Risk Seen in Lowering Level of Cholesterol
A growing body of evidence suggests that diets and drugs that lower blood cholesterol levels may indirectly raise the risk of certain types of violent death by producing personality changes--inducing anger, irritability, aggressiveness and increased risk-taking.
The studies have set off competing waves of concern and skepticism in the medical community. Although some are calling for a serious re-examination of across-the-board cholesterol-lowering tactics, many others say the results have been overblown and that much more evidence is needed to justify the reversal of such an established medical policy.
Several large studies in the United States and Europe have shown that lowering cholesterol reduces the risk of fatal heart attacks, as it was intended to do.
But, through mechanisms that scientists do not understand, the decrease in deaths from heart attacks has been offset by an increased risk of death from suicides, violence and accidents.
Some researchers now believe that lowering cholesterol levels can, in at least some individuals, affect the levels of brain neurotransmitters, the chemicals by which individual brain cells communicate with each other.
Experiments in monkeys show that lowering cholesterol reduces levels of the neurotransmitter serotonin. Low levels of serotonin in humans have previously been linked to an increased risk of suicide and aggressive behavior.
“These studies throw into a cocked hat the whole proposition that every American should lower his cholesterol levels,” said Dr. Hyman Engleberg, an internist at Cedars-Sinai Medical Center. “People who have had a heart attack, stroke or a very bad early history have far more to gain than to lose by lowering their cholesterol, but for a healthy population, they have as much to lose as to gain.”
“Right now is a good time for people to pause and look at the recommendations (for lowering cholesterol) more critically,” said Dr. Matthew Muldoon, a clinical pharmacologist at the University of Pittsburgh who conducted one of the major studies. “But not enough is known yet to say that the recommendations ought to be changed or that there are definite dangers.”
Many researchers think the concern is overblown. “I don’t dismiss it, but I don’t think the data for harm is particularly impressive,” said Dr. Basil M. Rifkind, chief of the lipid metabolism and atherogenesis branch at the National Heart, Lung and Blood Institute in Bethesda, Md. “To draw any conclusive results from (the data available) is hazardous and, perhaps, stretching the data beyond the point where it should be stretched.”
The questions about cholesterol are of more than academic interest. Coronary artery disease is the nation’s No. 1 killer. Each year, 1.5 million Americans suffer heart attacks and 550,000 die. The primary causes of heart disease include smoking, high blood pressure, obesity, family gene inheritance--and high cholesterol.
Cholesterol is a waxy, odorless substance that is synthesized by the body, but that is also present in a number of foods, especially meat and dairy products. It plays a variety of positive roles in the body, serving as a precursor for sex hormones, as a building block for the membranes that surround all cells in the body and as an aid to digestion.
The problems begin when the body has too much cholesterol, either because there is an overabundance in the diet or because the individual is genetically inefficient at removing it from the blood stream. In either case, the excess cholesterol is slowly deposited on the walls of blood vessels, building up over decades until an artery suddenly becomes too constricted to sustain blood flow.
If the artery is in the heart, the constriction usually leads to a heart attack, killing that part of the heart muscle fed by the artery. If the artery is in the brain, the individual has a stroke and brain tissue dies. If the artery is in the leg, the blockage may lead to its amputation.
Public health authorities have long urged Americans to lower the amount of fats in their diet to reduce cholesterol, but the campaign took on new urgency in 1985 with the launching of the National Cholesterol Education Program, sponsored by 26 organizations and funded by the National Heart, Lung and Blood Institute. Its goal is for Americans to know their cholesterol level and take action if it reaches dangerous levels: 200 milligrams of cholesterol per deciliter of blood if two other risk factors are also present or 240 mg/dl with no other risk factors.
More than 25% of the American population have levels higher than 240 mg/dl, as do fully half of those ages 45 to 65.
No one is questioning the value of reducing cholesterol in those individuals at high risk of heart disease. “The evidence is overwhelming that mortality . . . is diminished by lowering serum cholesterol after a heart attack,” said Dr. Thomas Chalmers, a clinical epidemiologist at the Harvard School of Public Health.
But regarding those who have not had a heart attack and whose cholesterol levels are low, he said, “There comes a time where the risk of dying of a heart attack is so small that any risk of side effects is large enough to worry about.”
The concern about such side effects arose as long ago as 1978 when physicians analyzed the results of a World Health Organization-sponsored trial of the cholesterol-lowering drug clofibrate in nearly 15,000 men age 30 to 59. The study showed that daily use of clofibrate reduced the number of deaths from heart attacks by almost half, but the number of deaths from other causes in the drug-taking group nearly doubled. Overall, the total number of deaths in the treated group was virtually identical to the number in the untreated group.
“We totally had no idea that this might happen, no prediction that it might occur and no hypothesis about why it happened,” said Dr. Michael F. Oliver, a cardiologist at the Wynn Institute for Metabolic Research in London who was in charge of the WHO study. “We were very worried about it.”
