Original Article

Molecular Psychiatry (2008) 13, 980–988; doi:10.1038/sj.mp.4002087; published online 25 September 2007

Association between the oxytocin receptor (OXTR) gene and autism: relationship to Vineland Adaptive Behavior Scales and cognition

All the authors contributed equally to the design, execution and preparation of this manuscript for publication.

E Lerer1, S Levi2, S Salomon3, A Darvasi4, N Yirmiya3 and R P Ebstein2,5

  1. 1Department of Human Genetics, Hebrew University, Jerusalem, Israel
  2. 2Department of Child Psychiatry, Hadassah-Hebrew University Medical School, Jerusalem, Israel
  3. 3Department of Psychology, Hebrew University, Jerusalem, Israel
  4. 4Faculty of Mathematics, The Alexander Silberman Institute of Life Sciences, Faculty of Science, Hebrew University, Jerusalem, Israel
  5. 5S Herzog Memorial Hospital, Jerusalem, Israel

Correspondence: Professor RP Ebstein, Department of Psychology, Scheinfeld Center of Human Genetics for the Social Sciences, Hebrew University, Jerusalem 91905, Israel. E-mail: ebstein@mscc.huji.ac.il

Received 22 February 2007; Revised 19 August 2007; Accepted 20 August 2007; Published online 25 September 2007.



Evidence both from animal and human studies suggests that common polymorphisms in the oxytocin receptor (OXTR) gene are likely candidates to confer risk for autism spectrum disorders (ASD). In lower mammals, oxytocin is important in a wide range of social behaviors, and recent human studies have shown that administration of oxytocin modulates behavior in both clinical and non-clinical groups. Additionally, two linkage studies and two recent association investigations also underscore a possible role for the OXTR gene in predisposing to ASD. We undertook a comprehensive study of all 18 tagged SNPs across the entire OXTR gene region identified using HapMap data and the Haploview algorithm. Altogether 152 subjects diagnosed with ASDs (that is, DSM IV autistic disorder or pervasive developmental disorder—NOS) from 133 families were genotyped (parents and affected siblings). Both individual SNPs and haplotypes were tested for association using family-based association tests as provided in the UNPHASED set of programs. Significant association with single SNPs and haplotypes (global P-values <0.05, following permutation test adjustment) were observed with ASD. Association was also observed with IQ and the Vineland Adaptive Behavior Scales (VABS). In particular, a five-locus haplotype block (rs237897-rs13316193-rs237889-rs2254298-rs2268494) was significantly associated with ASD (nominal global P=0.000019; adjusted global P=0.009) and a single haplotype (carried by 7% of the population) within that block showed highly significant association (P=0.00005). This is the third association study, in a third ethnic group, showing that SNPs and haplotypes in the OXTR gene confer risk for ASD. The current investigation also shows association with IQ and total VABS scores (as well as the communication, daily living skills and socialization subdomains), suggesting that this gene shapes both cognition and daily living skills that may cross diagnostic boundaries.


autism spectrum disorders, oxytocin receptor (OXTR), candidate gene, transmission disequilibrium, social adjustment, polymorphism



A large body of evidence shows that both oxytocin (OXT) and arginine vasopressin (AVP) peptides markedly modulate behavior and cognition across species.1 Interestingly, one facet of social memory, recognition of a familiar conspecific, is modulated by both OXT and AVP in lower mammals.2 This facet of social cognition may be especially relevant to autism spectrum disorders (ASDs). The development of an OXT knockout mouse and the deficits observed in its behavior underlines the role of this hormone in particularly molding social memory.3 Moreover, the absence of exposure to OXT during development was associated with abnormalities in the development of emotional behavior.4 The role of OXT and AVP in shaping social communication in non-human species resonates with the core deficits in ASD, especially impairment in communication skills and social interactions.

In humans and other mammals, OXT and AVP are both nonapeptides synthesized in the hypothalamus that are released into the bloodstream via axon terminals in the posterior pituitary or neurohypophysis. These two peptides are closely related structurally, differing at only two amino acids and have a common evolutionary origin.5 Two genome-wide scans highlight the 3p25 region, containing the OXTR gene, as a putative linkage site for ASD.6, 7 Recently Wu et al.8 genotyped four single-nucleotide polymorphisms (SNPs) in OXTR in 195 Chinese Han trios and demonstrated association with ASD with both individual SNPs and haplotypes. Jacob et al.9 genotyped two OXTR SNPs and observed significant association with a single SNP (rs2254298) in Caucasian children and adolescents with autism.

