A case for measuring both cellular and cell‐free mitochondrial DNA as a disease biomarker in human blood