When Screening Positive for Diabetes-Specific Disordered Eating Behaviors Depends on Who You Are: Unequal mSCOFF Flags by Race, Sex, and HbA1c

OBJECTIVE Given the need for equitable screening tools for disordered eating behaviors (DEBs) in youth with type 1 diabetes (T1D), the current study sought to 1) assess the factor structure and model fit of the modified SCOFF (mSCOFF) screener among Black and White youth with T1D and 2) assess systematic item bias in the mSCOFF as a function of HbA1c, race, and sex assigned at birth. RESEARCH DESIGN AND METHODS A total of 130 youth (ages 10–18 years; 54% female; 57% Black; mean [SD] HbA1c 10.89% [2.25%]) with T1D completed the mSCOFF screener for DEBs. Patient sex assigned at birth, race, HbA1c levels, and screener responses were obtained from medical records. Confirmatory factor analysis and differential item functioning (DIF) analyses were used to examine item response bias across patient demographics and HbA1c. RESULTS Analyses confirmed nonuniform DIF across most mSCOFF items, resulting in biased screening when established cutoffs were used and suggesting that items do not consistently measure DEBs across intersections of race, sex assigned at birth, and HbA1c. CONCLUSIONS These findings underscore the importance of using equitable, unbiased screeners to accurately identify youth with T1D who endorse DEBs. To support accurate risk assessment and early intervention, providers should interpret mSCOFF scores with caution, particularly in racially diverse populations.