Nowadays, measuring cognition is still unsatisfactory in longitudinal follow-up in HD. In the framework of the REPAIRHD project, Creteil’s centre coordinates the elaboration of an assessment battery of reference for surgical trial and was committed to the development of cognitive tests. We developed a digitalized battery of 7 tasks that were assessed in French, English and German during a one-year follow-up, in comparison with well-established tasks.Aims
To build the most efficient composite measure combining the best cognitive tasks (old and/or new).Methods
The analysed data set includes 36 participants (16 controls, 20 HD manifest TFC stage I or II) assessed at baseline, month 1 and month 12 in Creteil.Methods
We first computed for each cognitive test Signal to Noise Ratio (S/N ratio) as defined by the mean change from baseline to month 12 divided by the corresponding standard deviation.Methods
We then selected the tests with the highest S/N ratio in patients, different from S/N ratio in controls, to perform a principal component analysis (PCA) to define a composite variable as a linear combination of individual variables.Results
The PCA analysis identified Verbal fluency, Stroop interference, Subtraction, eCog-Executive, eCog-Language that accounted for most of the variability (60%) in the data set. They were used to build a composite score which showed a S/N Ratio higher than any individual measures.Conclusion
The longitudinal follow-up carried out in Creteil suggested that mixing eCog assessments and old tasks in a composite measure offers an optimal cognitive measure to be used in clinical trials, specifically with small cohorts. These results will be further validated when the one-year assessment will be completed in all REPAIRHD centers.