Key Takeaways
Extended Takeaways
- Demographic transparency remains a major limitation in the antidepressant trial literature: among 166 included US trials, sex was reported in 90.36%, but race in only 39.16% and ethnicity in 33.13%, which constrains subgroup interpretation even before representation is assessed.
- Using recruitment-region census data rather than national averages changes the benchmark for representativeness and may be more clinically relevant for US trial appraisal, especially when interpreting studies conducted in states with distinct racial and ethnic compositions.
- Not all racial groups were underrepresented to the same degree; White (73.10% vs 77.52%, P=.443), Black (20.21% vs 13.51%, P =.106), and NHPI (0.21% vs 0.16%, P=.069) enrollment did not differ significantly from state census benchmarks, so concerns about generalizability are driven by specific gaps rather than uniform underenrollment across all minority groups.
- Underrepresentation was concentrated in Asian, AIAN, multiracial, and Hispanic or Latino participants, with significant gaps for Asian (4.13% vs 5.67%, P<.001), AIAN (0.44% vs 0.89%, P<.001), multiracial participants (1.90% vs 2.26%, P=.031), and Hispanic or Latino participants (12.42% vs 17.20%, P <.001).
- The weighted average age was 46.04 years (CI: 44.67–47.60), so many trial samples may not reflect older adults who commonly receive antidepressants and may have different pharmacokinetic and tolerability profiles.