Psychometric properties of the symptom and nutrition assessment questionnaire: a preliminary investigation using real-world digital health data from WellTheory

Introduction Autoimmune conditions impose multidomain symptom burdens that existing validated instruments inadequately capture in digital health contexts, where proprietary patient-reported outcome measures are widely deployed but rarely formally evaluated. Methods This retrospective study examined the psychometric properties of the Symptom and Nutrition Assessment Questionnaire (SNAQ) in 734 adults enrolled on the WellTheory digital health platform, approximately one-third of whom carried a formal autoimmune diagnosis. Results Internal consistency of the final 10-item measure was acceptable to good (Cronbach's α = .83; McDonald's ω = .91; ω h = .63). Exploratory factor analysis in a training subsample supported a four-factor structure reflecting neuro/autonomic symptoms, fatigue and energy, musculoskeletal pain, and upper gastrointestinal symptoms, confirmed via confirmatory factor analysis in an independent held-out test subsample. Total scores showed small-to-moderate positive associations with depression, anxiety, fatigue severity, and pain interference, and a small negative association with life satisfaction, with domain-level patterns consistent with a priori hypotheses. Among 178 participants with paired assessments within 24 weeks, total SNAQ scores showed a statistically significant reduction (mean difference = -1.89 points; SRM = -0.45), with the largest domain-level change observed for Upper GI symptoms. Discussion These findings provide preliminary support for the SNAQ's psychometric properties as a multidomain symptom monitoring instrument and offer a methodological model for evaluating proprietary outcome measures in real-world digital health settings. Future work should prioritize discriminant validity testing, test-retest reliability, and establishment of clinically interpretable change thresholds.