* It does not show human harm, only cellular disruption.
* It uses an unnatural exposure method.
* It builds on epidemiological correlations that may be reverse causality.
* It does not account for systemic factors, metabolism, or adaptive responses.