As we saw in the previous post, low cholesterol levels are not associated with lower mortality.
But what about randomized trials? Do randomized trials support a mortality benefit for lowering cholesterol?
To answer these questions, I created the following table of eleven lipid-lowering treatments and their corresponding risk ratios for all-cause mortality:
According to the table, most cholesterol-lowering treatments — 91% of them — do not provide a survival benefit.
Thus, the data contradict the "lower the better" hypothesis.
Non-Statin Drugs and All-Cause Mortality
In 2021, Zhou, Stouffer, and Smith stated:
Failing to demonstrate survival benefit with up to 4-7 years of follow-up in the non-statin therapy trials raises reasonable question about their efficacy especially in survival benefits.
To emphasize this point, I conducted a further analysis of non-statin treatments using the major double-blinded trials:
Based on 16,606 deaths, a precise estimate, and up to seven years of follow-up, the results confirm that non-statin drugs do not reduce mortality.
What About Statins?
At face value, statins are the only lipid-lowering drugs that provide a mortality benefit.
But the size of this alleged benefit, on average, is small.
Byrne and colleagues (2022), for example, reported an estimated relative risk reduction of 9% and an absolute risk reduction of only 0.8%.
Although lipid-lowering supporters like to extrapolate these small benefits beyond the trial, this requires assumptions that may not be true:
We can only know about the benefit accrued during the trial. Benefit beyond the trial has to be modelled on the basis of assumptions [Hallas, Kristensen, Christensen (2015)].
But are these small benefits real?
For instance, a mortality benefit could be questioned in many populations, including primary prevention, the elderly, diabetics, and other specific populations (e.g., heart failure and kidney patients).
Some researchers also argue that statins appear beneficial only in early trials — before the enforcement of stricter clinical trial regulations.
To quote Okuyama and colleagues:
We have previously argued that high levels of LDL-C serve as a predictor of longevity and also that RCTs performed after 2004–2005 (when new regulations on clinical trials came to effect) failed to demonstrate that statins decrease objective measures of CVD, such as MI mortality and/or all-cause mortality.
In 2020, Savannah Browning analyzed this hypothesis and found that statins have less benefit in trials published after 2004.
As she concluded:
The major finding of this meta-analysis is that among studies conducted after the 2004 changes in clinical trial regulations, no differences were found for the primary outcome occurrence, cardio-related mortality incidence or all-cause mortality incidence comparing statin therapy to control interventions.
I also found a similar result in meta-analysis. Statin trials before 2005 show a small mortality benefit on average whereas trials from 2005 onward do not:
Even worse, the evidence is inconsistent for trials published before 2005. In some populations, statins may have a benefit, but in other populations statins may be ineffective or harmful (based on a prediction interval of 0.69 to 1.09).
Limitations
Randomized trials are designed to test "treatments" or "drugs," not cholesterol lowering or LDL lowering per se. This means that randomized trials do not actually test the "causal effect" of cholesterol/LDL lowering.
To compound this issue, the true effects of lipid-lowering treatments are not known. Trials of lipid-lowering drugs are often plagued by conflicts of interest, methodological problems (both internal and external validity), and lack of prospective registration.
Conclusions
The vast majority of clinical trials show a lack of association between drug-induced reductions in cholesterol levels and all-cause mortality.
Therefore, there is no credible evidence that a lower cholesterol level will prolong survival or provide a net health benefit.