Recent headlines have suggested that the long-term official health advice on dietary fat, and especially saturated fat, is not based on solid evidence and is no longer appropriate. Instead of reducing our fat intake, we are now being told that sugar is the real villain.
This type of controversy is music to the ears of the editors of newspapers, magazines and radio and TV programmes. It is not newsworthy to say that the advice from last year is still valid, but a challenge to that advice will fill a page or make for an exciting broadcast. Unfortunately, it will also confuse people. Who is telling the truth? Who should we believe? If the advice keeps changing, should we bother to change our diet?
Of course, advice on diet and health is subject to change as new research evidence comes to light. Different groups of experts may come to different conclusions from the same experimental data; there is always some element of fashion, and we should not ignore the possibility that industrial lobbies may attempt to influence the debate. Sugar, dairy produce, margarine and cooking oil are all large industries; many manufacturers support research in universities and their own laboratories and may sponsor symposia that give only a positive view of their products. Most scientific and medical journals now require the authors of research to declare any sources of funding and conflicts of interest. This means that the cautious reader can see where there is a possibility of bias.
This new debate on fat vs sugar is challenging a 30-year consensus that we should get no more than 30 per cent of our energy intake from fat (the western average is nearer 40 per cent). Furthermore, no more than one-third of that fat should be saturated (perceived as having adverse effects), which means increasing the proportion of mono- or polyunsaturated fats (seen as beneficial). There is also a consensus that sugar should provide no more than 10 per cent of our energy intake.
The association of fat intake with atherosclerosis (fatty plaque deposits inside the arteries) and coronary heart disease (CHD) was first suggested in the 1950s and 1960s as a result of epidemiological research; that is, the study of the health of defined populations. This research compared people in different countries, eating different diets, and with different patterns of disease, and concluded that people with a higher intake of fat had higher incidence of atherosclerosis, CHD and some cancers. Of course, many other factors may be involved, including ethnicity, environment and lifestyle. There were also studies of people who had migrated from one country to another, together with evidence of whether or not they had changed their diets, plus comparisons with first-degree relatives still living in the home country. These provided further evidence linking diet and the so-called diseases of affluence.
Studies of Yemeni Jews who migrated to Israel in the late 1940s and early 1950s suggested that a high intake of sugar (a rare luxury in Yemen but plentiful in Israel) was a factor in causing diabetes, atherosclerosis and CHD. These diseases were rare among the Jews in Yemen but common after they moved to Israel.
Long-term cohort studies, where a group of people are followed for many years, can also be useful. The oldest is a 1946 birth cohort study in the UK. Every child born in the second week of March 1946 is still being followed, and new cohorts have been added since. An American study has followed the population of the town of Framingham, Massachusetts, since the late 1940s. The Nurses’ Health Study and the Physicians’ Health Study (both in the USA) are following large numbers of health professionals. From diet records, lifestyle questionnaires and health records, many potentially important factors can be identified.
But epidemiology can only suggest links. It does not prove cause and effect. For this we need experimental evidence, either from laboratory animals or human intervention trials. It is relatively easy to feed rats or rabbits diets high in sugar or fat and look for signs of disease; but this diet is very foreign to the animals and it is difficult to extrapolate the findings to human beings. There is a problem with human intervention trials, too. Atherosclerosis begins to develop in early adult life, but does not become clinically important until 20 or 30 years later. This is where long-term cohort studies can help, but trials of changes in diet cannot be continued for long enough to give conclusive results. In 1987 there was a national intervention in Mauritius because of concern over high CHD rates. Only one (government-owned) factory produced cooking oil, and the decision was made to switch from palm oil (high in saturated fats) to soya-bean oil (lower in saturated and higher in mono- and polyunsaturated fats). Nine years later a report showed the expected decrease in blood-cholesterol levels nationwide, but to date there has been no report of a decrease in CHD. But it may still be too soon to see any effect on premature death and disease.
