Cholesterol and violent deathBMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6963.1228 (Published 05 November 1994) Cite this as: BMJ 1994;309:1228
Clinical importance is questionable
- D Owens
EDITOR, - I have several worries about the relation between cholesterol concentration, violent death, and mental disorder proposed by Ann Ryman.1 Firstly, although the violent deaths referred to are due to accident, violence, trauma, and suicide,2 attention rests on suicide and the reported connection between violent death and cholesterol concentration somehow becomes a link between depression and lipid neurochemistry.
Secondly, cohort studies examining cholesterol concentrations and violent death have yielded mixed findings. Selective reviews tend to put greater weight on studies that suggest that cholesterol concentration is a risk factor. When an association has been reported there is enormous scope for potential confounders, including alcohol, diet, existing mental illness, and social group.
Thirdly, although intervention studies of cholesterol lowering have minimised confounding, they have led to exaggerated certainty about causative connections. Although each of the six intervention studies systematically reviewed showed a small excess of violent deaths among participants who received the cholesterol lowering intervention (2, 3, 4, 6, 7, and 7 out of 1147 deaths in all),2 none reported that the excess was significant. Only when all the data were pooled in a meta-analysis did a significant excess of violent deaths emerge.2 What is the clinical importance of this finding?
Each trial …