Oxytocin is a mammalian hormone closely associated with labor and delivery. It is normally released by the mother's body in large quantities during childbirth in order to trigger uterine contractions, and often serves a role in breastfeeding. It is also believed to have an impact on social interaction, leading to a potential link between oxytocin and autism. Synthetic oxytocin also goes by the trade names Pitocin and Syntocinon. Synthetic oxytocin is frequently used to medically induce or augment labor.
Natural childbirth advocate Michel Odent, MD, has a pet theory that the use of synthetic oxytocin during labor could contribute to an increased risk of autism in the baby. The suggested mechanism is based on the role that oxytocin appears to play in social bonding and the idea that a significant increase in the amount of oxytocin the baby is exposed to could suppress the response of the baby's oxytocin receptors.
A recent study published in JAMA Pediatrics in August 2013 has shown a statistical link between children diagnosed with an autism spectrum disorder, and use of synthetic oxytocin to induce or augment labor. Despite the study's conclusion stating explicitly that "further investigation is needed to differentiate among potential explanations of the association including underlying pregnancy conditions requiring the eventual need to induce/augment", the media has seized on this study with headlines that will mislead and frighten the general population who don't understand how science (or reality) works.
The fact is that there is absolutely no conclusive evidence that synthetic oxytocin or labor induction/augmentation increases the risk of autism. While the science is still out on the matter, the current hysteria is based almost entirely on one study which showed a correlation, but no causal link and on the untested theories of one doctor who is already biased in favor of natural childbirth.
 Augmentation and induction
Pregnant women who do not naturally begin labor within a typical time frame, or who have complications which indicate a necessity for starting labor early, may have their labor induced through the administration of synthetic oxytocin. Additionally, women who are unwilling to wait for their bodies to decide when the fetus is done baking can schedule an induction by similar means. (Note that synthetic oxytocin is not the only way in which labor may be induced).
Women who do enter labor naturally, but whose labor then fails to progress, or who for medical reasons require labor to be expedited, can receive labor augmentation through synthetic oxytocin or other means. Oxytocin can trigger or strengthen uterine contractions, thereby speeding up the labor.
 Michel Odent's theory
Michel Odent is a noted medical doctor and advocate of natural pregnancy. He has theorized that increased and different exposure to oxytocin during labor could interfere with the child's natural oxytocin system. Because oxytocin is believed to be involved in social interactions, particularly in bonding, it is therefore suggested that such interference could lead to the type of social symptoms typical of autistic spectrum disorders.
While oxytocin is normally released naturally in large amounts from the mother's pituitary gland during labor, Dr. Odent suggests a number of ways in which synthetic and administered oxytocin behaves differently. Many of these differences affect the mother but not the child, such as the fact that administered oxytocin is injected intravenously and therefore cannot reach the mother's brain, unlike naturally released oxytocin. However, two major described differences which may affect the child are that administered oxytocin is given continuously, while naturally released oxytocin is released in waves; and that natural contractions are different from synthetically induced contractions, and induced contractions could limit blood flow to the child when too high a dose is given.
There has been nothing specifically put forth to contradict Odent's theory, but there has also been very little research on the subject, and most opponents of synthetic oxytocin for induction or augmentation of labor simply cite Odent.
 The JAMA Pediatrics North Carolina study
The study included 625,042 subjects born in North Carolina between 1990 and 1998, including "more than" 5500 who had been diagnosed with an autism spectrum disorder. From their analysis, they found that boys who were born by induced or augmented labor had a 35% higher incidence of being diagnosed with autism than those who were not. They also found increased incidence among girls only among those whose birth was augmented. The correlations held even "after controlling for potential confounders related to socioeconomic status, maternal health, pregnancy-related events and conditions, and birth year" (though it is not clear whether the specific figure of 35% is with or without this adjustment).
The study concludes with:
While these results are interesting, further investigation is needed to differentiate among potential explanations of the association including underlying pregnancy conditions requiring the eventual need to induce/augment, the events of labor and delivery associated with induction/augmentation, and the specific treatments and dosing used to induce/augment labor. 
It is important to note from this that the study's authors themselves explicitly deny that their study suggests a causal link between induction or augmentation and autism. They stress that the study only highlights an association (i.e., a correlation) between the two which may be explained by an underlying factor contributing to both.
Furthermore, the conclusion makes clear that their study did not revolve around the use of synthetic oxytocin, and in fact did not differentiate between different methods for induction or augmentation.
 Interesting Points
In the introduction to the study, the authors indicate that one in 88 children in the US are diagnosed with an autism spectrum disorder (ASD), yet the data they studied shows some variance from this. The introduction specifies that they studied 625,042 subjects, and of these, "more than" 5500 were diagnosed with ASD. Even if we assume that "more than" 5500 is as high as 5599, the overall incidence of autism diagnoses in the studied cohort is less than 1 in 111 (if "more than" 5500 is only 5501, the incidence is even lower, less than one in 113). This suggests that the group studied is not necessarily typical, and whether or not the correlations determined by the study apply in other cases is left unanswered.
Another point to keep in mind is that the study only considered subjects between the ages of 15 and 23 years old. One possible explanation is that induction or augmentation may cause individuals who are going to develop autism anyway to develop or display it earlier in their life than they otherwise would. This would lead to an increased rate of diagnoses for subjects in this age range who were induced or augmented, while the overall incidence of diagnoses for this group throughout their entire life remains the same as the general population. As with all current explanations for the association described by the study, this is only conjecture and one of many possible explanations.
