For a disease that robs boys of mobility before adolescence, having a reliable blood-based readout of disease progression could transform how clinicians track treatment response and adjust therapy. Duchenne muscular dystrophy currently lacks validated monitoring biomarkers, forcing reliance on functional motor assessments that are burdensome, age-sensitive, and variable — a gap this proteomic investigation begins to address.
Drawing on longitudinal serum samples from participants in the FOR-DMD randomized trial — which compared daily versus intermittent corticosteroid regimens in boys aged 4–8 at enrollment — investigators applied the SomaScan aptamer platform to profile approximately 1,500 circulating proteins over the trial's duration. Using linear mixed models, the team dissected two distinct analytic dimensions: whether boys with higher baseline protein levels tended to perform better on motor benchmarks (an across-patients signal), and whether within-individual protein fluctuations tracked that same boy's functional trajectory over time. Three standard motor assessments served as outcome anchors — Rise from Floor Velocity, 10-Meter Run/Walk Velocity, and the North Star Ambulatory Assessment. A penalized lasso mixed-model approach then tested whether the identified protein signatures improved predictive accuracy beyond a simpler age-and-treatment model, with performance evaluated by optimism-corrected root mean squared error.
This study contributes a methodologically rigorous dual-level analysis that most biomarker discovery efforts omit — separating population-level correlation from clinically relevant within-patient sensitivity to change. That distinction matters enormously: a protein that distinguishes severe from mild cases but fails to move as an individual patient improves or declines has limited trial utility. The cohort, while well-characterized, reflects a narrow pediatric age band and a single therapeutic context, limiting generalizability. Validation in independent DMD cohorts, including gene-therapy trial participants, will be essential before these protein signatures can function as surrogate endpoints. Incremental but methodologically meaningful, this work advances the infrastructure needed for more efficient DMD drug development.