As noted in a review by Insel et al.,10 OXT receptors are remarkably plastic, induced several fold after gonadal steroid administration and expressed in different brain regions even in closely related species. Moreover, OXT and AVP receptors are developmentally regulated and expressed more in the immature brain, and both peptides have been shown to have effects on neural development. OXT receptors are distributed in various brain areas11 that are associated with social behavior, including reproductive and parenting behaviors, affiliation and attachment, social memory and reactivity to social stress in non-human mammals.12, 13 Recent investigations extend the role of OXT in social cognition to healthy humans14, 15, 16, 17 as well as to subjects with ASD.18, 19, 20, 21, 22 A recent study shows that intranasal OXT in human modulates amygdala responses to facial expressions irrespective of their valence.23

The human OXTR is a 389-amino-acid polypeptide with seven transmembrane domains and belongs to the class I G protein-coupled receptor (GPCR) family.24 OXTR is located on chromosome 3p25 and spans ~19kbp and contains three introns and four exons.25 Several dozen SNPs have been identified within the gene by the National Center for Biotechnology Information database (dbSNP). To further explore the role of the OXTR gene in contributing risk to ASD, we undertook a comprehensive study using all 18 tagged SNPs across the entire OXTR gene region identified using HapMap data and the Haploview algorithm.26 Altogether 152 subjects diagnosed with ASD from 133 families were genotyped and both individual SNPs and haplotypes were tested for association with ASD. Moreover to better understand how variants in the OXTR mediate risk for ASD, we examined association between OXTR with intelligence quotient (IQ) and VABS (Vineland Adaptive Behavioral Scales) scores.27




Altogether 152 subjects with ASDs (128 men, 24 women) and their parents (133 mothers and 133 fathers) and unaffected siblings (146) were recruited. Nineteen of the families had two children diagnosed with ASD. Subjects were diagnosed with DSM IV autistic disorder (N=131) or pervasive developmental disorder-not otherwise specified (PDD-NOS; N=21). Families were recruited through treatment centers, special schools, the Israeli National Organization for Children with ASD and by acquaintance with other families who participated. Subjects were between the ages of 2 years and a month to 33 years and 8 months. All subjects tested negatively for Fragile-X using a PCR procedure with an ABI 310 DNA analyzer.28, 29, 30 In addition, none of the subjects was known (according to parents' reports) to have any other chromosomal abnormality or tuberous sclerosis.

Many of the subjects had a previous diagnosis within the ASD spectrum made by independent clinicians. However, for the purpose of this study, two trained clinicians confirmed the subjects' diagnosis of DSM IV autistic disorder or PDD-NOS. All subjects were diagnosed using the ASD Diagnostic Interview—Revised (ADI-R),31 and for 78 of them the ASD Diagnostic Observation Scale—Generic (ADOS-G)32 was also available.

The ADI-R is a standardized, semi-structured interview based on the ICD-10 definition of ASD (World Health Organization, 1993). Three areas of child functioning are assessed by this parent interview: communication and language; reciprocal social interaction; and repetitive, restrictive and stereotyped behavior. To receive a diagnosis of autistic disorder, one must meet the cutoff criteria for ASD on each of the three aforementioned areas and show evidence for developmental abnormality before the age of 36 months. To receive a diagnosis of PDD-NOS, one must meet the cutoff criteria for ASD for social interaction and for only one of the two other areas as proposed by Risi et al.33

The ADOS-G is a semi-structured, standardized assessment used for diagnosing children with ASDs. It is an assessment of the child's social interaction, communication, play and imaginative use of materials, through an interaction with a trained professional. Diagnostic algorithms are provided for ASD and for PDD-NOS. In the case of a discrepancy between the two diagnostic measures, one yielding a diagnosis of autistic disorder and the other of PDD-NOS, the final diagnosis was that of PDD-NOS, as proposed by Risi et al.33

In addition to the ADI-R, parents were also interviewed using the Vineland Adaptive Behavior Scales—Interview edition.34 The VABS27 is a structured interview administered to caregiver/s to assess the child's daily living skills on three domains: communication (receptive, expressive and written), daily living skills (personal, domestic and community), and socialization (interpersonal relationships, play and leisure time, and coping skills). In addition to the three domain scores, a total score reflecting overall functioning was also computed. The average total VABS scores were 52.21 (s.d.=19.88). The VABS has excellent levels of split-half, inter-rater and test–retest reliability for each domain used in the current work.27

The subjects' level of functioning was assessed using standard intelligence measures selected according to the subjects' age and abilities (that is, Wechsler Intelligence Scale for Children;35 Kaufman Assessment Battery for Children;36 Bayley Scales of Infant Development;37 Merrill–Palmer Scale of Mental Tests;38 Mullen Scales of Early Learning;39 Cattel Measurement of Intelligence of Infants and Young Children;40 Leiter International Performance Scale;41 or Stanford–Binet Intelligence Scale42). As some tests (for example, Mullen and Leiter) yield mental ages rather than IQ, mental age was transformed to an IQ estimate using the basic IQ formula: IQ=(Mental Age/Chronological Age)*100MA. Average IQ scores were 68.7 (s.d.=29.1).


