Canadians have never had more access to information about healthy eating, nutrition and the benefits of exercise. Yet for a significant portion of the population, staying physically healthy is an ongoing challenge.
According to the Canadian Health Measures survey released by Statistics Canada in 2018, 60 per cent of adults in Canada are either overweight or obese. And according to the non-profit National Initiative for Eating Disorders, approximately one million Canadians meet the diagnostic criteria for anorexia nervosa and other eating disorders, mental illnesses that can cause life-long mental and physical health struggles.
In an era dominated by social media, where unrealistic body images and fad diets abound, maintaining a healthy body weight has become even more challenging. Meanwhile, our population is aging, putting a further strain on our health care system.
Improved health outcomes need next-generation thinking. Researchers at Concordia University in Montreal are looking at health through a fresh lens, tackling some of the country’s most pressing health concerns using new concepts and tools. And while some of this work takes place in the lab, it has the potential to affect real world diagnoses and treatments in the near future.
Here is a look at some of the cutting-edge health research happening at Concordia, and why it matters.
Could obesity be considered an age-related condition?
People with obesity live, on average, about seven fewer years than those who are normal weight.
Sylvia Santosa, Canada research chair in clinical nutrition at Concordia University in Montreal, is proposing that the health community frame obesity as an age-related disorder.
“It’s a new way of looking at it,” says Santosa, who is also associate professor at the department of Health, Kinesiology and Applied Physiology at Concordia.
She notes that obesity effectively ramps up the aging process. For example, children with obesity are at greater risk of developing type 2 diabetes and hypertension than children who are normal weight. Historically, these adverse health conditions were only seen in adults and were very rare in children. Even in adults, having obesity combined with the passage of time puts people at higher risk of developing these diseases and others, including cancer.
Obesity has been diagnosed as a chronic disease by leading health organizations, according to Canadian charitable organization Obesity Canada. But Santosa’s research could change how doctors diagnose and treat overweight and obese patients over their lifetimes.
Santosa’s bigger-picture approach to obesity has been bolstered by her other research, which includes the use of magnetic resonance imaging (MRI) to analyze effects of fat deposits in people over time.
“Fat doesn’t behave the same around your organs and under your skin, in your upper body and your lower body,” says Santosa. Belly fat, for instance, is associated with higher health risks, and she’s looking to understand why from a physiological basis. Currently, Santosa is using imaging to look at fat around muscles in the legs.
“We want to see how fat and muscle communicate,” she says.
Understanding obesity is a key research priority across the country. The Government of Canada invests about $35-million annually on obesity and healthy body weights research, according to Health Canada. The research is used to help prevent obesity and also deal with complications, such as diabetes.
With a focus on the science of obesity, Santosa hopes to influence the way obesity is treated and how doctors assess disease risk as it relates to body fat and obesity. She says her lab’s work can contribute to future treatments by approaching obesity at a more sophisticated level.
“Right now, if someone has obesity, the only thing recommended is weight loss. But maybe some weight loss diet or approaches are more effective in some cases versus others,” Santosa says. “We can only achieve more individualized and effective treatments for obesity if we understand how the disease manifests differently in different people.”
Clues to eating disorder risk from social media
While the physiological roots of body weight issues like obesity are undeniable, mental health can play an equally significant role affecting physical well-being. For example, eating disorders like anorexia nervosa and bulimia can be complex, difficult to treat and sometimes deadly.
According to the National Eating Disorders Information Centre, they affect about 1.5 per cent of women aged 15-24 in Canada. (Eating disorders also impact men, but in lower numbers.) Shockingly, anorexia nervosa has the highest mortality rate of any psychiatric illness, with an estimated 10 per cent of people who develop it dying within 10 years of the onset of the disease.
But what if the early warning signs of eating disorders could be detected on social media?
To help health care professionals intervene early, Concordia professor Leila Kosseim has developed an artificial intelligence tool to detect anorexia in posts on social media platforms like Twitter, Facebook, Snapchat and more.
Kosseim, who is at the Gina Cody School of Engineering and Computer Science at Concordia, participated in an international project called eRisk. The eRisk project was organized by researchers from three European universities, who provided a data set collected from forum website Reddit, and set the task of identifying those at risk for anorexia. The Concordia team, comprised of Kosseim and two graduate students, developed a deep learning algorithm that looked at linguistic features of the posts, through a process called natural language processing.
“You’re looking for clues in how people talk,” Kosseim says.
The goal of the eRisk competition was to get to a conclusion quickly, and the Concordia team was able to do it faster than the other teams. “If you wait until 100 posts, it may be too late for real-life scenarios,” Kosseim explains.
This algorithm was so successful that Kosseim, her team and a researcher from a partnering university are collaborating with a hospital in Geneva, Switzerland, to look at data from real patients there. The deep learning algorithm will be adjusted to look for suicide risk among a group of patients using social messaging platform WhatsApp (patients will consent to this use).
“It can be used by doctors to just give them a warning,” Kosseim says.
Kosseim notes that while the AI can figure out which aspects of language lead to a mental health risk worthy of medical attention, the algorithm doesn’t tell researchers what those red flags are. “I can’t give you a list of words or patterns,” she says.
To change that, one of the graduate students working with Kosseim is looking to reverse-engineer the algorithm and uncover that information, which could then be used for more research or even as a training tool for healthcare professionals.
“People talk about the misuse of AI, but [we would like to show] it can be used for social good,” Kosseim says.
To learn more about the Health research being done by Canada’s leading researchers, visit Concordia’s Health Hub.
Advertising feature produced by Globe Content Studio. The Globe’s editorial department was not involved.