You are on page 1of 6

Research

JAMA Pediatrics | Original Investigation

Association of Sex With Recurrence


of Autism Spectrum Disorder Among Siblings
Nathan Palmer, PhD; Andrew Beam, PhD; Denis Agniel, PhD; Alal Eran, PhD; Arjun Manrai, PhD;
Claire Spettell, PhD; Gregory Steinberg, MD; Kenneth Mandl, MD; Kathe Fox, PhD;
Stanley F. Nelson, MD; Isaac Kohane, MD, PhD

IMPORTANCE Autism spectrum disorder (ASD) is known to be more prevalent among males
than females in the general population. Although overall risk of recurrence of ASD among
siblings has been estimated to be between 6.1% and 24.7%, information on sex-specific
recurrence patterns is lacking.

OBJECTIVE To estimate high-confidence sex-specific recurrence rates of ASD among siblings.

DESIGN, SETTING, AND PARTICIPANTS This observational study used an administrative database
to measure the incidence of ASD among children in 1 583 271 families (37 507 with at least 1
diagnosis of ASD) enrolled in commercial health care insurance plans at a large US managed
health care company from January 1, 2008, through February 29, 2016. Families in the study Author Affiliations: Department of
had 2 children who were observed for at least 12 months between 4 and 18 years of age. Biomedical Informatics, Harvard
Medical School, Boston,
Massachusetts (Palmer, Beam,
MAIN OUTCOMES AND MEASURES The primary measure of ASD recurrence was defined as the Agniel, Eran, Manrai, Kohane); Aetna
diagnosis of ASD in a younger sibling of an older sibling with an ASD diagnosis. Inc, Hartford, Connecticut (Spettell,
Fox); Independent Coauthor,
Dingmans Ferry, Pennsylvania
RESULTS Among the 3 166 542 children (1 547 266 females and 1 619 174 males; mean [SD]
(Steinberg); Department of
age, 11.2 [4.7] years) in the study, the prevalence of ASD was 1.96% (95% CI, 1.94%-1.98%) Pediatrics, Harvard Medical School,
among males and 0.50% (95% CI, 0.49%-0.51%) among females. When a male was Boston, Massachusetts (Mandl);
associated with risk in the family, ASD was diagnosed in 4.2% (95% CI, 3.8%-4.7%) of female Computational Health Informatics
Program, Boston Children’s Hospital,
siblings and 12.9% (95% CI, 12.2%-13.6%) of male siblings. When a female was associated Boston, Massachusetts (Mandl);
with risk in the family, ASD was diagnosed in 7.6% (95% CI, 6.5%-8.9%) of female siblings Department of Human Genetics,
and 16.7% (95% CI, 15.2%-18.4%) of male siblings. David Geffen School of Medicine,
UCLA (University of California,
Los Angeles) (Nelson); Department of
CONCLUSIONS AND RELEVANCE These findings are in agreement with the higher rates of ASD Psychiatry, David Geffen School of
observed among males than among females in the general population. Our study provides Medicine, UCLA (University of
more specific guidance for the screening and counseling of families and may help inform California, Los Angeles) (Nelson).
future investigations into the environmental and genetic factors that confer risk of ASD. Corresponding Author: Isaac
Kohane, MD, PhD, Department of
Biomedical Informatics, Harvard
JAMA Pediatr. 2017;171(11):1107-1112. doi:10.1001/jamapediatrics.2017.2832 Medical School, 10 Shattuck St,
Published online September 25, 2017. Boston, MA 02115
(isaac_kohane@harvard.edu).

