Association Between LDL Cholesterol and All-Cause and Cause-Specific Mortality in Denmark

This large Danish study reported a finding often quoted out of context: in the general population, the relationship between LDL cholesterol and all-cause mortality was U-shaped, with both very low and very high LDL associated with higher death rates than moderate levels (around 3.6 mmol/L).

Taken at face value, this might seem to challenge “lower is better.” But the interpretation requires care. The study is observational, and very low LDL in an untreated general population often reflects underlying illness — cancer, frailty, liver disease — that lowers cholesterol and raises mortality (reverse causation). This is fundamentally different from therapeutically lowering LDL in at-risk patients, which the randomized trials and genetic studies consistently show reduces events without a mortality floor.

The study is a legitimate observation about population epidemiology, but it does not contradict the causal evidence that lowering atherogenic lipoproteins with treatment is beneficial — a distinction critics sometimes blur.

We rate the evidence moderate. It is a large, well-conducted cohort, but its U-shaped association is prone to reverse-causation confounding and should not be read as evidence against treating high LDL.