. doi:10.1038/mp.2016.62PubMedGoogle Scholar
    59.
    Pandya  CD, Kutiyanawalla  A, Pillai  A.  BDNF-TrkB signaling and neuroprotection in schizophrenia.  Asian J Psychiatr. 2013;6(1):22-28.PubMedGoogle ScholarCrossref
    60.
    Weickert  CS, Hyde  TM, Lipska  BK, Herman  MM, Weinberger  DR, Kleinman  JE.  Reduced brain-derived neurotrophic factor in prefrontal cortex of patients with schizophrenia.  Mol Psychiatry. 2003;8(6):592-610.PubMedGoogle ScholarCrossref
    61.
    Chen  B, Dowlatshahi  D, MacQueen  GM, Wang  JF, Young  LT.  Increased hippocampal BDNF immunoreactivity in subjects treated with antidepressant medication.  Biol Psychiatry. 2001;50(4):260-265.PubMedGoogle ScholarCrossref
    62.
    Dunham  JS, Deakin  JF, Miyajima  F, Payton  A, Toro  CT.  Expression of hippocampal brain-derived neurotrophic factor and its receptors in Stanley consortium brains.  J Psychiatr Res. 2009;43(14):1175-1184.PubMedGoogle ScholarCrossref
    63.
    Chomiak  T, Hu  B.  Alterations of neocortical development and maturation in autism: insight from valproic acid exposure and animal models of autism.  Neurotoxicol Teratol. 2013;36:57-66.PubMedGoogle ScholarCrossref
    64.
    Tang  G, Gudsnuk  K, Kuo  SH,  et al.  Loss of mTOR-dependent macroautophagy causes autistic-like synaptic pruning deficits.  Neuron. 2014;83(5):1131-1143.PubMedGoogle ScholarCrossref
    65.
    Binder  DK, Croll  SD, Gall  CM, Scharfman  HE.  BDNF and epilepsy: too much of a good thing?  Trends Neurosci. 2001;24(1):47-53.PubMedGoogle ScholarCrossref
    66.
    Tuchman  R, Rapin  I.  Epilepsy in autism.  Lancet Neurol. 2002;1(6):352-358.PubMedGoogle ScholarCrossref
    67.
    Teng  HK, Teng  KK, Lee  R,  et al.  ProBDNF induces neuronal apoptosis via activation of a receptor complex of p75NTR and sortilin.  J Neurosci. 2005;25(22):5455-5463.PubMedGoogle ScholarCrossref
    ×