A
utism spectrum disorder (ASD) comprises a group of ASD cannot yet be explained using genetic approaches, sev-
common neurodevelopmental disabilities of mostly un- eral groups have sought to associate environmental exposures
known etiologic factors. Recent studies have sug- with ASD.11-13 However, the risk conferred by genomic varia-
gested that the prevalence of ASD in the general US population tion in combination with environmental exposures remains
may be greater than 1%,1 with approximately 4 times more males unknown. Understanding the rate at which ASD diagnoses are
affected than females. Autism spectrum disorder has a strong shared among siblings in a family will aid in the interpretation
genetic component, with heritability estimates ranging from of the genetic and environmental factors that increase the fa-
50% to 90%.2-4 Recent whole-genome and exome studies have milial risk for ASD, leading to improved screening and counsel-
implicated an increasing number of genes and pathways in ASD, ing at the point of care.
highlighting its extreme locus heterogeneity.5-9 At present, clini- The sibling recurrence rate, defined as the probability of
cal genetics evaluation is recommended after an ASD diagno- a child having ASD with 1 or more siblings with ASD, has been
sis and is estimated to result in an identified etiologic factor in estimated to be 6.1% to 18.7%.14,15 However, sex-specific re-
30% to 40% of patients.10 Because the cause of most cases of currence rates in the general US population remain largely

jamapediatrics.com (Reprinted) JAMA Pediatrics November 2017 Volume 171, Number 11 1107

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018


Research Original Investigation Association of Sex With Recurrence of Autism Spectrum Disorder Among Siblings

unknown; these data are particularly important for a sexu-


ally dimorphic disorder such as ASD, for which the strongest Key Points
risk factor is being male.16,17 Recent studies have examined
Question What are the sex-specific recurrence rates of autism
sex-specific ASD recurrence in siblings in small, targeted spectrum disorder among siblings?
cohorts14,18-21 or birth cohorts,3 but the sex-specific recur-
Findings In this population analysis of 1 583 271 families with 2
rence among siblings in the general US population remains to
children, a significantly increased risk of recurrence of autism
be quantified. Thus, we performed a large, retrospective popu-
spectrum disorder was found among males than among females.
lation analysis of the rate at which sibling pairs are com-
monly or singly diagnosed with ASD. The data for this analy- Meaning An older female sibling diagnosed with autism spectrum
disorder is associated with greater risk of recurrence in the
sis were derived from diagnostic billing codes at a commercial
younger sibling compared with an older diagnosed male sibling,
managed health care company in the United States, yielding
and male siblings are more likely to experience recurrence than
the largest such study to date, to our knowledge. female siblings regardless of the sex of the diagnosed sibling.