DNA was extracted by Master Pure kit (Epicentre, Madison, WI, USA).

Single-nucleotide polymorphisms were identified by searching through the dbSNP public database (http://www.ncbi.nlm.nih.gov/SNP/). Altogether 18 tagged SNPs were identified across the gene region (19.21kbp) by Haploview (http://www.broad.mit.edu/mpg/haploview/).26 In our sample, a single SNP was not in Hardy–Weinberg equilibrium and one SNP was in complete LD with the other SNPs. Therefore, 16/18 tagged SNPs were included in the association test. The location of the SNPs assayed and their location along the OXTR genomic region and their relationship to the exon/intron boundaries of the gene are shown in Table 1 and Figure 1.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Top: Schematic representation of chromosome 3 and the oxytocin receptor (OXTR) gene with the location of the 16 tagging single-nucleotide polymorphisms (SNPs). Bottom: SNPs and haplotypes in the OXTR tested for association with autism spectrum disorder (ASD), intelligence quotient (IQ) and Vineland Adaptive Behavior Scales (VABS) scores. The solid bars indicate only significant associations from single SNPs (window 1) to six-locus haplotypes. Both nominal and adjusted significance levels are presented. The significance levels following the permutation test (1000 iterations) are shown in parentheses. Black blocks represent ASD, medium-shaded blocks represent VABS scores and light-shaded blocks represent IQ.

Full figure and legend (134K)

Genotyping of all SNPs was performed using the SNaPshot Method (Applied BioSystems, Foster City, CA, USA). This method relies upon the extension of a primer immediately adjacent to the SNP using fluorescently labeled ddNTPs. The fluorescently labeled extension primers can then be visualized by electrophoresis on a capillary ABI PRISM 310 automated sequencer. Amplification of the OXTR SNPs regions was amplified using the pairs of primers as described in Table 2.

PCR cycling conditions in the SNaPshot Method were as follows: samples were initially heated at 94°C for 5min followed by 35 cycles of 94°C (30s), 55°C (30s) 72°C (90s) and a final extension step of 72°C for 5min. After the first PCR cycle, the PCR product was cleaned with ExoSAP for 37°C for 30min and then at 80°C for 15min. The conditions for the second PCR were as follows: 96°C (10s), 50°C (5s) and 60°C (30s) for 25 cycles. The second PCR product was cleaned using shrimp alkaline phosphatase (SAP) initially at 37°C for 1h followed by 72°C for 15min.

The extension primers for SNaPshot Method are shown in Table 2. In the case of Mendelian error, the genotyping for the entire family was rechecked and if the error persisted the genotyping for that SNP was not included in the analysis. The overall Mendelian error rate following our double-check procedure was <0.5%.

Statistical methods

We used the logistic-based variant of the transmission disequilibrium test (TDT) the so-called ETDT43 to assess association (and linkage) without confounding effect of population stratification. The TDT, in its simplest version, compares, for one allele, the number of times this allele is transmitted to the number of times in which this allele is not transmitted to an affected offspring. Note that only heterozygous parents are informative. This approach can be extended to haplotypes. The various tests are implemented in the latest version (3.0.9) of UNPHASED (http://www.rfcgr.mrc.ac.uk/~fdudbrid/software/unphased/). UNPHASED44 is a suite of programs for association analysis of multilocus haplotypes from unphased genotype data.

