Pre-diagnostic ALSFRS-R slope provides limited incremental information about the variability in subsequent functional decline in amyotrophic lateral sclerosis: a population-based study

Abstract Background The pre-diagnostic ALSFRS-R slope is frequently used in a clinical trial design and clinical practice under the implicit assumption that it reflects subsequent functional decline. We assessed the association between the pre-diagnostic ALSFRS-R slope and subsequent functional decline. Methods The Piemonte and Valle d’Aosta ALS Registry (PARALS) was used as the primary cohort. The associations between pre- and post-diagnostic slopes at 6, 12, and 18 months were examined using correlation coefficients and multivariable linear regression. Linear and quadratic mixed-effects models used all ALSFRS-R assessments obtained during the first 18 months to evaluate longitudinal trajectories and quantify the incremental contribution of the pre-diagnostic slope. All analyses were replicated in the Emilia Romagna Registry for ALS (ERRALS). Results A total of 2,222 patients from PARALS and 1,543 from ERRALS were included. In PARALS, the pre-diagnostic ALSFRS-R slope showed only a modest monotonic association with post-diagnostic functional decline (Spearman’s ρ = 0.19–0.30) and explained 6.1% of its variance at 6 months, with this proportion decreasing thereafter. Quadratic mixed-effects models estimated that, at 12 months, the difference in ALSFRS-R change between patients in the highest and lowest pre-diagnostic slope quartiles was 2.04 points in PARALS (95% CI 1.25–2.83) and 2.47 points in ERRALS (95% CI 1.07–3.86). However, the linear and quadratic interactions together increased the marginal R 2 by only 0.43 percentage points in PARALS and 1.68 percentage points in ERRALS. Conclusions The pre-diagnostic ALSFRS-R slope identifies meaningful average differences in subsequent decline between patient groups but provides limited information for predicting individual trajectories.