determined to not be significantly associated with ASD diag-


Methods nosis in the younger sibling (see Results) and was omitted from
Our study population was drawn from a deidentified admin- the final model. Recurrence probabilities and their respective
istrative database at Aetna, a commercial managed health care 95% CIs were calculated from the logistic regression. P values
company. At the time of analysis, the database contained describing the significance of the difference between pairs of
International Classification of Diseases, Ninth Revision (ICD-9) estimated recurrence rates and their associated SEs were cal-
diagnostic billing codes for more than 63 million individuals culated using a z-test on the log odds-scale values derived from
in the United States during a 98-month time frame from Janu- the regression model. P < .05 (2-tailed) was considered sig-
ary 1, 2008, through February 29, 2016. No race, ethnicity, or nificant. All calculations were performed using a combina-
socioeconomic data were present in the database. The tion of custom written queries in Microsoft SQL Server 2014
Harvard Medical School Institutional Review Board waived (Microsoft Corp) and R statistical software.22
the requirement for approval, as it deemed this analysis of the
database to not be human subjects research.
Using subscriber-to-member relationship information con-
tained in the enrollment data, we identified all families with
Results
exactly 2 children between the ages of 4 and 18 years and with Overview
at least 12 months of medical coverage. The children in these In a general population cohort of 3 166 542 children from
families constituted our study population. A small fraction of 1 583 271 sibling pairs, the overall prevalence of ASD was 1.25%
the designated siblings in these families may have been ge- (95% CI, 1.23%-1.26%); the prevalence was 1.96% (95% CI,
netically unrelated or half-siblings. Because there is no way 1.94%-1.98%) among males and 0.50% (95% CI, 0.49%-
to identify and exclude these siblings from our analysis, we ac- 0.51%) among females. The male to female ratio of the study
knowledge that this caveat may lead to underestimation of the population was 1.05 and of individuals with ASD was 4.1, con-
rates of recurrence. However, this limitation should be com- sistent with previous findings.17 Figure 1 summarizes sex-
parable for males and females. specific ASD diagnoses in the 4 types of sibling pairs studied.
In total, 3 166 542 children were evaluable from 1 583 271 Figure 2 shows the distribution of ASD diagnoses by age
families. Of these, 39 460 children from 37 507 families had re- in the study cohort. The timing of diagnosis peaked at 5 to 6
ceived a diagnosis of ASD (ICD-9 code in the 299 group), indi- years of age and decreased thereafter, consistent with previ-
cating an overall population incidence of 1.25% (95% CI, 1.23%- ous reports.1,23 Each child may have received multiple diag-
1.26%). The male to female ratio of the overall study population noses during the observation period, and diagnoses may have
was 1.05, and the male to female ratio of those with ASD was occurred before or after the required 12-month observation pe-
4.10. riod between 4 and 18 years of age. Because the rate of diag-
To examine the sex-specific association of an older sib- nosis varied with age, we examined the effect of including ad-
ling’s ASD diagnosis with diagnosis in the younger sibling (ie, justments for age when measuring the association of a sibling’s
recurrence), we performed a multiple logistic regression on the ASD diagnosis. Age was determined to not be a significant
21 074 two-child families in the study population whose old- covariate (odds ratio, 1.01; 95% CI, 1.00-1.02; P = .13) in the
est child had received a diagnosis of ASD. The younger sib- logistic regression. Inclusion of age in the model yielded no
ling’s diagnosis status was modeled as a function of the child’s statistically significant difference in the recurrence probabili-
mean age during the observation period and the interaction ties for any of the sibling pairs when computed at 4 years of
between the birth-ordered sexes of the sibling pair (male and age vs 18 years of age (the useful range of the model). Thus,
male, male and female, female and male, and female and fe- age was omitted from the regression analysis.
male). Because the database contains only birth year and not Figure 3 shows the distribution of study participants (in-
full birth dates, we could not with certainty identify twins. cluding older and younger siblings) for number of months of
Thus, any sibling pairs with the same birth year were ex- observation. The spikes at 12-month intervals can be attrib-
cluded from this portion of the analysis. The child’s age was uted to the data being derived from an administrative data-

1108 JAMA Pediatrics November 2017 Volume 171, Number 11 (Reprinted) jamapediatrics.com

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018


Association of Sex With Recurrence of Autism Spectrum Disorder Among Siblings Original Investigation Research

Figure 1. Sex-Specific Patterns of Autism Spectrum Disorder (ASD) Diagnoses Among Sibling Pairs

A Male and male siblings

395 220 Pairs

8447 Older siblings 386 773 Older siblings


with ASD without ASD
diagnosis diagnosis

1092 Younger siblings 7355 Younger siblings 5940 Younger siblings 380 833 Younger siblings
with ASD without ASD with ASD without ASD
diagnosis diagnosis diagnosis diagnosis

B Female and male siblings

400 249 Pairs

2017 Older siblings 398 232 Older siblings


with ASD without ASD
diagnosis diagnosis

339 Younger siblings 1678 Younger siblings 7183 Younger siblings 391 049 Younger siblings
with ASD without ASD with ASD without ASD
diagnosis diagnosis diagnosis diagnosis

C Female and female siblings

359 679 Pairs

1976 Older siblings 357 703 Older siblings


with ASD without ASD
diagnosis diagnosis

150 Younger siblings 1826 Younger siblings 1607 Younger siblings 356 096 Younger siblings
with ASD without ASD with ASD without ASD
diagnosis diagnosis diagnosis diagnosis

