The wellness industry has produced no shortage of platforms and protocols, but the long-term, peer-reviewed evidence has been slower to arrive.
InsideTracker just made a sizable contribution. The platform just put 20,000 users and four-plus years of data on the table, publishing a peer-reviewed study in PLOS Digital Health.
The health analytics company, which provides personalized recommendations based on blood analysis, DNA, lifestyle and nutrition habits, found that using the AI-powered platform was associated with sustained improvements across 39 blood biomarkers, including LDL cholesterol, HbA1c, glucose, vitamin D, hsCRP and testosterone.
Renée Deehan, PhD, senior vice president of science and AI at InsideTracker and one of the study’s authors, said the work was a logical next step for the company, especially as the platform’s preventive health lifestyle recommendations are all based on existing clinical studies.
“It’s really important to us that we also practice what we preach and also contribute to the scientific literature,” she said.“Credibility is a big part of it, and also just a lot of curiosity. What secrets are in there that you have to tell us? Is it what we would expect? Is it what we wouldn’t expect?”
The headline finding was that among users who started with suboptimal blood biomarker levels, the majority improved by their second blood draw, and many continued improving across subsequent tests. Initial gains in fasting glucose, cortisol, HbA1c, hsCRP and folate held over time. Triglycerides, vitamins B12 and D, iron, testosterone in males, and LDL and HDL cholesterol continued to improve after the second test.
A self-described classic cynical scientist, Deehan said the findings caught her off guard. Biomarker profiles tend to worsen with age, she noted, and she expected the data to reflect that, save for a handful of biomarkers.
“Surprise number one was actually that the vast majority of biomarkers improved over time,” she said. “If you look at a group level, I did not expect that at all.”
Deehan also pointed to an interesting finding for anyone trying to figure out which wearable metric to prioritize.
“If you need to focus on one number, on one thing, go make it VO2 max,” Deehan said. “Because probably if you improve that — and this was not a prospective study, we can’t say any of it’s causal, and we won’t — but VO2 max was correlated retrospectively with all these different biomarkers. So the hypothesis is that if you improve your VO2 max, that’s going to tug a lot of biomarkers along with it.”
A second wearable finding came from a smaller subgroup of users with high baseline total cholesterol. Those who improved their cholesterol levels increased their daily step count by an average of about 950 steps and showed higher REM sleep percentages than non-improvers. Deehan called the REM sleep result “the weirdest” the team saw.
“People who were able to improve their total cholesterol were getting better REM over the course of a year,” she said.“We didn’t see the same result with total sleep or deep sleep. It was specifically REM sleep, which was a little bit of a puzzle.”
The authors flag this finding as exploratory given the smaller sample size, but it reinforces the role of activity and recovery in cardiovascular health.
A few other threads from the study and Deehan’s insights are also notable, particularly at a time when supplements are booming and people are investing in their health in pursuit of some extra years.
For one, supplements may move the needle fastest, but not always for the most important biomarkers. Vitamin D, B12 and folate were among the biomarkers users most frequently improved, in part because supplementation is an easier behavior to adopt than dietary or exercise change, Deehan said. On the other hand, LDL cholesterol, HbA1c and other cardiometabolic markers were harder to move.
“LDL is one of the harder biomarkers to optimize,” Deehan said. “There isn’t a magic bullet supplement that’s going to pull your LDL down the way that a vitamin D supplement can really pull your vitamin D levels up.”
Genetics also plays a key role, a useful reminder for anyone who has been doing the work and wondering why their numbers won’t budge. Users with a higher polygenic risk score for high LDL or total cholesterol saw smaller improvements than users with lower genetic risk.
“Some people just cannot ‘lifestyle’ their way out of high cholesterol,” Deehan said. “They’re bumping up against a kind of genetic wall.”
A particularly relevant finding speaks to the importance of engagement, the foundation of any lasting behavior change. InsideTracker users who created an Action Plan, which includes personalized recommendations and in-app reminders, showed greater biomarker improvements at the first follow-up than users who did not create one.
Ultimately, committing to change is one part of the equation. Giving users the supportive tools and informed nudges to keep following through is the other, all of which adds up over time.
“By prioritizing and individualizing actionable health insights, it’s possible to help people make the kind of small, consistent lifestyle changes that lead to real, measurable improvements in their health,” Deehan said. “The fact that we saw these changes sustained over a period of years is particularly encouraging.”
