Left Brain Right Brain

A few months back an article appeared in the journal Pediatrics: Neonatal Jaundice, Autism, and Other Disorders of Psychological Development. It’s one of those articles I meant to blog and just never got around to.

Here’s the abstract:

Objectives The goals were to study the association between neonatal jaundice and disorders of psychological development in a national, population-based cohort and to study whether gestational age, parity, and season of birth influenced that association.

Methods A population-based, follow-up study of all children born alive in Denmark between 1994 and 2004 (N = 733 826) was performed, with data collected from 4 national registers. Survival analysis was used to calculate hazard ratios (HRs).

Results Exposure to jaundice in neonates was associated with increased risk of disorders of psychological development for children born at term. The excess risk of developing a disorder in the spectrum of psychological development disorders after exposure to jaundice as a neonate was between 56% (HR: 1.56 [95% confidence interval [CI]: 1.05–2.30]) and 88% (HR: 1.88 [95% CI: 1.17–3.02]). The excess risk of infantile autism was 67% (HR: 1.67 [95% CI: 1.03–2.71]). This risk for infantile autism was higher if the child was conceived by a parous woman (HR: 2.71 [95% CI: 1.57–4.66]) or was born between October and March (HR: 2.21 [95% CI: 1.24–3.94]). The risk for infantile autism disappeared if the child was conceived by a primiparous woman (HR: 0.58 [95% CI: 0.18–1.83]) or was born between April and September (HR: 1.02 [95% CI: 0.41–2.50]). Similar risk patterns were found for the whole spectrum of autistic disorders.

Conclusions Neonatal jaundice in children born at term is associated with disorders of psychological development. Parity and season of birth seem to play important roles.

This week, three response letters were published in Pediatrics. These originally appeared as online letters:

Effects of perinatal asphyxia plus bilirubin?

Jaundice-Autism Link Unconvincing

and

neonatal jaundice: should we go crazy?

The first article (Effects of perinatal asphyxia plus bilirubin) was supportive, ending with:

Thank you for publishing the report by Maimberg et al., which will hopefully help direct research on perinatal brain vulnerability as an important area of investigation

The second article, as you might guess, was not. “Jaundice-Autism Link Unconvincing”. This was written by Thomas B Newman and Lisa Croen. Lisa Croen has previously published on biliruben (the substance which causes jaundice). They present a number of confounders which should be considered. They also note the potential problems with an incomplete study:

Raising concerns about the danger of jaundice could lead parents or clinicians to treat jaundice more aggressively. However, the Danish study also lacks any information on phototherapy. Thus, if the association is real, it could be due to phototherapy, rather than to the jaundice itself, in which case frightening families about jaundice could lead to an increase in treatment, which would be particularly counter-productive.

The third response notes “In this regard, it should be quoted the paper by Croen at al, who studied neonatal bilirubin levels (although the presence of haemolytic diseases was unspecified) in relation with autism spectrum disorders and did not find any association”. Yes, that’s the same Croen as in the second response. There was no association in those data and that previous study.

Another online letter is even more direct: Neonatal Jaundice Does Not Cause Autism.

In the Response from authors, we see that the authors have performed a reanalysis:

Sir, it has come to our attention that table 3 in our paper (1) only includes patients treated in somatic Danish hospitals identified in the Danish National Hospital Register. If we also add the patients diagnosed with autism spectrum disorders (ASD) at the mental hospitals only (n=6171), the revised figures will look like this:

Association between neonatal jaundice and autism spectrum disorders (F84.0-F84.9):

All;

Crude HR=1.29, adjusted HR=1.07 (0.94-1.21)

Term=>37 weeks;

Crude HR=1.23, adjusted HR=1.06 (0.90-1.25)

Preterm< 37 weeks;

Crude HR=1.17,adjusted HR=1.05 (0.83-1.32)
As seen in the new analysis (including children from the mental hospital only)the association between jaundice and autism is substantially attenuated after adjustments which make us doubt the causal nature of our observation. If the association is causal it apparently only involves the more severe cases reaching the somatic hospital. Our findings are in line with what has been reported from Sweden (2) using a similar case mix. It is likely that our study and the Swedish study are confounded by co- morbidity.

