John Rodakis wasn't looking to launch an investigation into autism.
But that's what happened after the Dallas man began a quest to understand why his son's autism symptoms changed dramatically while taking an antibiotic for strep throat.
Three clues led Rodakis to question traditional autism research.
Matt could read before he could speak.
"I had suspected this when we would read stories and he would turn the pages," John Rodakis says. "But I had just assumed he just knows when it's time to turn the page."
Rodakis, who studied molecular biology and worked at a large medical venture capital firm, loves to run experiments. So Rodakis and his wife, Christi, tested their son. To protect the boy's privacy, KERA has agreed to call him Matt.
They wrote words like "run" and "dance" on a board and held it up in front of him. Without saying anything, Matt ran and danced when he saw the words.
The reading skill is called hyperlexia, and usually it's associated with autism.
Not long after, Rodakis and his wife noticed serious changes in their son.
"Within the course of a month, he stopped making eye contact," Rodakis says. "He became quite rigid, an insistence on sameness, and when things didn't go the way he wanted, he would throw fits and temper tantrums."
Matt, who's now 5, has autism spectrum disorder. An estimated one in 68 children has it. The disorder varies greatly among kids. Classic features are impaired communication, poor social engagement and repetitive behavior.
Autism is one of the most divisive medical conditions out there. It separates scientists from scientists, and parents from parents. After decades of research, we still don't know what causes it. And when you don't know the cause of a disorder, you don't know how to treat it.
You don't know how to help kids like Matt.
In the absence of answers to what's behind autism, theories have flourished.
"There's been many waves of popular thinking about what causes autism," Dr. Wendy Chung says. "Everything from problems or bad parenting to vaccines to a shift in terms of thinking that it's genes playing a large role."
Chung is the director of clinical research at the Simons Foundation Autism Research Initiative. The conventional view of autism today is that it's mostly a genetic disorder. It's something kids often inherit and are born with. Their brains are hardwired differently.
Rodakis and his wife accepted that idea. They began taking Matt to therapy. Speech therapy, occupational therapy and physical therapy for dozens of hours each week.
Then they started to question the conventional view of autism.
They began to notice their son's symptoms changed depending on environmental factors – like what he ate and what medicines he took. In particular, there were three clues that led Rodakis to believe his son could get better.
Clue No. 1: The Fever Effect
Every time Rodakis' son had a fever, he noticed transient improvements — in his speech, mood and eye contact.
"It was noticeable enough that I went to the medical literature to see if there was anything about this, and sure enough, there was," Rodakis says.
Now, most parents don't study publications like the New England Journal of Medicine or The Lancet in their free time. Rodakis is not like most parents. He's trained as a molecular biologist – and wedded to his hardcore scientific roots. In the living room of his Dallas home, Rodakis flips open a binder filled with the latest publications on autism research.
Data drives him. And he keeps track of his son obsessively. He tracks Matt's rigidity, how agreeable he is, his happiness and energy level. Rodakis exports this daily data to Excel and creates charts and graphs to monitor his son's progress.
Which brings us back to the first observation. When Matt had a fever, he was more agreeable, more talkative, and there was one doctor who had studied this: a pediatric neurologist named Andrew Zimmerman.
Zimmerman has cared for thousands of autistic patients and published more than 70 peer-reviewed papers and served as director of medical research at the Kennedy Krieger Institute in Baltimore.
Zimmerman has a name for this change in behavior among kids with autism. He calls it the "fever effect."
"If you ask people coming into an autism clinic if they've ever seen [the fever effect], about 20 to 30 percent will tell you they have," Zimmerman says.
In 2007, Zimmerman and a group of researchers evaluated 30 kids with autism from age 2 to 18 during and after fever.
Results published in the journal Pediatrics showed fewer autistic-like behaviors for children with fever compared to controls. This was compelling, but it was a small study, and Zimmerman wasn't about to advocate giving fevers to kids with autism.
Instead, he worked with colleagues at Johns Hopkins to isolate a molecule called sulforaphane that mimics the fever effect. They used that molecule to do a double-blind study. The results appeared in the journal Proceedings of the National Academy of Sciences.
"We found that sulforaphane improved two-thirds of men who received the drug compared to those on placebo," Zimmerman says.
How might a fever, and a molecule found in broccoli sprouts, reduce autism symptoms?
Here's how Zimmerman explains it: Along with the bad stuff, fever causes cells to produce more energy, to be more active (this is also known as a cellular stress response). He thinks that might help troubled cells ramp up production and work properly.
There's a big idea here – the idea that there are ways to improve autism symptoms without altering genes.
For dad John Rodakis, the fever study provided hope. But it took two more clues to convince him. It would take more than conventional research to understand kids like Matt.
Clue No. 2: The Antibiotic Effect
For the Rodakis family, Thanksgiving 2012 was supposed to be a quiet holiday. Instead, both Matt and his little sister got strep throat while visiting Grandma out of town. Both were prescribed antibiotics — a 10-day course of amoxicillin.
It was on day four that Rodakis noticed changes in Matt. The normally low-energy, lethargic 3-year-old was running around.
"And more importantly," Rodakis says, "he was happy and smiling and stringing words together in ways he never had before."
Each day got better, Rodakis says. "He was looking into my eyes and happy and telling me 'I'm here -- talk with me, play with me!' And the changes were pretty dramatic."
Then, on Day 6, his son hopped on a tricycle he'd been unable to ride and rolled down the hallway. They'd seen those mild improvements with the fever effect, but this was different.