D Male and female siblings

405 389 Pairs

8442 Older siblings 396 947 Older siblings


with ASD without ASD
diagnosis diagnosis

356 Younger siblings 8086 Younger siblings 1490 Younger siblings 395 457 Younger siblings The numbers at each vertex indicate
with ASD without ASD with ASD without ASD the total number of families observed
diagnosis diagnosis diagnosis diagnosis
to have a corresponding pattern of
diagnoses and births.

base at a US managed health care company, and commercial for the servicing health care provider. Among the remaining
health plans typically renew or change enrollment on an an- 85% of claims, the top 10 most frequently recorded special-
nual basis. The spike at 98 months represents the siblings who ties, accounting for greater than 90% of the claims with an
were enrolled for the entire 98-month observation period re- ASD diagnosis, were mental health professional, physical
corded in the data set. The quartiles of the distribution were rehabilitation professional, family practice, mental health
24, 44, and 72 months; 75% of the children had at least 24 provider, pediatrics, acute short-term hospital, dentistry,
months of observation, 50% had at least 44 months of obser- children’s hospital, neurology, and rehabilitation facility.
vation, and 20% had at least 72 months of observation. The remaining 57 recorded specialties each accounted
Approximately 15% of the claims that met the coding for 1% or less of all ASD diagnosis claims for the study
criteria for the study population did not record a specialty population.

jamapediatrics.com (Reprinted) JAMA Pediatrics November 2017 Volume 171, Number 11 1109

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018


Research Original Investigation Association of Sex With Recurrence of Autism Spectrum Disorder Among Siblings

Figure 2. Distribution of Autism Spectrum Disorder (ASD) Diagnoses Figure 4. Sibling Rates of Recurrence of Autism Spectrum Disorder (ASD)

300 000 20
Diagnoses Across Population, No.

250 000

Probability of Recurrence
15
200 000

150 000 10

100 000

5
50 000

0
0 2 4 6 8 10 12 14 16 18 20 22 24 0
Younger Male Younger Male Younger Female Younger Female
Age at Diagnosis, y With Older With Older With Older With Older
Female With Male With Female With Male With
ASD Diagnosis ASD Diagnosis ASD Diagnosis ASD Diagnosis
Individuals may have had multiple ASD diagnoses.
Sex

Probability of a diagnosis of ASD in a younger sibling with an older sibling


Figure 3. Distribution of Number of Study Participants During Observation diagnosed with ASD, stratified by sex and birth order. Recurrence was more
likely in a male sibling than in a female sibling regardless of whether the older
4500 sibling with the associated risk was male or female. Older female siblings with
4000 an ASD diagnosis were associated with higher rates of recurrence (in male or
female siblings) than were older male siblings. Whiskers indicate 95% CIs.
3500
Study Participants, No.

3000
ling with an ASD diagnosis appeared to be a stronger indica-
2500
tor of risk than a similarly diagnosed male sibling.
2000