Correction for multiple testing

Common methods used for controlling are the Bonferroni correction or more refined variations of the method. Applying the Bonferroni correction, which is a single-step procedure applied to a multiple-testing scenario with correlated hypotheses (for example, association analysis of multiple markers and gene–gene interaction), leads to conservative results. Thus, multistep procedures were developed to achieve higher power for correlated tests while controlling the family-wise error rate. We used the permutation test option as provided in UNPHASED to avoid spurious results and correct for multiple testing. Permutation test correction was performed using 1000 iterations (random permutations) and applied to correction of global P-values (the P-value generated by UNPHASED as an overall test of each block of haplotypes). In nuclear families, the permutations are generated by randomizing the transmission status of the parental haplotypes. In unrelated subjects, the trait values are randomly shuffled between subjects. The randomization is held constant over all analyses specified by the disease-, trait-, marker- and window options, as well as any tests selected by the individual, test-confounders and test-modifiers options. In each permutation, the minimum P-value is compared to the minimum P-value over all the analyses in the original data. This allows for multiple-testing corrections over all tests performed in a run. This procedure corrects for multiple testing but accounts for correlation between markers, hence is less conservative than a Bonferroni correction, which is appropriate for independent tests such as unlinked markers.

It should be noted that while this multiple-testing correction adjusts for the number of markers tested, it only adjusts the global P-value of the particular haplotype block tested within any window size. The permutation test we used does not correct for the overall number of haplotype blocks examined across all window sizes tested. Nevertheless, it should be considered that all haplotypes represented in sliding windows covering a given region are likely to be highly correlated with each other and are therefore unlikely to greatly inflate the number of independent tests undertaken in this study. Nevertheless, our results should hence be interpreted cautiously.



As shown in Table 1, single SNP analysis revealed two SNPs (rs2268494 intron 3 and rs1042778 exon 4 UTR) in nominal (no correction for multiple testing) association with ASD. Two SNPs were nominally associated with IQ (rs468301 intron 3 and rs1042778 exon 4) and two SNPs with total VABS scores (intron 3 UTR: rs468301 and exon 4 UTR: rs6770632). Five single SNPs showed significant association with the VABS communication subdomains, six SNPs with VABS daily living skills and one SNP with socialization (Table 1). One SNP (rs2254298) associated with VABS daily living skills and communication subdomain scores was previously identified to be associated with autism in an independent study by Jacob et al.9

We next proceeded to a multimarker association analysis using a sliding window approach and examined consecutive SNPs comprising haplotypes of increasing length from 2 to 8 SNPs along the gene and tested blocks of haplotpes for association with ASD, IQ and VABS scores. Haplotype analysis, as has been suggested on theoretical grounds,45 was more informative than single SNP analysis. A haplotype has a higher level of heterozygosity than individual single SNPs and the association study based on the haplotype may have an increased power for detecting disease association compared with SNP-based analyses. Significant haplotype association (significance levels for global P-values <0.05 following adjustment by permutation tests) was observed with ASD as well as between the two continuous measures, IQ and total VABS scores (Figure 1). Figure 1 shows only the significant (following permutation test adjustment) haplotype blocks and their location along the OXTR gene (see also Table 1 for exact SNP location). VABS' subdomain haplotypes are not presented for reasons of economy of presentation of data.

Since significant global P-value association was observed between haplotype blocks and ASD, we next proceeded to examine association between individual haplotypes of various lengths and ASD, IQ and VABS scores. Table 3 shows the association for individual haplotypes (within blocks) with ASD. Notably, for each significant haplotype the rs2254298 ‘G’ SNP was included in the haplotype. This is the SNP associated with autism in the study by Jacob et al.9 The longer (>3 locus) haplotypes associated with ASD were clustered close to the exon 3/intron 3 boundary (see Table 1 for locations) and showed strongest evidence for association. The five-locus haplotype A-T-T-G-A (rs237897-rs13316193-rs237889-rs2254298-rs2268494) showed highly significantly association with ASD (P=0.00005). Most of the longer haplotypes associated with ASD included as their first 5′ SNP, rs237897 that was the closest tagged SNP to the exon 3/intron 3 boundary. For eight-locus haplotypes nominal, although highly significant, association was observed between ASD (Likelihood ratio χ2=37.92, d.f.=15, global P-value=0.00093), IQ (Likelihood ratio χ2=52.09, d.f.=19, global P-value=0.00006) and total VABS (Likelihood ratio χ2=44.28, d.f.=19, global P-value=0.0009). Permutation tests were not successfully performed >6-locus haplotypes because of the computational demand of such an analysis. The first upstream 5′ SNP in all of these haplotype groups was rs237897. Individual haplotypes and their significance levels for eight-locus haplotypes associated with ASD are shown in Table 3.

