The Shape Trail Test (STT) was developed based upon the Trail Making Test, as a culture-neutral test for measuring processing speed and mental flexibility. This study aims to evaluate the accuracy and validity of this five-minute test for differentiating individuals with normal cognition (NC), subjective memory impairment (SMI), and mild cognitive impairment (MCI).

The study included 210 participants aged 50–80 years, with 70 participants in each group matched for age, education, and gender.

No significant difference in STT measures was found between the NC and SMI groups. In contrast, both the NC and SMI groups exhibited significantly better performance (shorter completion time in STT-A and STT-B and fewer STT-B errors) than the MCI group. No significant group differences were found in STT-A errors. Stepwise regression analysis identified three significant predictors for classifying the MCI group from the NC and/or SMI groups, including the STT-B completion time, the STT-A errors, and the interaction between STT-B completion time and STT-B errors. The composite score of these three predictors demonstrated good discriminatory power for classifying the MCI group from the other groups, with area under the curves (AUCs) of 0.76–0.79 (p < 0.001), sensitivities of 78.6%–80%, and specificities of 60%–61.4%. However, none of the STT measures or their interactions were significant predictors for differentiating the SMI group from the NC group. Besides, the STT measures were significantly correlated with age, education, and executive function measures.

The STT could be a culture- and language-free, reliable test for assessing executive function and a sensitive test for predicting MCI.