Subsequent studies in the United States and Europe with three other cholesterol-lowering drugs, as well as with diet alone, yielded similar findings.
In 1990, a team at the University of Pittsburgh combined the data from six large trials of cholesterol reduction involving more than 25,000 subjects. Using a sophisticated technique called meta-analysis, they found a significant reduction in deaths from coronary heart disease, but an equally large increase in deaths from accidents, suicides and other violence.
Similar results have been observed in two meta-analyses of other studies, one by Rifkind and his colleagues at the National Heart, Lung and Blood Institute and one by Dr. Richard Peto, an epidemiologist at the Imperial College in London.
“There are a lot of people in the medical profession, (in) the cholesterol field, who frankly don’t believe that there is any effect on these other mortality categories,” said Muldoon of the University of Pittsburgh. “In our opinion, that’s a difficult view to defend because . . . the statistics say it is not a fluke.”
It is also clearly not an effect of the drugs themselves, said Oliver, because identical results have been obtained with four drugs belonging to three widely disparate classes. Hence, the effect must involve some intrinsic properties of the cholesterol-lowering process itself.
Searching for a solution to the riddle, the Pittsburgh group teamed up with Dr. Jay R. Kaplan, a professor of comparative medicine at Bowman Gray School of Medicine in Winston-Salem, N.C., for a study of the effects of cholesterol reduction on 30 monkeys.
The monkeys were already being studied in a 22-month experiment to investigate dietary and behavioral influences on atherosclerosis, hardening of the arteries. Half received a “luxury” diet comparable to that consumed by most Americans. The other half received a “prudent” diet that contained only one-sixth as much cholesterol.
The monkeys’ behavior was monitored carefully over the course of the experiment. Kaplan and his colleagues reported late last year that the animals receiving the prudent diet displayed significantly more aggression than those on the luxury diet. That aggression included fighting, threatening gestures and grimacing and other facial expressions.
Previous studies in humans have linked aggression, as well as suicides, with lowered brain levels of the neurotransmitter serotonin. This led the Pittsburgh and Bowman Gray groups to study serotonin levels in the monkeys, using a blood test that provides an indirect measure of serotonin in the brain.
They reported last month in the journal Biological Psychiatry that serotonin levels in some of the “prudent” monkeys were, in fact, reduced. There was little difference in the overall serotonin average between the two groups, Muldoon said, “but several animals had very much lower levels.”
From these and other results, the researchers reason that lowering cholesterol in individuals who do not have high levels to begin with concurrently lowers the level of serotonin in the brain. That, in turn, causes some individuals to be more aggressive in their behavior toward others and to take more risks, thereby increasing the chance that they will die from accidents or violence. It also increases the risk that they will impulsively commit suicide.
In many cases, the effects might be more subtle but still damaging, argues Dr. David F. Horrobin, a pharmacologist at Scotia Pharmaceuticals Ltd. in Guildford, England. Reducing cholesterol levels could lead to “a more violent pattern of behavior, most of which would not result in death, but would lead to more aggression at work and in the family, more abuse of children and partners and generally more unhappiness,” he wrote recently in the British Medical Journal.
But such opinions are still in the minority. More representative is Rifkind, who argues that the cholesterol-lowering studies, including his own, were not designed to detect deaths from accidents and violence and that, even when submitted to meta-analysis, they are “far too small to draw conclusions with any confidence. By some estimates, you might have to have 75,000 people studied for five years or more to find out what happens to total mortality.”
Rifkind said: “We think it is something to keep watching, but in terms of giving real cause for concern, most of us feel that it is not a serious potential hazard of cholesterol control.”
Meanwhile, Muldoon and his colleagues are just beginning a study--sponsored by the National Heart, Lung and Blood Institute--examining the effects of cholesterol-lowering on mood and behavior in 400 patients. Kaplan is also beginning studies of the effects of low cholesterol in monkeys, especially during puberty and other stages of development. “There is a tremendous amount of virgin ground to be plowed,” Muldoon said. “There are a lot more things that we don’t know than that we do know about this phenomenon.”
Researchers are also eagerly awaiting the results of a new set of clinical trials of a new family of cholesterol-lowering drugs collectively called statins. Those studies have been designed with the serotonin question in mind, and results are expected in the mid-1990s. “We can’t say confidently that they will be definitive,” Rifkind said, “but perhaps they will give us some valuable new information.”
But the final word for now perhaps comes from Harvard’s Chalmers. “My own opinion of this whole business is that if you’ve had a heart attack or your likelihood of dying is high (because of risk factors), you ought to lower your cholesterol levels,” he said. “If you haven’t had a heart attack and your cholesterol isn’t high, you ought to eat a normal diet. But remember, moderation in all things.”
