TY - JOUR
T1 - The clinical picture of psychosis in manifest Huntington's disease
T2 - A comprehensive analysis of the enroll-HD Database
AU - Rocha, Natalia P.
AU - Mwangi, Benson
AU - Candano, Carlos A.Gutierrez
AU - Sampaio, Cristina
AU - Stimming, Erin Furr
AU - Teixeira, Antonio L.
N1 - Publisher Copyright:
Copyright © 2018 Rocha, Mwangi, Gutierrez Candano, Sampaio, Furr Stimming and Teixeira.
PY - 2018/11/6
Y1 - 2018/11/6
N2 - Background: Psychotic symptoms have been under-investigated in Huntington's disease (HD) and research is needed in order to elucidate the characteristics linked to the unique phenotype of HD patients presenting with psychosis. Objective: To evaluate the frequency and factors associated with psychosis in HD. Methods: Cross-sectional study including manifest individuals with HD from the Enroll-HD database. Both conventional statistical analysis (Stepwise Binary Logistic Regression) and five machine learning algorithms [Least Absolute Shrinkage and Selection Operator (LASSO); Elastic Net; Support Vector Machines (SVM); Random Forest; and class-weighted SVM] were used to describe factors associated with psychosis in manifest HD patients. Results: Approximately 11% of patients with HD presented history of psychosis. Logistic regression analysis indicated that younger age at HD clinical diagnosis, lower number of CAG repeats, history of [alcohol use disorders, depression, violent/aggressive behavior and perseverative/obsessive behavior], lower total functional capacity score, and longer time to complete trail making test-B were associated with psychosis. All machine learning algorithms were significant (chi-square p < 0.05) and capable of distinguishing individual HD patients with history of psychosis from those without a history of psychosis with prediction accuracy around 71-73%. The most relevant variables were similar to those found in the conventional analyses. Conclusions: Psychiatric and behavioral symptoms as well as poorer cognitive performance were related to psychosis in HD. In addition, psychosis was associated with lower number of CAG repeats and younger age at clinical diagnosis of HD, suggesting that these patients may represent a unique phenotype in the HD spectrum.
AB - Background: Psychotic symptoms have been under-investigated in Huntington's disease (HD) and research is needed in order to elucidate the characteristics linked to the unique phenotype of HD patients presenting with psychosis. Objective: To evaluate the frequency and factors associated with psychosis in HD. Methods: Cross-sectional study including manifest individuals with HD from the Enroll-HD database. Both conventional statistical analysis (Stepwise Binary Logistic Regression) and five machine learning algorithms [Least Absolute Shrinkage and Selection Operator (LASSO); Elastic Net; Support Vector Machines (SVM); Random Forest; and class-weighted SVM] were used to describe factors associated with psychosis in manifest HD patients. Results: Approximately 11% of patients with HD presented history of psychosis. Logistic regression analysis indicated that younger age at HD clinical diagnosis, lower number of CAG repeats, history of [alcohol use disorders, depression, violent/aggressive behavior and perseverative/obsessive behavior], lower total functional capacity score, and longer time to complete trail making test-B were associated with psychosis. All machine learning algorithms were significant (chi-square p < 0.05) and capable of distinguishing individual HD patients with history of psychosis from those without a history of psychosis with prediction accuracy around 71-73%. The most relevant variables were similar to those found in the conventional analyses. Conclusions: Psychiatric and behavioral symptoms as well as poorer cognitive performance were related to psychosis in HD. In addition, psychosis was associated with lower number of CAG repeats and younger age at clinical diagnosis of HD, suggesting that these patients may represent a unique phenotype in the HD spectrum.
KW - Behavior
KW - Enroll-HD
KW - Huntington's disease
KW - Machine learning
KW - Psychosis
UR - http://www.scopus.com/inward/record.url?scp=85056256922&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85056256922&partnerID=8YFLogxK
U2 - 10.3389/fneur.2018.00930
DO - 10.3389/fneur.2018.00930
M3 - Article
AN - SCOPUS:85056256922
SN - 1664-2295
VL - 9
JO - Frontiers in Neurology
JF - Frontiers in Neurology
IS - NOV
M1 - 930
ER -