It is well established that a higher level of blood cholesterol increases the likelihood of developing atherosclerosis. As the thickness of the fatty plaque increases, so the internal diameter of the blood vessel decreases. When this plaque is laid down in the coronary arteries that supply the heart with blood, the result is blockage of one or more arteries, leading to the death of the heart-muscle cells that they supply. This is what we call a coronary thrombosis or a heart attack.
Epidemiological studies suggested that not only the total amount of fat in the diet, but also the relative amounts of saturated, mono- and polyunsaturated fats, were significant. This led to a number of short-term trials. Volunteers were fed diets containing the same total fat, but with differing proportions of each type. The results showed that saturated fats increased blood cholesterol in proportion to the intake, while polyunsaturated fats decreased it proportionally. The obvious conclusion was that decreasing the first type of fat and replacing it with either of the other two types would reduce blood cholesterol, and it was therefore likely to reduce the risk of atherosclerosis, too.
We can now provide a biological mechanism to explain this. Most blood cholesterol is carried in low-density lipoprotein (LDL), and having a high concentration of this in the bloodstream is the underlying cause of atherosclerosis. LDL is normally cleared by the liver, by way of receptors on the surface of liver cells. But any LDL that has suffered chemical damage in the circulation (mainly oxidative damage) is not recognised for what it is by the liver receptor, so it stays in the bloodstream. It is then taken up by circulating macrophages (part of the body’s defence against bacteria) and these lipid-engorged macrophages burrow under the surface of blood vessel walls, where they break down the LDL. They are killed by the free cholesterol that they release, so laying down cholesterol-rich fatty plaque on the blood-vessel walls.
Most cholesterol absorbed from the diet, as well as cholesterol newly synthesised in the liver, is esterified (chemically combined) with fatty acids before being exported from the liver for uptake by tissues that need it. The remnants are left in LDL to be cleared by the liver. Saturated fatty acids are poor candidates for the esterification of cholesterol, while mono- and poly-unsaturated fatty acids work well. This means that if there are relatively large amounts of saturated fatty acids in the liver, this will result in more unesterified cholesterol in the liver, which reduces the effectiveness of the LDL receptors, leaving more LDL in the blood to be taken up by macrophages.
The total amount of fat in the diet is also important. Dietary fat is absorbed by the formation of chylo-microns in intestinal cells. These are large assemblies of fat and protein which are absorbed into the lymphatic system, then enter the bloodstream, where tissues will take up such fat as they require. The fat-depleted chylomicron remnants are then cleared by the liver using the LDL receptors. The more fat in the diet, the more chylo-microns are formed. That leaves more remnants to be cleared by the liver, competing with LDL for uptake. So an increased concentration of chylomicron remnants will lead to an increase in circulating LDL to be taken up by macrophages.
That’s the hard biological evidence for the damage caused by fat. Now what about sugar? Studies have also shown a high sugar intake will lead to an increase in LDL cholesterol. Table sugar is sucrose, a combination of glucose and fructose. The metabolism of glucose by the body is strictly controlled, and when energy needs have been met, any left over is stored in the liver and muscles in the form of the carbohydrate glycogen. With fructose, when enough has been metabolised to meet energy needs the excess becomes part of a process which forms fatty acids in the liver. These are then exported in very low density lipoprotein (VLDL). Tissues take up fatty acids from VLDL, and indirectly donate cholesterol to it and turn it into regular LDL. So a high intake of fructose (mainly from sucrose) will also lead to an increase in LDL.
Ultimately, the research shows that eating too much fat and eating too much sugar can trigger obesity. And even if short-term intervention trials do not provide evidence that cutting back on saturated fat and eating more mono- and polyunsaturated fat cut your risk of atherosclerosis and coronary heart disease, these biological mechanisms suggest that it is still highly prudent advice.
David A Bender is Emeritus Professor of Nutritional Biochemistry at University College London.