We also examined the role that IQ played in the observed association between ASD and OXTR. As shown in Table 1, when IQ was entered as a covariate (‘confounding factor’ in the UNPHASED program), significant association is nonetheless observed for the two SNPS (rs2268494 intron 3 and rs1042778) with ASD. Similarly, we also examined the five-locus haplotype association (starting with SNP rs237897) with ASD when IQ is entered as a covariate in the test. Highly significant association was nonetheless observed with IQ was a covariate (without IQ as cv: Likelihood ratio χ2=43.32, d.f.=12, nominal P-value=0.000019; with IQ as cv: Likelihood ratio χ2=41.37, d.f.=14, nominal P-value=0.00015). Therefore, it appears that the association between ASD and OXTR is not simply related to an effect of the OXTR on cognition but appears to be related at least partially to deficits in daily living skills.



The current investigation confirms and extends the findings of Wu et al.8 in a Han Chinese population and Jacob et al.9 in a Caucasian population who first demonstrated provisional association between OXTR and ASD. Whereas the Wu et al.8 genotyped only four SNPs and Jacob et al.9 two SNPs, we used Haploview to identify all tagging SNPs across the gene region towards implementing a rational and cost-effective genotyping approach. Interestingly, the rs2254298 SNP associated with autism (G allele overtransmitted) in the the study by Jacob et al.,9 although not by itself significantly associated with ASD in the current investigation (although G in our study is also overtransmitted), nonetheless was a core SNP in the haplotypes (and the G allele was the ‘risk’ allele) that showed significant association with ASD (Table 3). Additionally, this SNP was significantly associated with the VABS subdomain daily living skills scores. A single SNP, rs4564970, located in the 5′-flanking region in the vicinity of a number of transcription factors, was not associated with ASD. Sliding-window haplotype analysis by maximizing the genetic information extracted from the 16 genotyped tagged SNPS showed significant association between haplotypes and both ASD as well as IQ and VABS scores. The association between OXTR haplotypes and IQ and VABS scores suggests the notion that these aspects of cognition and daily living skills, respectively, contribute to the association between this gene and ASD. It should be noted, however, that the associations between OXTR haplotypes with IQ and the VABS are not necessarily specific to ASD and future research should address this issue by also examining other diagnostic groups such as mental retardation. Notably, the association demonstrated here between the OXTR gene and continuous measures including IQ and VABS scores suggests a molecular genetic validation to recent pharmacological manipulations using OXT in both syndromic18, 19, 46 and non-syndromic populations.14, 15, 16, 47

As shown in Figure 1 and Table 1, the OXTR gene spans 19kb and contains three introns and four exons. Exons 1 and 2 correspond to the 5′-non-coding region and contain a number of putative transcription factor sites. Exons 3 and 4 encode the amino acids, see review.24 Intron 3, which is the largest at 12kb, separates the coding region immediately after the putative transmembrane domain 6. Exon 4 contains the sequence encoding the seventh transmembrane domain, the COOH terminus, and the entire 3′-non-coding region, including the polyadenylation signals. Interestingly, the longer haplotypes showing significant association with DSM IV ASD tended to cluster in the 5′-flanking region of the gene at the exon 3 coding region–intron 3 boundary. It is tempting to speculate that these SNPs/haplotypes may be in linkage disequilibrium with a critical splice site polymorphism or enchancer element in this region. Moreover, Mizumoto et al.48 reported that a genomic element within the third intron of the human oxytocin receptor gene may be involved in transcriptional suppression of the gene suggesting that some of the intronic SNPs that we observed in association with ASD may alter methylation sites that regulate gene expression.

The association observed between OXTR haplotypes and ASD, IQ and total VABS scores (as well subdomains) are consistent with a role for OXT in cognition and social communication in both humans and other species. de Wied49 pioneered animal studies showing that OXT modulates behavior, learning and memory. These early studies have been substantiated by a large body of experimental findings showing that OXT enhances social interactions such as pair bonding12, 13 and maternal behavior50 and inhibits HPA-induced stress responses.47

A function for OXT in conferring risk ASD is further strengthened by the studies of Hollander et al.,18, 19 who have presented evidence that OXT infusions administered to participants (fulfilling DSM criteria for autistic disorder or Asperger's syndrome) might facilitate social information processing as well as reduce repetitive behavior. In a recent study with evident inferences for ASD research, Domes et al.16 showed that intranasally administered OXT improved Reading the Mind in the Eyes Test51 that examines the ability to infer the mental state of others from social cues of the eye region. Interestingly, despite individual variability and overlapping group distributions midday plasma OXT were significantly lower in subjects with DSM IV autistic disorder than the normal group.21, 22 Indeed, several reviews have underscored the theoretical basis for this neuropeptides involvement in contributing to DSM IV ASD.10, 21, 52, 53, 54

In summary, a role for OXT in conferring risk for DSM IV ASD is suggested by two genome-wide scans of ASD,6, 7 the role of this nonapeptide in molding social memory and communication in non-human species,3 recent human studies showing a role of OXT in social cognition in both non-syndromic14, 15, 16, 17 and syndromic18, 19, 20, 21, 22 subjects and the current family-based association study that replicates and extends a previous study in a Han Chinese population8 and a Caucasian population.9 Importantly, the potential of intranasal administration of this and related peptides as suggested by Born et al.55 coupled with a pharmacogenetic approach, made feasible by the current investigation, offers a worthwhile strategy to explore a novel therapeutic window into the treatment of some of the intractable core social deficits associated with the ASD syndrome.



