Can oxytocin help improve social skills in kids with autism? It's complicated
Oxytocin is a powerful hormone that plays an important role in social bonding. Nicknamed “the love hormone” because it’s released during hugging, it also acts as a chemical reward in the brain during various social interactions. However, researchers are learning that its effects are complex and situation-dependent.
Because of oxytocin’s connection to social behavior, researchers have investigated whether the hormone may play a role in autism spectrum disorder. Some research suggests that there are autistic kids who have less oxytocin than they should, and that boosting oxytocin could provide motivation to engage in appropriate social interactions, but the data have been inconsistent.
Still, this is not the end of the oxytocin story, says Daniel Geschwind, MD, PhD, the Gordon and Virginia MacDonald Distinguished Professor of Human Genetics, Neurology and Psychiatry at UCLA. In an editorial he wrote for NEJM discussing the clinical trial, Dr. Geschwind, who directs UCLA’s Center for Autism Research and Treatment, suggests that although the hormone isn’t effective as given off label in the community for widespread use, it may yet be useful in certain situations.
Q: First of all, what does oxytocin do in a non-autistic or neurotypical brain?
A: That’s a really good question. Oxytocin is a hormone that is involved in lactation. It's released when an infant is suckling on the mother's breast, and it heightens maternal-infant bonding.
Its other functions outside of that aren't that well known. Certain activities like cuddling, gazing into a loved one’s eyes, or petting a dog can increase people’s oxytocin levels. A recent study showed that giving people oxytocin makes them more likely to trust strangers. But its effects differ, depending on the context – for instance, when it's given to men versus women, or if it's given in a positive situation or a fearful situation. Some studies suggest that rather than being a “love” hormone, it actually amplifies whatever a person is already experiencing – the so-called “social reward”.
It really doesn't have a single effect, but it's thought of, especially during development, as important for social bonding.
Q: Is that where the idea came from to treat symptoms of autism with oxytocin?
A: There’s been a couple of decades of study of oxytocin in animals. Differences in oxytocin are related to differences in pair bonding in small mammals called voles. Oxytocin was found to be “prosocial” in the sense that it led to forming monogamous relationships, while changing its effects stopped the voles from forming monogamous pair bonds.
Because a fundamental deficit in autism is social cognition, the thought is that if you can increase social motivation and make social interactions more interesting, then that will improve social behavior.
Q: Has it been observed that people with autism have a deficit of oxytocin, or is there genetic evidence that they don’t make enough of it?
A: There is some genetic evidence, but it’s not super strong. Differences in some relevant genes, like the oxytocin receptor, do contribute to autism risk, but it's not a major risk factor.
When we've looked at mouse models which we've genetically engineered to have mutations that cause autism in people, those mice have abnormal social behavior, such as repetitive movements and lack of interest in playing with other mice. And a handful of those mouse models either respond to oxytocin, or have a deficit in oxytocin, or both.
Q: Looking at this study in the New England Journal of Medicine, which is a double-blinded, placebo-controlled clinical trial, what should people take away from this study?
A: This study was designed to answer the question: “if we give oxytocin twice a day in a nasal spray, would it show an effect on social behavior?” Previously, several much smaller trials had claimed an effect.
This study was large enough to have statistical power. It had a sufficient number of patients enrolled that they could detect the kind of changes that people were claiming. This trial showed that a broad effect wasn't there.
The bottom line is that using oxytocin the way it's being used in the community now is definitely not supported by this study. But that comes with a lot of caveats and is definitely not the last word.
Q: When you say ‘the way it’s being used in the community,’ what does that mean? If a parent wanted to try this for their child, they could get it?
A: Yeah, they could. That's why this study was important to do, because it shows it's not really helping. But oxytocin is not a dangerous drug. And so, sure, it's a waste to give it to people if it’s not helping, but you're not really hurting them either.
It’s possible that a fraction of people are actually benefiting from it, but it would be very hard to know from this study, because they’re testing people between ages 3 and 17. One important point about oxytocin is that its effects are different at different stages of development.
Q: Yeah, it seems obvious that if we're talking about a hormone that's released during mother-infant bonding, it's going to have a very different effect at age 17.
A: Exactly. And that's known from animal studies, too. There are a number of animal studies that show that for a day or two right after birth, oxytocin can have a long-lasting effect, even weeks and months later. But when you give it when the animal is an adult, or even just after a certain point, the effect lasts for maybe an hour and a half.
That’s one of the issues about dosing people. If you have a drug whose effect is under two hours, you need to dose it in such a way that you can keep the levels up all day. In the study, they gave it twice a day, but you’d need to give it maybe five times a day or more. Giving a short-acting drug twice a day is unlikely to have an effect.
Q: What effect is oxytocin supposed to be having in autistic people?
A: It’s all predicated on increasing social motivation.
Let's say you want to see if increasing social motivation would increase social engagement and learning appropriate behaviors. What you’d really want to do is couple oxytocin with optimal behavioral therapy, which is working on increasing social cognition and social skills. You’d get the oxytocin maybe half an hour before the therapy starts and then have therapy for an hour. That would be, in my view, a better way to give oxytocin and to test it.
Q: In your editorial, you say that ‘it’s premature to summarily reject oxytocin’ as a potential treatment for autism. Do you think some people with autism might benefit from oxytocin therapy?
A: This trial shouldn’t cause us to throw out the baby with the bathwater. It behooves researchers to think more clearly and more mechanistically about how oxytocin is working, and how to use it better. I'm not saying it's going to be a panacea, but it's still too early to throw it away.
A major issue is that autism isn't one thing. There are many pathways that can get you to social dysfunction. We now know that, like cancer, autism has many, many different genetic causes, and you won’t treat them all the same way. You don't treat all breast cancers the same: you look at the genetic mutations, and those lead you to different treatments.
It’s the same way in autism. There may be certain genetic abnormalities or biomarkers that signify that oxytocin is low. It’s possible there might be a few percent of autistic children who have a real oxytocin deficit due to a mutation. And there might be others in whom oxytocin will help for other reasons. But it's unlikely to be the majority of people with autism.
We're entering a stage of understanding the causes of autism and other psychiatric diseases in a totally new way. We’re entering an era of precision medicine. So, the best way to do a trial now would be to see if you can identify people who have a mutation that has been associated with an oxytocin problem. Rather than grouping all autistic people together, our hope is that we'll be able to identify specific subtypes that have more of a chance of responding to a particular precision therapy.
Caroline Seydel is the author of this article.