1500

1000
Discussion
500

0 Because these data were derived from insurance company–


12 24 36 48 60 72 84 96 based ICD-9 codes for a large sample population, we were able
Observation, mo
to amass a larger data set than in most studies in ASD in which
The spike at 12-month intervals reflects the period of annual commercial diagnosis was determined in a clinical research setting. Thus,
insurance plan renewals or enrollment changes. this study of a population of 37 507 families with ASD is larger
than than those in several studies published in the past de-
cade and therefore enables a broad-based estimate of recur-
ASD Recurrence Rates rence in the context of medical care. Despite the obvious meth-
We estimated sex-specific recurrence rates of ASD by fitting a odological differences in ascertainment of ASD cases, the risk
logistic regression on the 21 074 families with 2 children in the of recurrence seen in these data is generally similar to esti-
study population whose oldest child had received a diagnosis mates from recently reported retrospective studies3,24 but
of ASD. Figure 4 shows the estimated recurrence probabilities lower than the highest estimates found in a smaller prospec-
with 95% CIs for each of the 4 possible ordered sibling pairings. tive infant sibling cohort.14
The probability of ASD recurrence in a younger male sib- Consistent with some prior studies, we found a differ-
ling with an older diagnosed female sibling was the highest of ence in the rate of recurrence in individuals of the same sex
all 4 ordered sibling pairs (16.7%; 95% CI, 15.2%-18.4%). The depending on the sex of their sibling. For families with 2 chil-
next highest probability of recurrence was in a younger male dren, the rate of recurrence of ASD in a male child was ap-
with an older diagnosed male sibling (12.9%; 95% CI, 12.2%- proximately 1 in 7.7 if the other sibling with ASD was male. If
13.6%). The probability of recurrence in a younger female sib- the other sibling with ASD was female, the rate of recurrence
ling with an older diagnosed female sibling was 7.6% (95% CI, was 1 in 6, an approximately 1.3-fold higher risk of recur-
6.5%-8.9%). The lowest probability of recurrence was ob- rence. These rates are consistent with a model in which fe-
served for a younger female sibling with an older diagnosed males require a higher number of genetic or environmental risk
male sibling (4.2%; 95% CI, 3.8%-4.7%). All pairwise compari- factors to manifest ASD than do males.
sons of these sex-specific recurrence rates indicated a statis- This finding does not fit into a model of X-linked muta-
tically significant difference (P < .001). tions or germline mosaicism but may be consistent with mul-
When a female or male older sibling had a diagnosis of ASD, tiple alleles or mutations of modest effect and a protective en-
a younger male sibling had a higher probability of an ASD di- vironment for unexplained reasons in the developing female
agnosis than a female participant. Furthermore, a female sib- brain. The rate of recurrence of ASD among female children is

1110 JAMA Pediatrics November 2017 Volume 171, Number 11 (Reprinted) jamapediatrics.com

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018


Association of Sex With Recurrence of Autism Spectrum Disorder Among Siblings Original Investigation Research

7.6% if the affected sibling is female and 4.2% if male. Thus, ent. The sample size of our study enabled a high-confidence
the risk associated with a female sibling diagnosed with ASD analysis of sex-specific ASD rates of recurrence in the general
is higher than that associated with a male sibling, a finding that population, which prior studies were underpowered to detect.3
is in line with several genetic models. In the multivariate model These findings could be incorporated into screening and coun-
of Risch et al,25 a female sibling with a previous diagnosis of seling for families with no known ASD etiologic factors.
ASD was found to be associated with an increased rate of re-
currence in a male sibling, consistent with our observations. Limitations
This finding is also consistent with those of Robinson et al26 This study relied on data from an insurance claims database. Al-
of British and Swedish twin cohorts in which siblings of fe- though all data were based on patients’ medical information, the
male probands with ASD had significantly more autistic traits precise basis or severity of the ASD could not be determined from
than did siblings of male probands with ASD. Whether the fe- the provided data. Additional, precise genetic relationships could
male protective effect is inherent in differences in structure not be determined; some study members listed as siblings may
of the brain and hormonal milieu or in unidentified specific have been half-siblings or genetically unrelated.
genetic variants remains a matter of continued investigation.
Regardless of the mechanism, recent studies support the hy-
pothesis that the female protective effect results in female pro-
bands having a higher mutational burden, as determined by
Conclusions
exome sequencing in neurodevelopmental disorders.27 There- Autism spectrum disorder is a neurodevelopmental disorder
fore, males with a similar or lesser genetic burden acquired that is known to affect males at a higher rate than females in
through the family shared with the female proband may mani- the general population and to cluster in families. This study
fest the disorder but the converse may not be true. estimated the frequency with which the disease recurred in
Although some prior studies have not described a sex as- female and male sibling pairs and the rate of recurrence in sex-
sociation in ASD, our data set is the largest to date to report heterogeneous families. Our findings support the hypothesis
sex-specific recurrence rates among siblings and should be fac- that the disease clusters in families owing to shared genetic
tored into genetic counseling for families in the absence of a or environmental factors. Our findings also support previous
known explanatory cause of ASD in the family. Sandin et al3 observations that ASD affects males more than females even
identified a pattern of sex-specific adjusted relative risk of re- within families. For families that have a child who has been
currence for ASD that is consistent with our analysis. Al- diagnosed with ASD, the findings of this research provide bet-
though that study found the difference in relative risk of re- ter likelihood estimates of disease recurrence among subse-
currence to not be statistically significant, the sex-specific rates quent offspring and may help guide counseling and support
of recurrence reported in our study are significantly differ- services.

