John Hussman: Cracking the autism code

April 26, 2011: 5:00 AM ET

Investor John Hussman has used intense research to build a strong record at the $6 billion Hussman Strategic Growth fund. Now he's using those same skills to take on autism.

By Scott Cendrowski, writer-reporter

John Hussman

Hussman's breakthrough findings were published in a journal called Molecular Autisim.

John Hussman lives by research. It's how he analyzed markets as an economics professor. It's how he returned 5% annually over the past decade (and beat the S&P 500) at his eponymous $6 billion mutual fund. It's how he predicted a recession in late 2007 (and why he thinks stocks are overpriced today). And it's how he's attacking an unusual adversary for a fund manager: autism.

Over the past 15 years Hussman, 48, has thrown himself into understanding the disorder -- ever since his son, J.P., was diagnosed with it as a child. Because the vast majority of cases show a family link (a relative on J.P.'s mother's side is autistic), Hussman has focused on its genetic causes. Almost completely self-taught, he's spent nights and family vacations poring over obscure scientific textbooks (he says he's Amazon.com's best customer) and research abstracts. He not only funded a genomics center at the University of Miami, but has spent years collaborating with scientists there.

Finally, this year the science journal Molecular Autism published his findings in a paper that top researchers are calling a breakthrough. The dense, 16-page article features Ph.D.-level mathematics and abstruse genetic analyses. (We'll spare you the details, but you can check it out here.)

"I found it really interesting," says Dan Geschwind, a leading autism researcher at UCLA's School of Medicine. Joseph Buxbaum, another top researcher at Mount Sinai School of Medicine, says, "People are beginning to think about other areas where the mathematics and statistics have really evolved, and then bringing them back to genetics. This is a really neat example." Buxbaum himself is using methods borrowed from astrophysics to research autism.

Hussman's advance, researchers say, consists of the patterns he has found across hundreds of autism cases. Each person is the result of millions of genetic coin flips that occur across the human genome. Hussman's algorithm can detect unusual clusters of flips that may be linked to autism. There's a second benefit too. If you look at each coin flip separately to find the ones causing autism, you miss something because the flips are interconnected by genetic material -- one flip can influence another. Until now, autism researchers haven't been able to fully incorporate that influence into their research. Hussman created a way to do that, using complex statistics he learned from studying markets. "In finance, when you see the same signal across two different markets or countries, you take that as a stronger signal of information than if you only saw that signal in one market," he says. "It's similar in genetics."

In his Maryland office, Hussman is surrounded by eight computer screens, which he uses to monitor markets and do the trading for his three funds, which have a total of $8 billion in assets. Despite being well-known in certain investing circles for his prescient warnings before the 2008 crisis and his pointed critiques of Fed Chairman Ben Bernanke, Hussman generally avoids interviews with the press. He harbors disdain for some popular financial news, once writing "to watch a half hour of CNBC today is like watching an old episode of Gomer Pyle." His long, cerebral weekly reports are usually the only communication he has with investors.

But autism is different. Hussman talks for hours about the subject. He first got interested when his son J.P., now 17, was diagnosed as a two-year-old. At the time, Hussman was teaching economics at the University of Michigan. He visited a parade of doctors and therapists, but he often left disappointed. They suggested medication and dietary changes for J.P. But no one could tell Hussman the causes of autism. In fact, in those days many simply blamed the mother. Hussman was frustrated, so he did what he knew best: He began researching.

Around 2001, he met Margaret Pericak-Vance, a well-regarded geneticist who discovered one of the main genes implicated in Alzheimer's disease. Hussman later began funding her research on autism at the University of Miami with profits from his fund business. They talked about analyzing genetic data in new ways, and Hussman decided to tackle the project.

Not a Kandinsky: Hussman's map of proteins thought to be involved in autism

Genetics are known to play a large factor in autism. Scientists say between 70% and 80% of cases show heritability. But even though studies have indentified 100 or more genes contributing to autism, many genetic variations aren't shared between people. Hussman's statistical prowess has allowed him to zero in on those that may be most commonly linked to the disorder in addition to those rare variations.

When researchers search for genetic links to autism, they don't come across one "autism gene." Instead, a lot of signals pop up. Hussman wanted his computer code to filter out some of those signals to focus on regions giving off many signals. That, he hoped, could lead him to genes that function in the same biological process -- say, the way cells adhere to each other -- and ultimately to a set of genes strongly associated with autism.

Eventually, Hussman analyzed two data sets of hundreds of families with autism. He discovered that a lot of genes associated with autism interact in a pathway in the brain that regulates how neurons sprout to reach their synaptic partners. Scientists had suspected a link, but Hussman's paper provides proof.

The benefits of his research might not be realized for some time. Others need to replicate the results on independent data sets. If the same genes pop up, researchers have to go through the laborious task of sequencing each gene in hopes of finding specific molecular deficits within those stands of DNA. The end goal is to use a list of troublesome genes for therapeutic benefits like early intervention with children or even to predict risk of the disorder, says Mount Sinai's Buxbaum.

Meanwhile, Hussman is becoming a popular figure in the autism community. Recently he was in New York to attend the premiere of Wretches and Jabberers, a documentary about two autistic adults that was funded in part by Hussman's Foundation. After the premiere, at a private reception attended by actress Heather Graham and Autism Society chief Lee Grossman, Hussman was introduced as the project's "puppet master" and cheered by the crowd.

The big question -- what commonly causes autism? -- remains unanswered. That still creates a nightmare for parents of autistic children, who can find new autism studies each week on news sites. Much of them have little scientific support. "The problem with not knowing enough is not only that it doesn't provide enough useful avenues for intervention," says Hussman, "but it opens the door to a lot of snake oil." Just this year, a British medical journal concluded that an influential 1998 study linking childhood vaccines to autism was an "elaborate fraud."

Hussman realizes there's a long way to go: "As you go through as a parent, one of the things that strikes you about autism is that no one knows nearly as much about it as you could possibly imagine wanting," he says. Hussman's son J.P. is now 17 and in high school, and he knows his research will pay off more for other kids in the future than for his own son today. While there are simply many things that can't be known today about autism, he's focused on those that can.

Also from Fortune:

Join the Conversation
Current Issue
  • Give the gift of Fortune
  • Get the Fortune app
  • Subscribe
Powered by WordPress.com VIP.