Lastly, consider the distinction between those who have autism, and those who are diagnosed with autism. Depending on a wide variety of factors, these two numbers could be different. In some cases, people with autism (especially high functioning autism) may not be diagnosed, or may be misdiagnosed with another disorder. In other cases, people who do not actually have autism may be misdiagnosed as having autism due to things like increased attention on the disorder, pressure from parents or teachers, etc.
This type of discrepancy can contribute to misleading correlations. For instance, one possible explanation for the association is that mothers who had induced or augmented births a more likely to have a higher regard for modern/western medicine and are therefore more likely to have their child tested for autism, whereas mothers who opted for at-home or natural births may be more likely to resist having their child tested or diagnosed. Alternatively, the fact that the labor needed to be induced or augmented could leave the parents feeling more anxious or concerned about the wellness of the child, and therefore lead to an increase in medical monitoring and intervention, thus leading to an increased chance of diagnosis. One must also consider the possibility of a self-fulfilling prophecy at work, in which the belief that induction or augmentation leads to an increased risk of autism will encourage mothers whose labor was induced or augmented to have their children tested, once again leading artificially to an increased incidence of diagnoses.
 Fallacies, inaccuracies, and misleading statements (oh, my!)
As with any emotionally charged issue (or for that matter, any issue at all), the field is ripe with flawed reasoning, data and quote mining, and misleading statements. Here are a few examples to give you an idea of what to look for.
- The most common and important fallacy associated with this debate (and many others) is the mistaken belief that a correlation between two data implies that one causes the other. Even the authors of the much cited JAMA study describe above specifically state that "further investigation is needed to differentiate among potential explanations of the association". In other words, even if the correlation between induction and autism turns out to be genuine, it could be because an underlying cause contributes to both. In other words, something that causes the child to be predisposed to developing autism may also interfere with natural labor, thus increasing the chances that labor will need to be induced or augmented.
- This article from USA Today actually does a good job with this. They reference the lead author of the JAMA paper when they say that "The increased autism risk in the JAMA study likely stems from an underlying problem with the pregnancy, rather than any of the methods used to jump-start labor", and go on to explain it with "It's possible that 'infants destined to develop autism are less likely to send out the correct biochemical signals for normal progression of labor,'" quoted from Tara Wenger, a pediatric genetics fellow at the Children's Hospital of Philadelphia. They finish up with: "Authors of the JAMA study note their research doesn't definitively prove a link betwen [sic] labor induction and autism. They say doctors shouldn't change the way they manage labor and delivery based on their study".
- The headline of this article from Natural News says simply "Inducing childbirth causes autism", and sites the JAMA Pediatrics study discussed above. Despite the opening paragraph of this article cites the study as "suggesting inducing childbirth increases risk for developing autism" (emphasis added), it also goes on to say that "Results from the analysis proves that" induced children have increased odds of developing autism (emphasis added to quote).
- In reality of course, the JAMA study doesn't prove anything, it merely provides statistical analysis of the data. Furthermore, whatever evidence the study does provide does not even suggest that induced birth increases the risk of autism, it merely shows a correlation between the two. Even if there is a causal link between induction and increased risk, it's still a leap to go from there to saying that induction causes autism (which strongly implies that if you induce labor, the baby will develop autism).
- This article from WND repeatedly confounds having a disorder with being diagnosed with the disorder, claiming that in "1980, one child in 5,000 had autism", while in "in 2010, one in 110 children had autism", in order to show a correlation between increased incidence of autism and the use of synthetic oxytocin during labor.
- In reality, it is hard to know whether the actual occurrence of autism has increased during this time, or simply the rate of diagnosis. One thing than can be said categorically is that if actual occurrence has increased during this time, it is not as dramatically as the numbers suggest, since the definition for autism has broadened significantly during that time period. Other possibly reasons for increased diagnosis include more effective ways of identifying the disorder, more awareness so that more people are seeking diagnosis, and the fact that autism has been "trendy", leading to more doctors looking specifically for symptoms of the disorder, and an increased chance of incorrect, dubious, or border-line diagnoses.
- The same article from WND changes their statistics to make it look like the trend is stronger than it may actually be. After citing the rates of children "with" (read "diagnosed with") autism during several different years over a thirty year period, they conclude with the rate of boys (diagnosed) with autism spectrum disorders in 2013.
- The cited statistics compare a rate of 1 in 110 children (0.91%) in 2010 with a rate of 1 in 70 boys (1.4%) just three years later, seeming to conclude with a 53% increase in just three years. However, with the information provided, such a conclusion is unjustified, as the incidence among boys may have always been higher than the average among all children. Without more information, it is impossible to conclude anything for the cited statistics, except that the author of the article either has a very poor grasp on statistics, or was being deliberately misleading.
- Furthermore, the latter statistic (2013) has broadened the category under consideration. While the 2010 statistic was cited for subjects with autism, the 2013 statistic for subjects who "are ASD" (ASD meaning autism spectrum disorder). Without a definition allowing us to compare autism with ASD, the comparison is again meaningless.
- ↑ 1.0 1.1 1.2 http://www.wombecology.com/?pg=oxytocin
- ↑ http://theperfectbirth.wordpress.com/2011/10/25/a-woman-mdob-who-believes-in-odent-oxytocin-physiology/
- ↑ 3.0 3.1 3.2 3.3 3.4 "Association of Autism With Induced or Augmented Childbirth in North Carolina Birth Record (1990-1998) and Education Research (1997-2007) Databases"; Gregory, Anthopolos, et al; JAMA Pediatrics, 12 Aug 2013
- ↑ 4.0 4.1 4.2 http://www.webmd.com/baby/news/20130812/induced-labor-linked-to-raised-risk-of-autism-study-suggests