Glucagon-Like Peptide-1 Based Therapies and the Risk of Severe Gastrointestinal Motility Adverse Events: A Cohort Study.

AIMS: Evidence showing the association between glucagon-like 1 (GLP-1) based therapies and severe gastrointestinal motility events is limited and possible treatment effect heterogeneity by demographic and clinical characteristics is unknown. We aimed to compare rates of severe motility-related gastrointestinal adverse events between initiators of GLP-1 based therapies versus sodium-glucose co-transporter 2 inhibitors (SGLT-2i).

MATERIALS AND

METHODS

This new-user active-comparative cohort study used two de-identified US commercial healthcare databases to identify adults with type-2 diabetes without a prior history of major gastrointestinal conditions who initiated GLP-1 based therapies (GLP-1 receptor agonists or tirzepatide) or SGLT-2i. The main outcome was a composite of severe constipation, gastroparesis, and gastrointestinal obstruction. We estimated incidence rates, hazard ratios (HRs), and 95% confidence intervals (CIs) after propensity score matching.

RESULTS

Among 313 342 matched pairs of GLP-1 based therapies and SGLT-2i initiators (mean age 60 years; 45% female), followed on treatment for a median of 5.2 months, incidence rates of the composite outcome were 1.02 among GLP-1 based therapies versus 0.75 among SGLT-2i initiators per 100 person-years. The risk was higher among GLP-1 based therapies versus SGLT-2i initiators (rate difference 0.27 per 100 person-years; HR 1.37, 95% CI 1.30, 1.45) for the composite event and each individual component. This relative increased risk was consistent across subgroups stratified by GLP-1 based therapies agent, age, sex, BMI, frailty level, diabetes severity, or opioid use.

CONCLUSIONS

The absolute risk of motility-related gastrointestinal adverse events was 1% or less but was higher among initiators of GLP-1 based therapies compared to SGLT-2i. These findings can inform the risk benefit assessment by clinicians before initiating these treatments.