ARTICLE INFORMATION years—Autism and Developmental Disabilities 9. De Rubeis S, He X, Goldberg AP, et al; DDD
Accepted for Publication: June 27, 2017. Monitoring Network, 11 Sites, United States, 2010. Study; Homozygosity Mapping Collaborative for
https://www.cdc.gov/mmwr/preview/mmwrhtml Autism; UK10K Consortium. Synaptic,
Published Online: September 25, 2017. /ss6302a1.htm?s_cid=ss6302a1_w. Published transcriptional and chromatin genes disrupted in
doi:10.1001/jamapediatrics.2017.2832 March 28, 2014. Accessed May 7, 2017. autism. Nature. 2014;515(7526):209-215.
Author Contributions: Dr Palmer had full access to 2. Ronald A, Happé F, Bolton P, et al. Genetic 10. Schaefer GB, Mendelsohn NJ; Professional
all the data in the study and takes responsibility for heterogeneity between the three components of Practice and Guidelines Committee. Clinical
the integrity of the data and the accuracy of the the autism spectrum: a twin study. J Am Acad Child genetics evaluation in identifying the etiology of
data analysis. Adolesc Psychiatry. 2006;45(6):691-699. autism spectrum disorders: 2013 guideline
Study concept and design: Palmer, Beam, Spettell, revisions. Genet Med. 2013;15(5):399-407.
Mandl, Fox. 3. Sandin S, Lichtenstein P, Kuja-Halkola R, Larsson
Acquisition, analysis, or interpretation of data: H, Hultman CM, Reichenberg A. The familial risk of 11. Atladóttir HO, Henriksen TB, Schendel DE,
Palmer, Beam, Agniel, Eran, Manrai, Spettell, autism. JAMA. 2014;311(17):1770-1777. Parner ET. Autism after infection, febrile episodes,
Steinberg, Nelson, Kohane. 4. Hallmayer J, Cleveland S, Torres A, et al. Genetic and antibiotic use during pregnancy: an exploratory
Drafting of the manuscript: Palmer, Eran, Fox. heritability and shared environmental factors study. Pediatrics. 2012;130(6):e1447-e1454.
Critical revision of the manuscript for important among twin pairs with autism. Arch Gen Psychiatry. 12. Harrington RA, Lee LC, Crum RM, Zimmerman
intellectual content: Palmer, Beam, Agniel, Manrai, 2011;68(11):1095-1102. AW, Hertz-Picciotto I. Prenatal SSRI use and offspring
Spettell, Steinberg, Mandl, Nelson, Kohane. 5. C Yuen RK, Merico D, Bookman M, et al. Whole with autism spectrum disorder or developmental
Statistical analysis: Palmer, Beam, Agniel, Eran, genome sequencing resource identifies 18 new delay. Pediatrics. 2014;133(5):e1241-e1248.
Manrai, Fox. candidate genes for autism spectrum disorder. Nat 13. Zerbo O, Yoshida C, Gunderson EP, Dorward K,
Obtained funding: Kohane. Neurosci. 2017;20(4):602-611. Croen LA. Interpregnancy interval and risk of autism
Administrative, technical, or material support: spectrum disorders. Pediatrics. 2015;136(4):651-657.
Spettell, Kohane. 6. Yuen RK, Thiruvahindrapuram B, Merico D, et al.
Study supervision: Mandl, Kohane. Whole-genome sequencing of quartet families with 14. Ozonoff S, Young GS, Carter A, et al. Recurrence
autism spectrum disorder. Nat Med. 2015;21(2): risk for autism spectrum disorders: a Baby Siblings
Conflict of Interest Disclosures: None reported. 185-191. Research Consortium study. Pediatrics. 2011;128(3):
7. Krumm N, Turner TN, Baker C, et al. Excess of e488-e495.
REFERENCES
rare, inherited truncating mutations in autism. Nat 15. Grønborg TK, Schendel DE, Parner ET.
1. Autism and Developmental Disabilities Genet. 2015;47(6):582-588. Recurrence of autism spectrum disorders in full-
Monitoring Network Surveillance Year 2010 and half-siblings and trends over time:
Principal Investigators. Prevalence of autism 8. Iossifov I, O’Roak BJ, Sanders SJ, et al. The
contribution of de novo coding mutations to autism a population-based cohort study. JAMA Pediatr.
spectrum disorder among children aged 8 2013;167(10):947-953.
spectrum disorder. Nature. 2014;515(7526):216-221.