  1. Engelmann M, Wotjak CT, Neumann I, Ludwig M, Landgraf R. Behavioral consequences of intracerebral vasopressin and oxytocin: focus on learning and memory. Neurosci Biobehav Rev 1996; 20: 341–358. | Article | PubMed | ISI | ChemPort |
  2. Dantzer R, Bluthe RM, Koob GF, Le Moal M. Modulation of social memory in male rats by neurohypophyseal peptides. Psychopharmacology (Berl) 1987; 91: 363–368. | Article | PubMed | ChemPort |
  3. Winslow JT, Insel TR. The social deficits of the oxytocin knockout mouse. Neuropeptides 2002; 36: 221–229. | Article | PubMed | ISI | ChemPort |
  4. Winslow JT, Hearn EF, Ferguson J, Young LJ, Matzuk MM, Insel TR. Infant vocalization, adult aggression, and fear behavior of an oxytocin null mutant mouse. Horm Behav 2000; 37: 145–155. | Article | PubMed | ISI | ChemPort |
  5. Acher R. Chemistry of the neurohypophysial hormones: an example of molecular evolution. In: Knobil E, Sawyer W (eds). Handbook of Physiology, vol. 4. American Physiological Society: Washington, DC, 1974, pp 119–130.
  6. Lauritsen MB, Als TD, Dahl HA, Flint TJ, Wang AG, Vang M et al. A genome-wide search for alleles and haplotypes associated with autism and related pervasive developmental disorders on the Faroe Islands. Mol Psychiatry 2006; 11: 37–46. | Article | PubMed | ISI | ChemPort |
  7. McCauley JL, Li C, Jiang L, Olson LM, Crockett G, Gainer K et al. Genome-wide and ordered-subset linkage analyses provide support for autism loci on 17q and 19p with evidence of phenotypic and interlocus genetic correlates. BMC Med Genet 2005; 6: 1. | Article | PubMed | ChemPort |
  8. Wu S, Jia M, Ruan Y, Liu J, Guo Y, Shuang M et al. Positive association of the oxytocin receptor gene (OXTR) with autism in the Chinese Han population. Biol Psychiatry 2005; 58: 74–77. | Article | PubMed | ISI | ChemPort |
  9. Jacob S, Brune CW, Carter CS, Leventhal BL, Lord C, Cook Jr EH. Association of the oxytocin receptor gene (OXTR) in Caucasian children and adolescents with autism. Neurosci Lett 2007; 417: 6–9. | Article | PubMed | ISI | ChemPort |
  10. Insel TR, O'Brien DJ, Leckman JF. Oxytocin, vasopressin, and autism: is there a connection? Biol Psychiatry 1999; 45: 145–157. | Article | PubMed | ISI | ChemPort |
  11. Landgraf R, Neumann ID. Vasopressin and oxytocin release within the brain: a dynamic concept of multiple and variable modes of neuropeptide communication. Front Neuroendocrinol 2004; 25: 150–176. | Article | PubMed | ISI | ChemPort |
  12. Carter CS. Neuroendocrine perspectives on social attachment and love. Psychoneuroendocrinology 1998; 23: 779–818. | Article | PubMed | ISI | ChemPort |
  13. Young LJ, Wang Z. The neurobiology of pair bonding. Nat Neurosci 2004; 7: 1048–1054. | Article | PubMed | ISI | ChemPort |
  14. Kosfeld M, Heinrichs M, Zak PJ, Fischbacher U, Fehr E. Oxytocin increases trust in humans. Nature 2005; 435: 673–676. | Article | PubMed | ISI | ChemPort |
  15. Heinrichs M, Meinlschmidt G, Wippich W, Ehlert U, Hellhammer DH. Selective amnesic effects of oxytocin on human memory. Physiol Behav 2004; 83: 31–38. | Article | PubMed | ISI | ChemPort |
  16. Domes G, Heinrichs M, Michel A, Berger C, Herpertz SC. Oxytocin improves 'mind-reading' in humans. Biol Psychiatry 2007; 61: 731–733. | Article | PubMed | ISI | ChemPort |
  17. Kirsch P, Esslinger C, Chen Q, Mier D, Lis S, Siddhanti S et al. Oxytocin modulates neural circuitry for social cognition and fear in humans. J Neurosci 2005; 25: 11489–11493. | Article | PubMed | ISI | ChemPort |
