I can see this research being reported as "Bisphenol-A causes autism", but this isn’t what the work claims to have found. 

This study looks at possible links between Bisphenol-A (BPA) and autism. BPA is a chemical used to manufacture plastics and epoxy resins as well as various other industrial uses. It is one of several chemicals in the Bisphenol family - other members include Bisphenol-S (BPS) and Bisphenol-F (BPF).

The authors first found a possible association between autism and BPA exposure in data from a previous study. However, that study, was initially based on self-reporting (later assessed by a clinician) and the BPA was tested by a single measurement in the mother’s urine. If BPA is in the urine, it has been excreted and it can’t be assumed that what is in the urine is directly related to the exposure of the unborn child. Neither is a single measurement representative of the entire pregnancy.

The authors then undertook mouse and cell to explore the possible link. This work was detailed and showed a plausible biological mechanism by which the effect could occur. However, just because something can occur, does not mean that it does so. The levels of BPA in this study were higher than we are generally exposed to. Results from mice and cells only give an indication of what might happen in humans and in this case, there were large overlaps between the behaviours seen in the treated and untreated mice, which means the results are not conclusive.

While this is interesting work no single study, no matter how well conducted can be counted as proof. Further work would need to show similar results. BPA is one of the most well-studied chemicals on earth with thousands of papers published on it and while still the topic of a lot of debate nobody has shown any effects of this compound at the levels to which we are exposed.

Autism is a neurodevelopment disorder with a wide spectrum of behavioural and cognitive changes. Autism results from a complex interaction between genes and the environment, and the nature of the environmental interactions is still largely unclear. Bisphenol A (BPA) is plasticiser used in a variety of plastics and industrial applications. BPA can enter our foods in small amounts from contact with plastic wrapping, plastic bags and the plastic lining on tins. While it is considered an environmental estrogen, BPA is between 10,000 to 100,000 times weaker.

A recent small study in 43 children suggested that male children who had low levels of the enzyme aromatase, an enzyme which converts brain androgens to brain estrogens, and high exposure to BPA based on urinary levels, were more likely to have an autism diagnosis or autism spectrum behaviours. However, the results of the small study were very variable, and the interpretation not straightforward.

This new study has done a series of careful experiments to investigate whether BPA exposure could be modulating aromatase activity to cause long term behavioural changes.  Animals whose genes for the aromatase enzyme had been deleted showed repetitive behaviours, suggesting that changing aromatase levels could induce some autism-like behaviours. This is a technically excellent study with too many aspects to fully summarise, but the key ones follow.

In a tissue culture study, exposure of a model human brain cell line to BPA reduced aromatase expression. However, the lowest concentration used in the study was roughly 25 times higher than plasma concentrations found in the general population and is not likely to realistically model brain exposure.

In mice studies where the mother was treated with 50 μg/kg/day of BPA for around four days mid pregnancy, there were changes in the number of brain cells and their structure in the male offspring. However, the majority of changes had huge overlaps between treated and untreated, and while statistically significant, it is not clear if this is biologically relevant. Furthermore, the doses were the maximum permissible, which most people are not going to be exposed to. The FSANZ did studies looking at human exposure to BPA and concluded that in Australia, human exposures are well below the maximum permitted dose of 50 μg/kg/day of BPA.

What is more, the doses were given subcutaneously, which bypasses the metabolic systems BPA encounters when taken orally. As most human exposures are oral, and humans efficiently metabolise and excrete BPA taken orally, the exposure of the mice to BPA will be higher than humans.
Given that the BPA treated mice have higher exposure than humans could be expected to have, the marginal effect of BPA on the autism surrogate marker, time spent grooming, this is less than compelling evidence that BPA exposure is involved in autism.

Bisphenol A is an industrial chemical that is widely distributed in the environment and known to be present in the bloodstreams of most of us. It mimics natural hormones and can disrupt their action, especially in developing babies and children.

BPA’s most common uses are in plastics—epoxy resins and polycarbonate—and in that form it is quite safe. However, BPA can escape from these plastics and that’s how it is able to roam free in the environment and from there as a trace contaminant in food it can enter our bodies.

Many studies have shown associations between the presence of BPA and a range of abnormal developments. ‘Association’ cannot automatically be assumed to be ‘causation’ but many scientists are working on this question. Suspicions about its toxic effects are strong, but the effects are usually weak and hard to detect, requiring extensive population studies, so it has proved extremely difficult to see exactly how BPA exerts its effects—something that’s need to link cause and effect. 

The researchers who did the work reported in this paper found an association between higher BPA blood levels and the development of autism spectrum disorders in male children. What’s really new about their results is that they were able to pin the effect to a biological pathway that is important in brain development.  In other words, BPA is acting as a ‘rogue’ hormone to out-compete the natural hormone that is usually involved in this pathway.

The case is compelling, they write, and supports broader evidence on the need to further reduce BPA exposure, especially in pregnancy.

In addition to locating the pathway of BPA action, they have found in laboratory studies that an unusual fatty acid (10HDA to avoid a difficultly long name!) competes with BPA and can be administered to alleviate its effects.

The children involved in this study were from Geelong. The authors—all 45 of them—were mostly Australian with just a few Canadian and American contributors.

This is an interesting study suggesting that exposure to bisphenol A derived from plastics in the environment can impact estrogen levels in male infants who may then have a higher risk for being diagnosed with autism. 
 
In mice, the effects of higher levels of bisphenol A on the brain were improved by fatty acid (10HDA) treatment. 
 
Future work to carefully measure bisphenol A levels over time during pregnancy to clarify these findings  as well as further assessing the effects of the 10HDA fatty acid will be informative. The effects of diet may also be important in these findings. 
 
This is interesting research worthy of further investigation but it’s important to understand there are many other genetic variations that are possible contributors to autism that have similar amounts of evidence. Ultimately we still don’t know for sure what causes autism for most people and a normal healthy diet and lifestyle advice should be followed during pregnancy.