The downstream treatment burden of commonly used glucose-lowering therapies in type 2 diabetes mellitus (T2DM) remains incompletely characterized in routine clinical practice. A retrospective propensity score–matched cohort study published in Diabetes, Obesity and Metabolism compared downstream pharmacologic and clinical outcomes following initiation of glucagon-like peptide-1 (GLP-1) receptor agonists versus sodium-glucose cotransporter 2 (SGLT2) inhibitors.
The analysis used a multicentre electronic health record network and included adults with T2DM who initiated either therapy between January 1, 2017, and June 1, 2025. After 1:1 matching, approximately 164,000 individuals were included in each group. Outcome-specific cohorts with standardized washout periods yielded sample sizes ranging from approximately 124,000 to 164,000 per outcome. Primary outcomes assessed initiation of seven downstream medication classes, while secondary outcomes included gastrointestinal and nutritional diagnoses and health-care utilization.
Across all seven medication classes, downstream pharmacotherapy initiation occurred more frequently among GLP-1 initiators. The largest differences were observed for symptom-driven medications, including antidepressants (absolute risk difference 3.38%; 95% CI −3.61 to −3.15; HR 0.78, 95% CI 0.76-0.80), antiemetics, and sedatives or hypnotics. Proton pump inhibitors, laxatives, histamine-2 receptor antagonists, and antidiarrheal agents were also initiated more frequently, although the differences were smaller. Gastroesophageal reflux disease and esophagitis were more frequent following GLP-1 receptor agonist initiation, whereas peptic ulcer disease and endoscopic procedures occurred more often with SGLT2 inhibitors. Nutrient deficiency and unspecified anemia were also more frequent with GLP-1 therapy. Overall health-care utilization was similar between groups, except for higher inpatient acute care use among SGLT2 inhibitor initiators.
These findings indicate that GLP-1 receptor agonist initiation is associated with greater downstream use of symptom-driven medications without corresponding increases in overall health-care utilization. Medication initiation patterns may provide a sensitive measure of treatment burden in routine diabetes care.
