Deep brain stimulation improved mood and anxiety symptoms in patients with severe anorexia nervosa, says Dr. Blair Woodside. (CBC)

Surgically implanting pacemaker-like devices into the brains of people with severe anorexia might help improve their symptoms, a small Canadian study suggests.

Anorexia affects an estimated 15,000 to 20,000 people in Canada, mainly young women who face a high risk of premature death. The mortality rate is between six to 11 per cent.

About 60 to 70 per cent of people with anorexia recover fully with traditional treatments, said Dr. Blake Woodside, medical director of the eating disorders program at Toronto General Hospital.

But in Wednesday's online issue of the medical journal The Lancet, Woodside and his co-authors describe using deep brain stimulation to treat six women with severe anorexia that did not respond to treatment.

The treatment involves surgery to implant the electrical stimulators. It's considered minimally invasive and the stimulation can be turned off. In the pilot study, the average age of the women at diagnosis was 20 and they ranged in age from 24 to 57 when the surgery was performed. Five had a history of repeated hospital admissions for eating disorders.

While the study was meant to test the safety of the procedure, not its effectiveness, Woodside's team found three of the six patients achieved and maintained a body mass index greater than their historical level.

"Deep brain stimulation in this group of six patients with chronic and treatment-refractory anorexia nervosa was generally safe," the study's authors concluded.

During the surgery, the implanted electrodes are connected with thin wires into the upper chest where they are hooked up to a pacemaker-like device.

One woman, Kim Rollins, had a seizure about two weeks after surgery — a serious adverse event requiring an admission to hospital. She also had obsessive compulsive disorder and post-traumatic stress disorder, Woodside said. The medical team wasn't sure if the seizure resulted from how physically sick she was or the stimulation.

When the stimulation was turned back on at a lower level, Rollins's mood and anxiety symptoms improved. The Kitchener, Ont., resident also responded to traditional hospital treatment, showing a normal weight for seven months.

At age 28, Rollins had a heart attack. She also had two strokes and repeatedly broke the bones in her legs and feet while exercising due to osteoporosis

Now 36, Rollins said she's able to eat and live normally. "I feel absolutely fabulous, I have energy now."

The researchers said MRI shows how the treatment targets a specific area of the brain —a bundle of white matter below the corpus callosum, the thick bundle of nerve fibres dividing the left and right sides of the brain. Doctors have previously aimed deep brain stimulation at the area to treat patients with depression.

The study's senior author, Dr. Andres Lozano, a neurosurgeon in the field of deep brain stimulation, called the findings encouraging because they seemed to point to a genuine therapeutic effect rather than a placebo or hunger-increasing effect.

In a journal commentary accompanying the research, Prof. Janet Treasure and Prof. Ulrike Schmidt of King's College London's Institute of Psychiatry said resistance to treatment in anorexia might be caused in part by starving the brain during key periods of adolescence when decision making, social communication and new learning are impaired by low weight.

It's thought that treatment could be more difficult as anorexia becomes entrenched since brain functions that are needed for successful treatment can also be impaired by low weight, the commentators said.

Much larger, blinded studies are needed to confirm that the approach works and if so, how well.

"Many open questions remain in relation to use of DBS [deep brain stimulation] in anorexia nervosa, including the choice of targets, mechanisms of action, and practical issues such as patient selection and acceptability of the treatment to patients and their families. Nonetheless, the findings of this proof-of-concept study are promising and will give hope to patients with especially pernicious forms of the disorder and their families," Treasure and Schmidt wrote.

Other adverse effects in the study included pain, nausea, panic attacks and an air embolus.

The research was funded the Klarman Family Foundation Grants Program in Eating Disorders Research and the Canadian Institutes of Health Research.

With files from CBC's Pauline Dakin