Real-world EHR analysis shows tirzepatide yields 14.7% mean weight loss versus 10.8% for semaglutide over two years

The PNAS Nexus observational study defined response strata by maximum percentage weight loss and found tirzepatide nearly doubled the proportion of high responders (>15% loss) while recording lower rates of gastrointestinal, headache, and fatigue events. Demographic gradients persisted across both agents, with female and White patients achieving greater loss and Black and Hispanic patients over-represented in the <5% category. These patterns were not explained by dosing or adherence variables captured in the records.

Prior randomized trials (SURMOUNT-1 and STEP-1) reported larger absolute differences under protocol-specified titration and lifestyle intervention; the real-world attenuation suggests variable persistence and titration speed outside trials. The absence of randomization leaves residual confounding by indication and unmeasured socioeconomic factors that may influence both drug selection and weight-loss capacity.

Future head-to-head RCTs with standardized dosing algorithms and longer follow-up are required to isolate pharmacologic effect from prescribing bias. Payer-mandated step therapy and formulary placement decisions should await such data rather than extrapolating from this matched but non-randomized comparison.