“which make us doubt the causal nature of our observation.”.

I’m rather glad I didn’t blog this at the time. This looks like the science community working to self-correct itself.

Real data on autism prevalence around the world is scant. Many countries do not track accurately (including the US) and few countries have reported on autism prevalence.

So, while this is only a “pilot study”, I was intrigued to read that someone was reporting on autism prevalence in Brazil:

Brief Report: Prevalence of Pervasive Developmental Disorder in Brazil: A Pilot Study.

Paula CS, Ribeiro SH, Fombonne E, Mercadante MT.

Developmental Disorder Post Graduation Program, Mackenzie Presbyterian University, Rua da Consolação, 930. Edifício 38 Secretaria de Pós-Graduação em Distúrbios do Desenvolvimento, São Paulo, SP, 01302-907, Brazil, csilvestrep09@gmail.com.
Abstract

This pilot study presents preliminary results concerning the prevalence of Pervasive Developmental Disorder (PDD) in South America. It was a three-phase study conducted in a typical town in Southeast Brazil. Case definition was based in a combination of standardized instruments and clinical evaluations by experts. The prevalence of PDD was 27.2/10,000 (95% CI: 17.6-36.8) and some hypotheses were raised to explain this low frequency. Clinical findings of PDD cases were consistent with previous data, such as, male preponderance, more children diagnosed with PDD-NOS than with autistic disorder, and half of them born from older mothers. Moreover, the study raised concerns about treatment of cases, because identification of PDD had been late and access to services has been very limited.

The prevalence is low: about 0.27% for all PDD’s. About 1/4 that of the US or the UK.

Just in this year (2011), 4 other papers have come out on autism prevalence:

Denmark and Australia: 0.685% and 0.51%, respectively.

Autism prevalence varies by race, even within a single US municipality (Atlanta). In the San Francisco Bay area, autism prevalence is about 0.47%, and again varies with race/ethnicity. Yes, the San Francisco bay area has a prevalence about 1/2 the national average.

While more of a review, this paper discusses how autism prevalence rates vary by country

Autism prevalence estimates vary. They vary with time, they vary with place, they vary with ethnicity, they vary with socio-economic status. I don’t think it is hopeless to try to pull causation inferences from prevalence data, even administrative prevalence data like education data or the CDDS data. It is hard though. There are obviously a number of factors at play. Many social factors included. And that makes it really hard to interpret what all these data mean.

As a final aside, for those trying to work these latest data from Brazil into the “vaccine hypothesis”, the vaccine schedule for Brazil is here. They have a lot of vaccines, including the birth dose of HepB and thimerosal containing vaccines. The low prevalence in this pilot study do not support the hypothesis.

Most of what we hear (especially on this blog and other autism focused sources) about the Autism and Developmental Disabilities Monitoring Network (ADDM) is, well, autism related. Mostly we get the prevalence estimates for the CDC from this source. Sometimes I admit, I forget that “and Developmental Disabilities” is in there.

A recent paper shows that the ADDM is more broad than just autism. And, at the same time, gives some interesting autism information. The paper is: Prevalence and functioning of children with cerebral palsy in four areas of the United States in 2006: A report from the Autism and Developmental Disabilities Monitoring Network

Here is the abstract:

Prevalence and functioning of children with cerebral palsy in four areas of the United States in 2006: A report from the Autism and Developmental Disabilities Monitoring Network.

Kirby RS, Wingate MS, Van Naarden Braun K, Doernberg NS, Arneson CL, Benedict RE, Mulvihill B, Durkin MS, Fitzgerald RT, Maenner MJ, Patz JA, Yeargin-Allsopp M.

Department of Community and Family Health, College of Public Health, University of South Florida, 13201 Bruce B. Downs Blvd, MDC56, Tampa, FL 33612, United States.
Abstract

AIM: To estimate the prevalence of cerebral palsy (CP) and the frequency of co-occurring developmental disabilities (DDs), gross motor function (GMF), and walking ability using the largest surveillance DD database in the US.

METHODS: We conducted population-based surveillance of 8-year-old children in 2006 (N=142,338), in areas of Alabama, Georgia, Wisconsin, and Missouri. This multi-site collaboration involved retrospective record review at multiple sources. We reported CP subtype, co-occurring DDs, Gross Motor Function Classification System (GMFCS) level, and walking ability as well as CP period prevalence by race/ethnicity and sex.

