The issue here is that large doesn't make good. That's a particular retrospective kind of study that suffers from failure to document, record and have an accurate estimate, and the kind of symptoms that often get reported never show up in medical records, so—
On May 13th, 2019. See this statement in context.