But mevalonate is a precursor of other substances as well, substances that have important biological functions (Toxicol Lett, 1992; 64/65: 1-15). Al-though the metabolic pathways are not known in all of its details, reduced amounts of mevalonate may explain why simvastatin makes smooth muscle cells less active (Atherosclerosis, 1992; 95: 87-94) and platelets less inclined to produce thromboxane (Circulation, 1992; 85: 1792-8). One of the first steps

in the process of atherosclerosis is the growth and migration of smooth muscle cells inside the artery walls, and thromboxane is a substance that is necessary for blood to clot.

By blocking the function of smooth muscle cells and platelets, simvastatin may benefit cardiovascular disease by at least two mechanisms, both of which are independent of cholesterol levels. In one experiment by Japanese researcher Dr Yusuke Hidaka and his team, the inhib-itory effect of simvastatin on muscle cells could not be abolished by adding LDL cholesterol to the test tubes (Ath-erosclerosis, 1992; 95: 87-94) and, in experiments that compared the statins with several different cholesterol-lowering agents, thromboxane production was inhibited only by the statins, indicating that the effect was not due to cholesterol reduction itself, but to something else (Circulation, 1992; 1792-8).

The protective effects of simvastatin were also demonstrated in animal experiments. In one such study by Dr B.M. Meiser and colleagues in Munich, Germany, hearts were transplanted into rats (Transpl Proc, 1993; 25: 2077-9). Normally, the function of such grafts gradually deteriorates because the coronary vessels become narrowed by an increased growth of smooth muscle cells in the vessel walls. This condition is called graft vessel disease, a condition with many similarities to early athero-sclerosis.

In Dr Meiser's experiment, however, rats that received simvastatin had considerably less graft vessel disease than control rats, which did not receive simvastatin, and this was not due to cholesterol reduction because simvastatin has been shown not to lower cholesterol in rats. In fact, LDL cholesterol was highest in the rats that received simvastatin.

In another experiment, Dr Maurizio Soma and his colleagues in Milan, Italy, placed a flexible collar around one of the carotid arteries in rabbits (Atherosclero-sis, 1993; 25: 2077-9). After two weeks, the arteries with collars were narrowed, but less so if the rabbits had received simvastatin. Again, the effect was not related to the rabbits' cholesterol levels.

Thus, the statins in some way protect against cardiovascular disease, but their effect is not due to cholesterol reduction. The proponents of the cholesterol hy-pothesis have simply had incredible luck in finding a substance that prevents cardiovascular disease and, at the same time, lowers cholesterol.

Statistical manipulation

But why bother about pharmacological mechanisms. Isn't it wonderful that the statins work? Should we all be taking statins?

To answer that question, it is necessary to look at the figures from the

trials. Coronary mortality in these studies was lowered by 19 per cent to 41

per cent-most in the 4S trial and least

in the CARE trial. These are the so-called relative risk figures, used by most doctors and by the drug companies in their ads, which reflect the probability or