Autism spectrum disorders, or ASD, appear to be increasing in prevalence, if statistics reported by the Centers for Disease Control and Prevention are to be believed. In 2014 one in 68 children in the US was diagnosed with ASD, with boys four times as likely as girls to be affected. That's up from one in 88 children in 2012. Educators, public health officials and most frantically, parents are all trying to discern the antecedents. Now, as Rick and I report on PodMed this week, a very clever Swedish study reported in JAMA may shed some light on the issue.
Investigators examined data from almost 2 million children born in Sweden between 1982 through 2006. Twin pairs, both identical (monozygotic) and fraternal (dizygotic), full sibling pairs, half-sibling pairs with either the same mother or the same father, and cousin pairs were all identified. A total of 14, 516 children were diagnosed with ASD, of whom 5689 had the most severe form, referred to in this paper as 'autistic disorder.'
This study introduced me to a new acronym for assessing disease risk within a genetic model: RRR for relative recurrence risk. Briefly, this statistic purports to pinpoint familial aggregation of disease by calculating the relative risk of a subject with a sibling or cousin with the particular disease or condition as compared to the risk seen in a subject who has no affected sibling or cousin. After adjustments for age, birth year, gender, age of parents and their psychiatric history, RRR was calculated for both ASD and autistic disorder. Since most of us, (including me!) don't really appreciate RRR as informing our understanding of disease risk, instead of those numbers I cite others the authors calculate: For individuals with a full sibling with ASD, the cumulative probability of an ASD diagnosis at age 20 years was estimated to be 12.9% compared with 1.2% for individuals without. The cumulative probability of an ASD diagnosis at age 20 years was 59.2% for monozygotic twins, 12.9% for dizygotic twins, 8.6% for maternal half siblings, 6.8% for paternal half siblings, and 2.6% for cousins.
There, that sums up nicely, I think, the kind of risk explanation parents might find useful when attempting to understand the risk of having a second affected child when they've already had one. The other number this study provides is an estimate of the impact of genetic factors versus environmental factors in ASD: 50%. That is, based on this data the authors calculate that 50% of the blame, if you will, rests within the genes, while the remaining 50% can be attributed to environmental influences.
Here's one really interesting statement the authors make: The male:female ratio was 2.7 for ASD cases and 2.4 for autistic disorder cases. This statement seems to support one of Rick's concerns, that because this study is limited to Swedish children, we may not be able to extrapolate the conclusions to more genetically diverse populations, since the male:female ratio here in the US is reported as 4:1. We agree that the study is important but clearly begs the question of the identification of significant environmental influences.
This week we also look at increasing rates of diabetes in US youth, in JAMA, and reduced mortality when health insurance is mandatory in Annals of Internal Medicine, as well as vaccinating pregnant women for pertussis, also in JAMA. Until next week, y'all live well.