  18. Hollander E, Novotny S, Hanratty M, Yaffe R, DeCaria CM, Aronowitz BR et al. Oxytocin infusion reduces repetitive behaviors in adults with autistic and Asperger's disorders. Neuropsychopharmacology 2003; 28: 193–198. | Article | PubMed | ISI | ChemPort |
  19. Hollander E, Bartz J, Chaplin W, Phillips A, Sumner J, Soorya L et al. Oxytocin increases retention of social cognition in autism. Biol Psychiatry 2007; 61: 498–503. | Article | PubMed | ISI | ChemPort |
  20. Modahl C, Fein D, Waterhouse L, Newton N. Does oxytocin deficiency mediate social deficits in autism? J Autism Dev Disord 1992; 22: 449–451. | Article | PubMed | ISI | ChemPort |
  21. Green L, Fein D, Modahl C, Feinstein C, Waterhouse L, Morris M. Oxytocin and autistic disorder: alterations in peptide forms. Biol Psychiatry 2001; 50: 609–613. | Article | PubMed | ISI | ChemPort |
  22. Modahl C, Green L, Fein D, Morris M, Waterhouse L, Feinstein C et al. Plasma oxytocin levels in autistic children. Biol Psychiatry 1998; 43: 270–277. | Article | PubMed | ISI | ChemPort |
  23. Domes G, Heinrichs RW, Glascher J, Buchel C, Braus DF, Herpertz SC. Oxytocin attenuates amygdala responses to emotional faces regardless of valence. Biol Psychiatry 2007 [E-pub ahead of print].
  24. Gimpl G, Fahrenholz F. The oxytocin receptor system: structure, function, and regulation. Physiol Rev 2001; 81: 629–683. | PubMed | ISI | ChemPort |
  25. Inoue T, Kimura T, Azuma C, Inazawa J, Takemura M, Kikuchi T et al. Structural organization of the human oxytocin receptor gene. J Biol Chem 1994; 269: 32451–32456. | PubMed | ISI | ChemPort |
  26. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263–265. | Article | PubMed | ISI | ChemPort |
  27. Sparrow SS, Balla DA, Cicchetti DV. Vineland Adaptive Behavior Scales. American Guidance Services: Minneapolis, MN, 1984.
  28. Haddad LA, Mingroni-Netto RC, Vianna-Morgante AM, Pena SD. A PCR-based test suitable for screening for fragile X syndrome among mentally retarded males. Hum Genet 1996; 97: 808–812. | Article | PubMed | ISI | ChemPort |
  29. Larsen LA, Gronskov K, Norgaard-Pedersen B, Brondum-Nielsen K, Hasholt L, Vuust J. High-throughput analysis of fragile X (CGG)n alleles in the normal and premutation range by PCR amplification and automated capillary electrophoresis. Hum Genet 1997; 100: 564–568. | Article | PubMed | ISI | ChemPort |
  30. O'Connell CD, Atha DH, Jakupciak JP, Amos JA, Richie K. Standardization of PCR amplification for fragile X trinucleotide repeat measurements. Clin Genet 2002; 61: 13–20. | Article | PubMed | ISI | ChemPort |
  31. Lord C, Rutter M, Le Couteur A. Autism Diagnostic Interview—Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord 1994; 24: 659–685. | Article | PubMed | ISI | ChemPort |
  32. Lord C, Risi S, Lambrecht L, Cook Jr EH, Leventhal BL, DiLavore PC et al. The autism diagnostic observation schedule—generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord 2000; 30: 205–223. | Article | PubMed | ISI | ChemPort |
  33. Risi S, Lord C, Corsello C, Chrysler C, Szatmary P, Cook EH et al. Combining information from multiple sources in the diagnosis of autism spectrum disorders. J Am Acad Child Adolesc Psychiatry 2006; 45: 1094–1103. | Article | PubMed | ISI |
  34. Sparrow SS, Balla DA, Cicchetti D. Vineland Adaptive Behavior Scales, Interview Edition. American Guidance Service: Circle Pines, MN, 1984.
  35. Wechsler D, Psychological Corporation. WISC-II. [kit] Wechsler Intelligence Scale for Children, 3rd edn. Psychological Corporation Harcourt Brace Jovanovich: San Antonio, TX, 1991.
  36. Kaufman AS, Kaufman NL. Kaufman Assessment Battery for Children (K-ABC). American Guidance Services: Minneapolis, 1983.