"It's like the veil of autism was lifted magically," Rodakis says.
It turns out many children with autism have behavior changes with antibiotics — both positive and negative.
Dr. Richard Frye is a pediatric neurologist and director of research at Arkansas Children's Hospital Autism Center. Every so often, he hears from parents who notice their kids behaving differently while on antibiotics.
"One of the issues with somebody saying I gave my child this antibiotic and they got better is that they're giving [the antibiotic] for a reason, the child is sick, so the child may have a fever, may have other types of immune processes going on," Frye said. "So it's important for any type of treatment that it is done systematically."
The only scientific study that's looked at antibiotics and autism is one from the late '90s, published in the Journal of Child Neurology.
It was meant to be a pilot study, but raising money for a follow-up proved too difficult. Ten boys with autism were given the antibiotic Vancomycin. Eight had noticeable improvements.
"They would now look at people that spoke to them, they would let their mother kiss them, things that just wouldn't happen before," Dr. Sydney Finegold says.
Finegold has been researching antibiotics for half a century. His trial with Vancomycin was controversial, and he had trouble getting it published, he says, because the journals were reluctant to believe the improvements took place.
When Finegold brings it up at meetings, some pediatricians still slough it off. But others have embraced the idea that antibiotics can influence autism symptoms.
How? By affecting the gut.
That brings us to clue No. 3 for John Rodakis. Remember Clue No. 1 -- the fever effect. Clue No. 2 -- the antibiotic effect. Then there was the food factor.
Clue No. 3: The Food Factor
Antibiotics, like food, influence the microorganisms that live in our guts. After Rodakis saw the change from the strep throat medicine, he decided to try changing his son's diet.
"The first thing we tried was we took gluten out of his diet," Rodakis says. "We didn't really see any improvements. Later, we tried taking dairy out of his diet, and within about three days we saw pretty dramatic improvements in speech and eye contact."
Studies have shown that children with autism spectrum disorders commonly have gastrointestinal, or GI, symptoms. Some experts have suggested that gluten-and casein-free diets could help improve GI symptoms and behavioral problems.
After seeing Matt improve with a fever, then antibiotics, then removing dairy, Rodakis says he came to the conclusion, at least for his kid, that there is more going on than simply genetics.
Rodakis is trying to getmainstream medicine to think about autism in new ways. He never mentioned vaccines as a cause of autism and isn't advocating antibiotics as a cure. Instead, his thinking reflects a larger shift in autism research. A shift away from genetics as the sole explanation for autism, towards a complex interaction between the environment and genes.
"I think there's been the promise that genetics will tell us so much about medical diseases," Dr. Richard Frye says. "If you look at heart disease, diabetes and such, there's no gene – there's no diabetes gene really, but what we do know is that there's a big influence between susceptibility factors and environment."
Think about it: Your genes may predispose you to heart disease. Really, though, your lifestyle and what you eat -- those factors can determine if you'll actually get heart disease.
An emerging idea for what's behind some cases of autism isn't that different. The thinking goes that what we put in our stomachs — pills, pizza or carrots — alter the bacteria that live in and on us, and that alters our behavior.
So when Rodakis, the Dallas dad, noticed those changes in his son, he offered to send a fecal sample to a microbial ecologist in Canada.
Emma Allen Vercoe studies the communities of bacteria, called microbiota, that live in our guts. Vercoe likes to think of the microbiota in our gut as a rainforest.
"And the beauty of a rainforest ecosystem is it is very stable ecosystem, the stability comes from the fact that this is a diverse ecosystem, many species all acting to give a net effect of the output of the ecosystem," she says.
Vercoe says when we change our diet – like Rodakis did with his son — or take antibiotics – like strep throat medication — it can clear-cut swaths of the rainforest. It can disrupt the entire ecosystem.
"What's really exciting is we've completely overlooked the fact that microbes and humans talk to each other," Vercoe says. "There are specific pathways that relay information from the microbes in the gut to the brain and have effect on the gut and the brain, and all of that's been completely overlooked."
A major focus of research in autism has been on finding the underlying genetic causes, with less emphasis on potential environmental triggers or causes.
Another microbiologist in Australia is also looking at the microbiome's connections to autism.
In 2012, Enzo Palombo compared more than 50 children with autism and 50 unaffected siblings, and found that there were no big differences in microbiota. Before you jump to conclusions, consider what he has taken away from the study, published in the journal Autism Research:
"I think the jury is still out whether individual microbes are contributing to autism spectrum disorder," Palombop says. "Our gut microbiota is very complex. There are numerous species, both bacterial, fungal -- even viral -- microbes in the gut, and depending on how they talk to each other and communicate with the host, they do produce compounds, chemicals, metabolites which may themselves be the important factors rather than the microbes themselves."
Regardless of whether the microbiome turns out to explain autism for certain children, Palombo and other researchers agree that research on environmental factors is important.
"This is helping to add another layer to the understanding," Palombo says.
Rodakis couldn't agree more. He's on a mission to raise money for autism research that looks beyond genetics. After going back to school to take advanced biology classes and publishing an article in the journal Microbial Ecology, he started a nonprofit, called N of One.
"We favor research that treats these phenomena as clues -- following the clues that are in front of us," Rodakis says.
Clues like a fever, a new medication, a different diet.
Clues not always visible through a traditional research lens, but that often come into focus through the eyes of a parent.