jamapediatrics.com (Reprinted) JAMA Pediatrics November 2017 Volume 171, Number 11 1111

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018


Research Original Investigation Association of Sex With Recurrence of Autism Spectrum Disorder Among Siblings

16. Ruzich E, Allison C, Chakrabarti B, et al. Sex and genetic liability for autism spectrum disorders in 25. Risch N, Hoffmann TJ, Anderson M, Croen LA,
STEM occupation predict autism-spectrum multiplex families and twins. Mol Autism. 2015;6:27. Grether JK, Windham GC. Familial recurrence of
quotient (AQ) scores in half a million people. PLoS 21. Wood CL, Warnell F, Johnson M, et al. Evidence autism spectrum disorder: evaluating genetic and
One. 2015;10(10):e0141229. for ASD recurrence rates and reproductive environmental contributions. Am J Psychiatry.
17. Lai MC, Lombardo MV, Auyeung B, Chakrabarti stoppage from large UK ASD research family 2014;171(11):1206-1213.
B, Baron-Cohen S. Sex/gender differences and databases. Autism Res. 2015;8(1):73-81. 26. Robinson EB, Koenen KC, McCormick MC, et al.
autism: setting the scene for future research. J Am 22. R: a language and environment for statistical A multivariate twin study of autistic traits in
Acad Child Adolesc Psychiatry. 2015;54(1):11-24. computing [computer program]. Vienna, Austria: 12-year-olds: testing the fractionable autism triad
18. Xie F, Peltier M, Getahun D. Is the risk of autism R Foundation for Statistical Computing; 2015. hypothesis. Behav Genet. 2012;42(2):
in younger siblings of affected children moderated 245-255.
23. Shattuck PT, Durkin M, Maenner M, et al.
by sex, race/ethnicity, or gestational age? J Dev Timing of identification among children with an 27. Jacquemont S, Coe BP, Hersch M, et al. A higher
Behav Pediatr. 2016;37(8):603-609. autism spectrum disorder: findings from a mutational burden in females supports a “female
19. Messinger DS, Young GS, Webb SJ, et al. Early population-based surveillance study. J Am Acad protective model” in neurodevelopmental
sex differences are not autism-specific: a Baby Child Adolesc Psychiatry. 2009;48(5):474-483. disorders. Am J Hum Genet. 2014;94(3):
Siblings Research Consortium (BSRC) study. Mol 415-425.
24. Constantino JN, Zhang Y, Frazier T, Abbacchi
Autism. 2015;6:32. AM, Law P. Sibling recurrence and the genetic
20. Werling DM, Geschwind DH. Recurrence rates epidemiology of autism. Am J Psychiatry. 2010;167
provide evidence for sex-differential, familial (11):1349-1356.

1112 JAMA Pediatrics November 2017 Volume 171, Number 11 (Reprinted) jamapediatrics.com

© 2017 American Medical Association. All rights reserved.

Downloaded From: on 11/19/2018

You might also like