  37. Bayley N. Bayley Scales of Infant Development, 2nd edn. Harcourt Brace & Company: San Antonio, 1993.
  38. Stutsman R. Manual for the Merrill–Palmer Scale of Mental Tests. Western Psychological Services: Los Angeles, CA, 1948.
  39. Mullen EM. Mullen Scales of Early Learning. Western Psychological Services: Los Angeles, CA, 1997.
  40. Cattell P. The Measurement of Intelligence of Infants and Young Children—Revised. Psychological Corporation: New York, 1960.
  41. Roid GH, Miller LJ, Lieter RG. Leiter International Performance Scale—Revised: Examiners Manual. Stoelting: Wood Dale, IL (620 Wheat Ln., Wood Dale 60191), 1997 xiii, 1 v. (various pagings).
  42. Thorndike RL, Hagen EP, Sattler JM. The Stanford–Binet Intelligence Scale, Fourth Edition: Guide for Administering and Scoring. Riverside Pub. Co: Chicago, IL. (8420 Bryn Mawr Avenue, Chicago 60631), 1986 192pp.
  43. Sham PC, Curtis D. An extended transmission/disequilibrium test (TDT) for multi-allele marker loci. Ann Hum Genet 1995; 59(Part 3): 323–336. | Article | PubMed | ISI | ChemPort |
  44. Dudbridge F. Pedigree disequilibrium tests for multilocus haplotypes. Genet Epidemiol 2003; 25: 115–121. | Article | PubMed | ISI |
  45. Crawford DC, Nickerson DA. Definition and clinical importance of haplotypes. Annu Rev Med 2005; 56: 303–320. | Article | PubMed | ISI | ChemPort |
  46. Bartz AJ, Hollander E. The neuroscience of affiliation: forging links between basic and clinical research on neuropeptides and social behavior. Horm Behav 2006; 50: 518–528. | Article | PubMed | ISI | ChemPort |
  47. Heinrichs M, Baumgartner T, Kirschbaum C, Ehlert U. Social support and oxytocin interact to suppress cortisol and subjective responses to psychosocial stress. Biol Psychiatry 2003; 54: 1389–1398. | Article | PubMed | ISI | ChemPort |
  48. Mizumoto Y, Kimura T, Ivell R. A genomic element within the third intron of the human oxytocin receptor gene may be involved in transcriptional suppression. Mol Cell Endocrinol 1997; 135: 129–138. | Article | PubMed | ISI | ChemPort |
  49. de Wied D, Diamant M, Fodor M. Central nervous system effects of the neurohypophyseal hormones and related peptides. Front Neuroendocrinol 1993; 14: 251–302. | Article | PubMed | ChemPort |
  50. Carter CS, Altemus M. Integrative functions of lactational hormones in social behavior and stress management. Ann NY Acad Sci 1997; 807: 164–174. | Article | PubMed | ChemPort |
  51. Baron-Cohen S, Wheelwright S, Hill J, Raste Y, Plumb I. The 'Reading the Mind in the Eyes' Test revised version: a study with normal adults, and adults with Asperger syndrome or high-functioning autism. J Child Psychol Psychiatry 2001; 42: 241–251. | Article | PubMed | ISI | ChemPort |
  52. Hammock EA, Young LJ. Oxytocin, vasopressin and pair bonding: implications for autism. Philos Trans R Soc Lond B Biol Sci 2006; 361: 2187–2198. | Article | PubMed | ChemPort |
  53. Carter CS. Sex differences in oxytocin and vasopressin: implications for autism spectrum disorders? Behav Brain Res 2007; 176: 170–186. | Article | PubMed | ISI | ChemPort |
  54. Young LJ, Pitkow LJ, Ferguson JN. Neuropeptides and social behavior: animal models relevant to autism. Mol Psychiatry 2002; 7(Suppl 2): S38–S39. | Article | PubMed | ISI |
  55. Born J, Lange T, Kern W, McGregor GP, Bickel U, Fehm HL. Sniffing neuropeptides: a transnasal approach to the human brain. Nat Neurosci 2002; 5: 514–516. | Article | PubMed | ISI | ChemPort |


This research was partially supported by the Israel Science Foundation founded by the Israel Academy of Sciences and Humanities (NY and RPE).