BMI over time beats genetics in predicting obesity

Although the latest methods of genetic risk profiling can help patients know if they're more susceptible to becoming obese, new research from the Michigan Medicine Frankel Cardiovascular Center and the Massachusetts General Hospital Cardiovascular Research Center suggests focusing on body mass index (BMI), a measure of weight and height, may be more beneficial.

The study, published in JAMA Cardiology, found that a person's BMI measurement from 25 years ago was a better predictor of their current BMI than a polygenic risk score.

Researchers evaluated health data from a National Institutes of Health-sponsored study of more than 2,500 young adults from across the United States, who 25 years ago volunteered to participate in a longitudinal study. The data in Coronary Artery Risk Development in Young Adults (CARDIA) was collected between 1985 and 2010 to explore the development of cardiovascular disease.

The research team used a modern polygenic risk score, a composite measure of genetic risk of obesity, to calculate genetic risk of obesity for each person in their subset of the CARDIA study and compare it to the measurements taken during the 25 years of the study.

Baseline BMI in young adulthood explained 52.3 percent of a person's BMI 25 years later when it was considered in combination with age, sex and history of a parent ever being very overweight. The prediction could explain up to around 80 percent of BMI variation after following someone's BMI over time, rather than just at baseline and 25 years later.

Those same combinations of age, sex, and parental weight history, when considered with a polygenic risk score instead of BMI, were also associated with BMI but in a weaker association that only explained 13.6 percent of BMI in midlife.

The PRS was also more effective at predicting future BMI in the 1,608 white individuals than the 909 black individuals. The researchers noted there's more genetic data available in European populations for constructing genetic risk profiles, leading to some concern about methodology for determining polygenic risk scores for non-white patients.

Venkatesh Murthy, MD, PhD, lead author and a cardiologist at the Michigan Medicine Frankel Cardiovascular Center, said the data serve as a reminder that human genetics might be interesting in large population studies, but that caution is still needed toward incorporating them when providing clinical care and advice to patients.

However, he acknowledges clinicians are seeing more patients who have already purchased a genetic report from a direct-to-consumer company and want to go over it with their doctor. It's important for clinicians to be aware of the strengths and limitations of those direct-to-consumer products, Murthy said.

Murthy said the rising interest in genetic risk scores also brings up the idea of how incorporating them into clinical practice could change behavior. If someone is told they were born more likely to become obese, for example, how will that change their behavior today or this year, or 25 years down the road? Conversely, will people who learn they're less disposed to obesity become more motivated to lose that stubborn weight that's been difficult to shed?

"We don't know those answers very well yet," Murthy said in a statement. "However, some data says, whether based on a real genetic score or not, people may perform better in fitness tests if they're told they're genetically more likely to be fit. And we run a risk of demotivating some people if we tell them the genetics are against them, even though we now know other associations with BMI are stronger than genetics."

The good news, Murthy said, is calculating your BMI is significantly more affordable than purchasing a genetic test. Physicians should already have weight and height records for their patients over time, Murthy said, and the conversations around modifiable risk factors relating to BMI should already be happening during patient visits.