AI nutrition recommendations improved anxiety, depression

Key takeaways:

• Nutrition recommendations from an AI appeared to improve IBS, depression, anxiety and type 2 diabetes risk.
• The findings support the deeper integration of data-driven precision nutrition, researchers wrote.

Artificial intelligence-powered nutrition recommendations based on individuals’ molecular data were associated with significant improvements in anxiety, depression and more, according to studies by a company developing the technology.

Participants in the studies by Viome Life Sciences who received the recommendations also saw significantly improved clinical outcomes in irritable bowel syndrome and type 2 diabetes, supporting “the integration of data-driven precision nutrition into the standard of care,” Janelle Connell, RDN, and colleagues wrote in the American Journal of Lifestyle Medicine.

The Viome Precision Nutrition Program uses AI to give food and supplement recommendations based on molecular data from an individual’s gut microbiome. The program performs transcriptomic analysis of a stool (and possibly blood) sample, then uses algorithms based partly on clinical studies to give personalized nutrition recommendations.

The researchers, most of whom are employed by Viome, conducted non-conventional studies evaluating the efficacy of the program in IBS (n = 105), depression (n = 410), anxiety (n = 490) and type 2 diabetes. The type 2 diabetes trial compared a group that self-reported low adherence (n = 1,456) with a high-adherence group (n = 1,456), while the other trials each had an intervention arm with no control group.

Participants were U.S. adults invited to complete a “wellness survey” consisting of clinically validated surveys for the four conditions. They filled the survey out again after receiving precision food and supplement recommendations.

Statistically significant survey score improvements were seen in the second survey across mild, moderate and severe levels of IBS, depression and anxiety. At severe disease levels:

• depression, according to PHQ-9 score, decreased a mean 31% (P < .0001);
• anxiety, according to GAD-7 score, decreased a mean 31% (P < .0001); and
• IBS, according to IBS-SSS score, decreased a mean 39% (P < .0001).

High adherence to the recommendations was associated with a lower risk of type 2 diabetes. Compared with the low adherence group, the high adherence group saw a statistically significant improvement (P = 1.99e-05) in a type 2 diabetes risk score, developed in another Viome-funded study, based on stool metatranscriptomic analysis.

Future randomized controlled trials of the precision nutrition program will address the lack of control arms and examine patients over longer time periods, Connell and colleagues wrote.