RESULTS: CP prevalence was 3.3 (95% confidence interval [CI]: 3.1-3.7) per 1000 and varied by site, ranging from 2.9 (Wisconsin) to 3.8 (Georgia) per 1000, 8-year olds (p<0.02). Approximately 81% had spastic CP. Among children with CP, 8% had an autism spectrum disorder and 35% had epilepsy. Using the GMFCS, 38.1% functioned at the highest level (I), with 17.1% at the lowest level (V). Fifty-six percent were able to walk independently and 33% had limited or no walking ability.

INTERPRETATION: Surveillance data are enhanced when factors such as functioning and co-occurring conditions known to affect clinical service needs, quality of life, and health care are also considered.

I will admit that I did not know that the prevalence of ASD amongst those with cerebral palsey was so high, 8%. It is interesting to note that there is a geographic variation in the prevalence estimates of CP, from 2.9 to 3.8 per 1000. This is not quite as large as the spread in autism prevalence estimates by state, but it is pretty big.

In Prevalence of autism spectrum disorders and influence of country of measurement and ethnicity , the authors look at ASD prevalence by country and ethnicity. They hypothesize that “methodological factors, socioeconomic variables, and bias” play a role in the variability in autism prevalence.

The disparities by geography and by ethnicity within the data reported within the U.S. has been a big concern of mine for some time. Clearly there is not an obvious difference between, say, New Jersey (with an estimated prevalence of 10.6/1,000) and Alabama (with an estimated prevalence of 6/1,000) to account for the large difference in estimated prevalence.

Background
The prevalence of autism spectrum disorders (ASD) is generally somewhat lower in countries outside of North America and Europe. While there are culture-specific patterns of social cognitive processing, the influence of such patterns upon ASD prevalence has yet to be fully explored.
Methods
A comprehensive literature search for original articles reporting ASD prevalence was undertaken. Data across studies were compared with a particular focus on variables of geographic residence and ethnicity.
Results
ASD prevalence varies across countries in a manner that appears to suggest that the greatest influence is due to methodological variables. The nature of a potential influence of culture-specific patterns of cognitive processing upon prevalence remains unknown. The available little data concerning the association between ethnicity and prevalence are limited to studies within the United States (US) showing differences in children of Hispanic descent relative to Whites, a finding for which a definitive explanation is lacking.
Conclusions
Available evidence suggests that methodological factors are largely responsible for differences in ASD prevalence across studies. The much discussed increase in prevalence in ASD has been observed worldwide, suggesting that the refinement of diagnostic methodology and/or broadening diagnostic concept is not limited to Western countries. Within individual countries, only in the US has the influence of ethnicity upon ASD prevalence been examined in depth. In the US, children of Hispanic descent have the lowest prevalence of ASD, while Whites tend to have the highest prevalence of ASD. Hypothesized etiological factors for such prevalence differences include methodological factors, socioeconomic variables, and bias.

In Racial Disparities in Community Identification of Autism Spectrum Disorders Over Time; Metropolitan Atlanta, Georgia, 2000-2006. African American (non Hispanic Black) students are less likely than Non Hispanic White students to be identified with “less severe” ASD’s. This even after controlling for socioeconomic status.

Racial Disparities in Community Identification of Autism Spectrum Disorders Over Time; Metropolitan Atlanta, Georgia, 2000-2006.

Jarquin VG, Wiggins LD, Schieve LA, Van Naarden-Braun K.

From the Division of Birth Defects and Developmental Disabilities, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, GA.
Abstract

OBJECTIVE: Past research indicates that non-Hispanic black (NHB) children are less likely than non-Hispanic white (NHW) children to have an autism spectrum disorder (ASD) diagnosis, even if they seem to meet criteria for the disorder. This study examined differences in community identification of ASDs between NHB and NHW children identified by a population-based surveillance system.

METHODS: Participants were identified as an ASD surveillance case by the Metropolitan Atlanta Developmental Disabilities Surveillance Program in surveillance years 2000, 2002, 2004, and 2006. Health and education records were abstracted and reviewed to determine ASD surveillance case status; community identification was defined by a documented ASD diagnosis, special education eligibility, and behaviors noted in records. Children were placed in 1 of 5 mutually exclusive categories on the basis of ASD specificity.

