ABSTRACT

Objective:
Coffee is one of the most widely consumed beverages globally and has been linked to favorable health outcomes. However, its system-wide relationships with human biology and the underlying mechanisms remain poorly characterized. This study aimed to investigate the relationship between coffee consumption and continuous glucose monitoring (CGM) metrics and other biological systems in healthy adults.

Research design and methods:
In the Human Phenotype Project, 8666 generally healthy Israeli adults provided two weeks of real-time dietary logs, from which coffee intake was estimated. Participants wore CGM devices throughout this period, and multimodal data spanning 11 additional systems (e.g., gut microbiome, serum lipidomics, and body composition) were collected. We employed machine learning approaches to quantify the extent to which each system reflected coffee intake. We performed linear regression to identify individual traits associated with coffee intake, with false discovery rates < 0.05 considered significant.

Results:
This cross-sectional study identified continuously-monitored glucose regulation and gut microbial composition as the most reflective systems of coffee intake, with further analyses revealing favorable glycemic profiles spanning diverse aspects of glucose regulation with increasing coffee intake, and Clostridium phoceensis (i.e., Lawsonibacter asaccharolyticus) as the most significant species positively associated with coffee intake. Additionally, coffee intake was favorably associated with traits across body composition, serum lipidomics, and hepatic, hematopoietic, and renal systems.

Conclusions:
This study found that habitual coffee intake was linked to multifaceted favorable glucose control captured by CGM and favorable profiles across multiple biological systems, providing mechanistic insights that may guide precision nutrition strategies for diabetes prevention.

‍