Skip to content

Is Maternal Obesity Linked to Autism? A Look at the Research

May 15, 2012

Today’s Autism Week post comes from Kala, a friend and regular reader and commenter on Fierce Fatties, who’s grasp on statistical analysis I frequently rely on to ensure that I’m accurately describing the research I’ve analyzed. I’ve asked Kala to read this study and comment on it because this issue is too important for amateurs to handle. 

Disclosure: It is not my purpose today to debunk this study, or to nitpick the methods used, because I personally believe it to be a good study, written by responsible scientists. I feel that the bulk of the problem surrounding the title issue is the attention it caught from the media, and the misinterpretations that reverberated through many circles because of that attention.

What I hope to accomplish with this article is to inform readers on the study that was released in early April in the journal of Pediatrics, which examined maternal obesity, among other variables, during pregnancy as a risk factor for autism and developmental delays. This paper spurred a sudden outbreak of media attention on the connection between obesity and autism. You may remember such pithy headlines as:

Such headlines might seem at first glance to indicate that maternal obesity and autism are inextricably linked, but further reading of the actual articles showed a more nuanced, if not complete, understanding. (Warning: The vast bulk of the comments on any of these articles were unenlightened, and I don’t recommend reading them for those who are triggered by insulting language and general negativity).

The big question that motivated this study is whether the inflammatory response in a mother’s body, spurred by metabolic disorders, can affect the neurological development of children in utero. The question that follows is whether the increasing incidence of metabolic disorders in Americans is partly responsible for the rising rates of diagnoses for autism spectrum disorders (ASDs) and developmental delays (DDs). It’s common knowledge that the health of a mother at the time of pregnancy is a factor regarding the health and subsequent development of her child, but we are far from understanding everything is on this topic.

This particular study was a statistical analysis on the correlation between ASDs and DDs and three metabolic disorders: type 2 and gestational diabetes, hypertension, and obesity (defined as a BMI over 30 prior to the pregnancy). The study looked at the prevalence of these developmental disorders (the focus of my article) and also specific scores for development within groups of children with and without these disorders (not our focus today).

It’s worth noting the debate as to whether or not obesity can be considered a metabolic disorder or whether the BMI category of obesity itself is significant, but that debate is not the topic of my article. However, it does lead us toward two major points which indicate that this article is a very preliminary study only. The most important point, and one that isn’t clear until the end of the paper, is that obesity was chosen because of its high correlation with insulin resistance. Thus, obesity is used as a stand-in for insulin resistance. Of course, we already know that there’s currently a chicken and egg argument surrounding the relationship between insulin resistance and obesity, making this proxy association a bit misleading to those who aren’t well-versed in this area.

However, the reason proxies were chosen was because the biological measurements that would indicate insulin resistance were not available for the mothers who participated in this study. The lack of biological measurements leads us to the second important point: there are no distinctions made between mothers with well-managed diabetes or insulin-resistance, and those whose metabolic disorders were poorly managed. Along with these two major limitations, there are a few other author-acknowledged weaknesses to the study, but the two I’ve outlined above are the most important points that weaken much of the media hype.

Setting aside the major weaknesses, what were the actual conclusions from the study? The association between diabetes and ASDs did not reach statistical significance, which means that this study did not find a correlation between the two. Statistical significance is a metric that indicates whether a given result occurred from random chance. For DDs, those with diabetes were 2.3 times more likely to have children with developmental disorders. With respect to hypertension, no significant correlation was seen between either ASDs or DDs. Finally, a correlation was found between obesity and both ASDs and DDs. Obese mothers were 1.67 times more likely to have children with ASDs and 2.08 times more likely to have children with DDs. All of these measurements were made relative to the rate of incidence seen in the control (typical development) group.

These relative risks have to be taken into perspective. If something is twice as likely to happen to a group of interest in comparison to a control, then the risk is multiplied by the actual prevalence in the control group. That prevalence might be 1 in 1,000,000, or 1 in 1,000, or 1 in 100. I have seen several values pointed out for the prevalence of autism in the general population, but all numbers were on the low end of the order of magnitude of 100 (meaning somewhere in the low single digits or 1-2%).

Based on this one particular study, there is absolutely no reason to believe that obesity is in any way causative to autism. There’s no reason to believe that well-managed diabetes or insulin resistance is causative to autism. This research merely sought to examine whether there was a significant correlation between metabolic issues in mothers and developmental disorders in their children. It’s nothing more than a starting point for further research.

No realistic conclusions made because of it.

No public health campaigns launched because of it.

No criticism of your body or your lifestyle justified because of it

And if anyone says differently, they’re talking out of their ass.

3 Comments leave one →
  1. vesta44 permalink
    May 15, 2012 12:47 pm

    In other words, it’s a starting point. It gives them an idea of what to look for, what to control for, who to include, how to define parameters, etc when designing further studies, if they decide to do more studies on this.

    • Kala permalink
      May 15, 2012 12:50 pm

      Pretty much. The main advantage that the author states in the conclusions, over previous work, is the creation of a representative population and the consistent and well defined autism and DD diagnoses (because this data is from CHARGE [Childhood Autism Risks from Genetics and the Environment], the focus was on these issues and they were handled well).

  2. May 15, 2012 11:27 pm

    Thank you so much for this exquisitely simple analysis.

    In short, this research is weak-ass cheese and we can hold on the panic until they bring back some more substantial research.

    The problem isn’t the research, it’s the media response that is the problem. Giving a non-scientific public this kind of information, when most people only catch the headlines, is adding fuel to the fire.

    I’m so glad you did this, Kala. It’s such an awesome foundation to build on Autism Week. 🙂


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: