A-197 Trail Making Test Normative Data from the National Alzheimer’s Coordinating Center: A Comparison of Regression and Non-Regression Methods

Abstract

Abstract Objective Various methods are used to derive normative data for neuropsychological assessment measures. The National Alzheimer’s Coordinating Center (NACC) regression norms (2011) are frequently used within Alzheimer’s Disease Research Centers, yet little research exists comparing these norms to similar regression/non-regression-based norms. Normative data for the widely-used Trail Making Test (TMTA&B) was compared between two regression-based normative data methods (Mitrushina and NACC) and traditional normative methods (i.e., Mayo’s Older Americans Normative Studies [MOANS]). Method TMTA&B data were obtained from 14,873 cognitively normal NACC participants (ages: 50–89; education: M = 15.81, SD = 2.90). T-scores for TMTA&B scores based on Mitrushina, NACC, and MOANS norms were computed. T-scores for regression norms were bounded from 20 to 80. Cohen’s d effect sizes were used to compare differences in scores. Results MOANS and Mitrushina norms produced similar normative scores in the overall sample (TMTA T-score difference = 1.06, d = 0.112, TMTB T-score difference = 1.4, d = 0.144). NACC-derived scores were ~ 0.5 SD below MOANS (TMTA T-score difference = −5.3, d = 0.553, TMTB T-score difference = −4.1, d = 0.427) and Mitrushina scores (TMTA T-score difference = −6.3, d = 0.736, TMTB T-score difference = −5.6, d = 0.615). NACC data produced the lowest average TMTA&B T-scores across all 5-year age band groups. Conclusions Significantly lower TMT scores were obtained using NACC’s regression-based norms compared to both MOANS norms and Mitrushina regression norms. These results underscore that the choice of norms may under and/or overestimate cognitive functioning, and comparison norms must be chosen carefully when making decisions about patients’ cognitive status. Future research applying NACC’s regression-based norms to other commonly used neuropsychological measures is warranted.