(075) The Use of Potentially Interacting Supplement–Drug Pairs in Adults with Type 2 Diabetes: A Large Population-Based Cohort Study in the UK Biobank

Background: Dietary supplements are widely used among adults with type 2 diabetes. However, there has been substantial concerns about the potential interactions between these supplements and conventional drugs.

Objectives: We aimed to examine the patterns of use of potentially interacting supplement–drug pairs in adults with type 2 diabetes in real-world settings and to explore the effect of potentially interacting supplement–drug pairs on downstream health outcomes.

Methods: Participants with type 2 diabetes in the UK Biobank were included. Potentially interacting supplement–drug pairs were identified from four tertiary databases and were classified according to potential pharmacokinetic or pharmacodynamic interactions. We categorized the potential pharmacodynamic interactions into different clinical types according to their related outcomes and explored their associations with incident outcomes using Cox proportional hazards models.

Results: This analysis included 26,394 participants (mean age 59.9 ± 7.0 years, 37.6% women), of whom 48.5% used supplements. Most (85.0%) of the supplement users were taking potentially interacting supplement–drug pairs. More supplement users were taking supplement–drug pairs with potential PD interactions (n=10,360, 81.0%) than potential PK interactions (n=5,386, 42.1%). The most common classes of interacting drugs were lipid-modifying agents (n=9,001, 70.4%), glucose-lowering drugs (n=7,757, 60.6%) and antithrombotic agents (n=6,796, 53.1%). The potential pharmacodynamic interactions were related to clinical outcomes, which includes reducing the effects of glucose-lowering drugs (50.7%), hypotension (49.8%), bleeding (50.4%) and hepatotoxicity (34.8%). Exploratory analyses found that the use of potentially interacting supplement–drug pairs was associated with new-onset hepatic diseases (hazard ratio=1.26, 95% confidence interval 1.10–1.44, P< 0.001), but not with hypotension (P=0.409) or bleeding (P=0.438).

Conclusions: Real-world data suggests that most adults with type 2 diabetes who concurrently used supplements and drugs were on potentially interacting supplement–drug combinations, with the potential of causing various types of adverse outcomes. Our findings reinforce the importance of communicating with patients about their supplement use and to assess the potential risk of supplement–drug interactions in clinical settings. From a regulatory perspective, manufacturers of supplement products should conduct post-marketing surveillance studies to monitor the supplement products throughout their life cycle.