Causation or correlation? For instance, perhaps a poor diet encourages the presence of certain oral bacteria over others, and also is causative for pancreatic cancer without the bacteria having any causative effect on the cancer. (I do believe that the microbiome has a massive effect on health, but this is either lazy science or lazy science reporting in jumping to a conclusion of causation.)