The journal papers in the table in previous post (and included in here), are the ones I use to argue ALS median is ordinarily slightly below three years for symptom onset.

Table of ALS datasets