Response to letter to the editor: Use natural logarithms not base 10 logarithms to compare group means (re: Increased thymus weight in sudden infant death syndrome compared tocontrols: The role of sub-clinical infections).