Confounding factors can lead to biased results, making it appear as though there is an association when there is none (or masking a true association). In epidemiological research, failing to address confounding can result in incorrect conclusions, misinforming public health policies and interventions.