Background The REGISTRY study is one of the largest longitudinal observational studies in Huntington’s disease. One major assessment point in Registry is the UHDRS cognitive battery, including the Stroop test, the Symbol Digit Modalities Test (SDMT), and the verbal fluency test for letters.

Aims Our aim was to investigate whether the UHDRS cognitive tests were sensitive to longitudinal change in HD patients and differentiated between premanifest and manifest subjects. Further, we were interested in the differential development of premanifest and manifest patients in these tests over time.

Methods Patients were split in two groups by diagnostic confidence level (DCL), resulting in a premanifest group with DCL 1–3 (n = 670) and a manifest group with DCL 4 (n = 1634). Annual change in test scores was analysed using linear mixed models with fixed effects time, DCL at baseline and the interaction term timeXDCL.

Results Mean overall score difference between DCL 1–3 and DCL 4 was significant for all tests, ranging from 14.87 points (p < 0.0001) in the Stroop word reading test to 8.3 points (p < 0.0001) in the verbal fluency test. In the DCL 1–3 group, annual change was significant for Stroop word reading (–0.99 points, p = 0.014) and for the Symbol Digit Modalities Test (–0.55 points, p = 0.008). In the DCL 4 group, annual change was significant for all UHDRS cognitive tests, changes ranging from – 0.56 (p = 0.0001) points in the verbal fluency test to –3.36 points (p < 0.0001) in the Stroop word reading test.

Conclusions All UHDRS cognitive tests in the REGISTRY study were able to differentiate between the premanifest and manifest patient groups. In the premanifest group only Stroop word reading and the SDMT depicted annual cognitive decline, while in the manifest group all tests did.