Digital Speech Markers for Monitoring ALS in Spanish Speakers

Overview

The goal of this observational study is to learn how speech and breathing change over time in Spanish-speaking individuals with amyotrophic lateral sclerosis (ALS) compared to age- and gender-matched individuals without ALS. The main questions it aims to answer are: Can speech and breathing measures collected through a smartphone application serve as reliable digital biomarkers to track bulbar disease decline in Spanish-speaking people with ALS? How do these measures differ between individuals with ALS and those without ALS? Researchers will compare Spanish-speaking participants with ALS to age- and gender-matched healthy controls to see if specific speech and breathing features can identify or predict bulbar decline. Participants will: Use a Spanish-language smartphone application to record speech and breathing tasks over time. Complete assessments of speech, breathing, and functional abilities (e.g., ALS Functional Rating Scale). Provide data that will be compared to caregiver reports and clinical outcomes to validate new digital biomarkers.

BACKGROUND AND RATIONALE Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disorder characterized by the loss of upper and lower motor neurons, leading to paralysis, speech and swallowing decline, and respiratory failure. Approximately 90% of people living with ALS (pALS) experience bulbar symptoms, which severely affect communication, nutrition, and overall quality of life. Despite their clinical significance, bulbar symptoms remain poorly characterized, particularly in non-English-speaking populations. The absence of validated Spanish-language digital biomarkers represents a significant gap in the ability to monitor disease progression, design inclusive clinical trials, and improve access to care. This observational study aims to address that gap by developing and validating speech- and respiratory-based digital biomarkers for monitoring bulbar disease decline in Spanish-speaking individuals with ALS, using a smartphone application designed for remote data collection. STUDY DESIGN AND OBJECTIVES This is a prospective, longitudinal observational study involving approximately 30 Spanish-speaking individuals with ALS (pALS) and 5 age- and gender-matched healthy controls. The study leverages a Spanish-language Everything ALS smartphone application for data collection. Primary Objectives: Develop and validate speech and respiration-based digital biomarkers for tracking bulbar disease decline in Spanish-speaking pALS. Assess the sensitivity of maximum phonation time (MPT) and related parameters for detecting early respiratory and bulbar changes. Secondary Objectives: Compare digital biomarker profiles between ALS participants and matched controls to determine specificity. Evaluate correlations between digital biomarker features and established clinical outcomes, including the Spanish ALS Functional Rating Scale (ALSFRS-R). Assess feasibility, adherence, and participant satisfaction with remote digital data collection methods. PARTICIPANT ACTIVITIES Participants will complete the following study activities: Use the Spanish-language Everything ALS smartphone application to record structured speech and breathing tasks (e.g., sustained vowel phonation, reading passages, breathing exercises) at scheduled intervals. Participate in monthly engagement check-ins with a trained student proctor via Zoom or telephone, based on participant preference. Receive email reminders prior to each recording session to promote adherence. Access their individual dashboard ("My Data / Mis Datos") within the app to review past recordings and engagement history. ALS participants who attend multidisciplinary clinic visits may complete optional concurrent in-person assessments aligned with standard clinical care to reduce burden. Healthy control participants will follow a parallel smartphone-based protocol on a comparable timeline to generate reference data. DATA COLLECTION AND MANAGEMENT All data will be captured through the secure Everything ALS platform. Recordings are automatically encrypted during upload and stored on HIPAA-compliant servers. Each participant is assigned a unique identifier to ensure confidentiality. Quality Assurance and Data Validation: Quality Assurance Plan: Automated procedures verify audio completeness, recording duration, and signal quality upon upload. Range and Consistency Checks: The database includes predefined limits for key acoustic and respiratory measures (e.g., phonation duration, frequency range, signal-to-noise ratio). Out-of-range data trigger review by the data management team. Source Data Verification: Random data samples will be cross-checked with clinical assessments and ALSFRS-R records for validation. Data Dictionary: A detailed data dictionary defines all study variables, their sources, and coding standards, including MedDRA terminology for any adverse events or technical issues. Standard Operating Procedures (SOPs): SOPs govern all aspects of registry operations, including participant onboarding, remote task completion, data monitoring, analysis workflows, and management of missing or uninterpretable data. QUALITY CONTROL AND OVERSIGHT Continuous monitoring ensures data completeness, reliability, and participant safety. The Everything ALS data management team performs periodic audits and routine system checks. Trained proctors provide participant support and document engagement metrics to reduce attrition. No medical interventions or experimental treatments are administered. Although adverse events are not anticipated, any participant-reported concerns will be reviewed and documented according to institutional oversight procedures. SAMPLE SIZE AND STATISTICAL ANALYSIS PLAN The target enrollment is approximately 30 Spanish-speaking ALS participants and 5 age- and gender-matched healthy controls. This sample size provides adequate power (\>80%) to detect moderate correlations (r = 0.4-0.5) between digital biomarkers and clinical outcomes over time. Planned Analyses Include: Longitudinal mixed-effects models to assess within-subject change in digital biomarkers. Between-group comparisons (ALS vs. controls) using ANCOVA or nonparametric equivalents. Correlation analyses linking digital biomarkers to ALSFRS-R subscores and disease phenotype. Sensitivity testing of maximum phonation time and related features as indicators of bulbar decline. PLAN FOR MISSING DATA Missing data will be addressed using a predefined strategy: Missing or low-quality recordings will prompt re-collection via automated reminders. Analyses will use mixed-model approaches, which accommodate missing-at-random data. Sensitivity analyses will estimate the impact of incomplete or inconsistent recordings. ETHICAL CONSIDERATIONS AND PARTICIPANT ENGAGEMENT All participants will provide electronic informed consent before participation. Spanish-language materials and instructions will be used to ensure comprehension and accessibility. In response to feedback from a pALS focus group held at the Everything ALS Biomarkers Summit (August 2024), participant engagement and retention strategies include monthly virtual contact and access to personalized app-based feedback. These elements were designed to improve engagement and minimize attrition during the remote study period. RELEVANCE AND IMPACT This study represents the first effort to validate Spanish-language digital speech and respiratory biomarkers of bulbar disease progression in ALS. By enabling inclusive, remote, and noninvasive monitoring, it will improve early detection of bulbar involvement and enhance equity in ALS research participation. The project supports the NIH's "Accelerating Access to Critical Therapies for ALS Act (ACT for ALS)" and aligns with NINDS strategic priorities to expand access, improve data diversity, and enhance quality of life for people living with ALS.