Symbol Digit Modalities Test: Regression-Based Normative Data and Clinical Utility

Abstract

Abstract Objective The purpose of this study was to provide regression-based normative data for the written, oral, and incidental recall trials of the Symbol Digit Modalities Test (SDMT). Method Regression-based normative equations for the written and oral trials were derived from 536 healthy men and women between the ages of 18 and 91. Normative equations for the incidental recall trial are provided for a subset of the normative sample (age range = 60–91). The clinical utility of the newly developed norms was examined by comparing mean performance and rates of impaired scores for participants with traumatic brain injury (TBI), mild cognitive impairment (MCI), and dementia. Within-group analyses were used to compare the new norms to the original published norms. Results Age, education, and sex were all significant predictors of written trial performance, age and education were significant predictors of oral trial performance, and only age predicted incidental recall trial performance. As expected, the TBI group demonstrated the highest rates of impaired performance on both written and oral trials. Participants with dementia showed the highest rate of impaired scores on the incidental recall trial, followed by participants with amnestic MCI. Compared to traditional norming methods, the regression-based norms classified more clinical participants as impaired on both the written and oral trials. Conclusions Comprehensive regression-based normative equations with demonstrated clinical utility are provided to improve the detection of cerebral dysfunction using the SDMT. A calculator with the normative equations is provided so that raw scores can be easily converted to demographically-corrected standardized scores.