Modifiable factors associated with Huntington's disease progression in presymptomatic participants

Abstract

Objective: Huntington's disease (HD) is a neurodegenerative disorder characterized by progressive motor, cognitive, and psychiatric symptoms. Our aim here was to identify factors that can be modified to slow disease progression even before the first symptoms appear.

Methods: We included 2636 presymptomatic individuals (comparison with family controls) drawn from the prospective observational cohort Enroll-HD, with more than 35 CAG repeats and at least two assessments of disease progression measured with the composite Huntington's disease rating Scale (cUHDRS). The association between sociodemographic factors, health behaviors, health history, and cUHDRS trajectory was assessed with a mixed-effects random forest using partial dependence plots and Shapley additive explanation method.

Results: Participants were followed by an average of 3.4 (SD = 1.97) years. We confirmed the negative impact of age and a high number of CAG repeats. We found that a high level of education, a body mass index (BMI) <23 kg/m² before the age of 40 and >23 kg/m² thereafter, alcohol consumption of <15 units per week, current coffee consumption and no smoking were linked to slow disease progression, as did no previous exposure to antidepressants or anxiolytic, no psychiatric history or comorbidities, and being female. Other comorbidities or marital status showed no major association with HD evolution.

Interpretation: Reducing modifiable risk factors for HD is one way to support the presymptomatic population. A high level of education, low-to-moderate alcohol consumption, no smoking, and BMI control are likely to slow disease progression in this population.

Figure 1

Analysis of population selection flowchart. The extended flowchart in Figure S1.

Figure 2

Summary plot of the SHAP values for each prediction organized by level of relevance of each variable. The left‐hand panel shows the relative importance of variables, expressed as mean (|shap| values) in the fully adjusted mixed‐effects random forest model. The right‐hand panel shows the SHAP values for each of the individuals, represented by single dots, per variable and as a function of variable values. The color of each dot represents the value of each variable. The redder the dot, the higher the value of the variable and the bluer, the lower the value. Variable ranges are age [18–80], number of CAG repeats [36–55], education level [0–6 ISCED levels], BMI [18–36], pharmacological depression and anxiety treatment [0 = No‐1 = Yes], units of alcohol per week [0–30], tobacco and drugs consumption [0 = No‐1 = Yes], coffee consumption (0 = non, 1 = currently, 2 = more 3 cups per day), sex [0 = man‐1 = woman], all comorbidities history [0 = No‐1 = Yes], marital status [0 = single, 1 = in couple], and residence in a city [0 = No‐1 = Yes]. Adjusted linear mixed model with the first 9 variables is described in detail in Table S2 (p‐value <0.035 for all variables except for history of psychiatric morbidity, and interaction between educational level and age).

Figure 3

Partial dependence and SHAP values of the most relevant variables for prediction and cUHDRS. Mean partial dependency from MERF fully adjusted model. SHAP values of the predictions of fully adjusted mixed‐effects random forest model. Ranges of feature value: age [18–80], number of CAG repeats [36–55], education level [0–6 ISCED levels], BMI [18–36], pharmacological depression and anxiety treatment [0 = No‐1 = Yes], units of alcohol per week [0–30], tobacco and drugs consumption [0 = No‐1 = Yes], coffee consumption (0 = non, 1 = currently, 2 = more 3 cups per day), sex [0 = man‐1 = women], all comorbidities history [0 = No‐1 = Yes], marital status [0 = single, 1 = in couple], and residence in a city [0 = No‐1 = Yes]. PD: partial dependency. Adjusted linear mixed model with the first 9 variables is described in detail in Table S2 (p‐value <0.035 for all variables except for history of psychiatric morbidity, and interaction between educational level and age).

Figure 4

cUHDRS partial dependence plot with SHAP values of age and CAG repeats number with most relevant trajectories of the interacting variables in the mixed‐effects random forest model. SHAP values of the predictions of fully adjusted mixed‐effects random forest model. CAG repeat number blue<42≤red. Body index mass (BMI) blue = [12.5–23.1], purple = [23.2–27.0], and red= [27–50]. Education level blue<4≤red. SHAP interaction values in Figure S5.