RESULTS: Total ASD prevalence was higher for NHW than NHB children, but NHB children were more likely than NHW children to have autistic disorder and autism eligibility at a public school documented in records. NHB children were less likely than NHW children to have pervasive developmental disorder-not otherwise specified and Asperger’s disorder documented in records, even after controlling for socioeconomic status. NHB children were more likely than NHW children to have co-occurring intellectual disability.

CONCLUSION: NHB children were less likely than NHW children to have been identified with less severe ASDs, which might have prevented or delayed intervention services that would have catered to their needs. This study illustrates the need for continued professional education, particularly concerning milder ASDs in minority groups.

As a society, we have decided that autistic students often need educational supports distinct from those of children with other disabilities. Clearly if we are to serve our students appropriately, we should be accurately identifying each student’s disability (where they exist). The fact that we are not uniform in identifying autistic students indicates that we have far to go in this regard.

There has been a fair amount of discussion in the media and online about the possibility of an autism cluster amongst the Somali-American population in Minneapolis, Minnesota. A report came out last year, Autism Spectrum Disorders Among Preschool Children Participating in the Minneapolis Public Schools Early Childhood Special Education Programs, presenting data on the administrative prevalence of autism in the public schools. They found that, indeed, a higher fraction of Somali-American children were in the autism programs than children of other racial/ethnic backgrounds.

They analyzed the data in a number of ways, but here is one results table:

The administrative prevalence is about 1% in this analysis, which is significantly higher than that for the other subgroups. Many have commented that the Somali-American administrative prevalence is about the same as that for older children in the general population within the Minnesota public school system. This raises the question as to whether the prevalence is actually higher amongst Somali-Americans, or whether it could be that Somali-Americans are identified earlier.

Rather than wait a few years for that question to sort itself out, NIMH, Autism Speaks and the Centers for Disease Control are teaming up to study the situation. They will use the same methodology as used for the CDC’s autism prevalence figures: the Autism and Developmental Disabilities Monitoring (ADDM) Network methods. The ADDM uses a review of medical and educational records, not a direct examination of the children involved. But, this is an established methodology.

Here is the post from the NIMH. (copied in full because, well, it’s the government and therefore public domain):

Members of a Federal Advisory Group on Autism Mobilize in Response to a Mother’s Plea

NIMH, along with three other NIH Institutes, will be supporting a joint effort with the Centers for Disease Control and Prevention (CDC) and Autism Speaks, a private advocacy organization, to investigate reports of elevated prevalence of autism spectrum disorders (ASD) among children born to Somali immigrants living in Minneapolis, Minn.

At the October 2010 meeting of the Interagency Autism Coordinating Committee (IACC), Idil Abdull, a Somali parent and founder of the Somali American Autism Foundation, described the disproportionate numbers of Somali-American children enrolled in preschool ASD special education programs—up to seven times higher than their non-Somali peers, according to a 2009 report by the Minnesota Department of Health. Committee members, which include NIH, CDC, and Autism Speaks, responded immediately by identifying ongoing research that may be expanded to help answer why such a disparity appears to exist, as well as to determine the service needs of children with ASD in Minneapolis and their families.

“This shared effort demonstrates how members of the IACC can respond quickly and cooperatively to an issue brought to the Committee by the public. An increased prevalence of ASD among this specific Somali population would represent both a scientific opportunity and an urgent public health need,” stated Thomas R. Insel, M.D., NIMH Director and chair of the IACC.

With plans to start in early 2011, the three partnering organizations will provide support to the Minnesota Department of Health to collect prevalence data according to the CDC’s established surveillance methods, as used by the Autism and Developmental Disabilities Monitoring (ADDM) Network.

The IACC is a Federal advisory committee that coordinates all efforts within the Department of Health and Human Services concerning ASD. In addition to NIMH, the NIH Institutes supporting this research initiative are:

* Eunice Kennedy Shriver National Institute of Child Health & Human Development
* National Institute of Environmental Health Sciences
* National Institute on Minority Health and Health Disparities