Fair point; it's not that changing diagnostics mean that there can't be environmental/dietary changes contributing to increased autism cases, it's that the article just entirely skips over that part. It describes an outcome, describes a potential cause that could explain that outcome and then says, "and obviously also diets too."