A family with a high cholesterol - revision

                Canal side, Amsterdam, Netherlands

[Apologies for poor formatting in previous post]

Familial hypercholesterolaemia (FH) is a condition that is genetically determined, and it causes a high level of cholesterol in the blood.  The families affected have experienced a high incidence of coronary heart disease (CHD) with early deaths. This has been a serious problem for them during the second half of the 20th century, and as a consequence there has been a great deal of attention given to these families. But the connection between FH and CHD is not entirely straightforward.

The seriously flawed diet-cholesterol-heart hypothesis of the cause of CHD has dominated medical “non-thinking” up to the present time. It has a variant, the genetic-cholesterol-heart hypothesis. It has become clear to most informed people that diet cannot be the cause of CHD. It has also become clear that the great majority of cholesterol in our bodies is synthesised within us and has not been taken in the diet.

CHD, with or without a high cholesterol level in the blood, has run in families very strongly. It has been much more common than would be explained by simple Mendelian inheritance. However FH is genetically determined and it has a pattern of Mendelian inheritance, as we will see.

Gregor Mendel, the father of genetics

A high level of cholesterol in the blood can lead to deposition of cholesterol in body tissues, but this does not necessarily include the coronary arteries on the surface of the heart. We have seen the example of the high blood levels of cholesterol typical of the liver disease primary biliary cirrhosis (PBC) and it does not lead to an increased risk of CHD. It is important to appreciate that CHD is not due to “cholesterol”, and that cholesterol within the arterial wall is a secondary phenomenon, secondary to an underlying inflammatory process.

CHD is not a dietary disease and it is not a metabolic disease. It is an inflammatory disease. This was first recognised by Rudolf Virchow in the late 19th century but then, eclipsed by the diet-cholesterol-heart hypothesis, it remained forgotten until the late 20th century.

Rudolf Virchow, 1821-1902

Although CHD clearly runs strongly in families, this is a form of non-genetic inheritance, probably the inheritance of a micro-organism rather than genes. However there is clearly an important genetic component in familial hypercholesterolaemia (FH). CHD is worse in these families, for reasons that I hope will become clear. Also, that in the long term things are not altogether bad for those people with FH.

The epidemic of CHD

During the 20th century there was an epidemic of CHD and this is beyond dispute, that is by people who have taken the trouble to read the original papers which have described very clearly its emergence and its decline. 

Figure 1: The epidemic of coronary heart disease in the UK

During this epidemic people with FH fared very badly and had earlier deaths than the average population. Their standardised mortality ratio (SMR) reached almost twice that of the general population.

Very full records have been identified for a large family in The Netherlands, including ages at death over a long period of time. It is now known that the members of this family have FH. From the records, mortality rates have been compared with the population average (mean). If we compare an individual with the average then we are looking at his or her standardised mortality ratio (SMR). The average reference level is 100 and an SMR less than 100 indicates a survival advantage. An SMR of more than 100 is what you do not want, a high death rate. It is similar to IQ but with IQ a level greater than 100 is considered to be an advantage. 

The size of the family enabled 250 people to be studied, starting with a single pair of ancestors in the early 19th century. In those tested in more recent years the blood cholesterol levels were measured and were in excess of 10 mmol/L. Even more recently they have been identified as being carriers of the V408M mutation (see next Post). 

Cholesterol

The average blood level of cholesterol in the general population is about 5.9 mmol/L, slightly higher in young men with CHD. We can see this in Figure 2. The distribution curve of blood levels of cholesterol is shown, for young men without CHD between the ages of 30 and 50 years. If we then look at the men of the same age with CHD, we see that the curve has shifted slightly to the right, to a slightly higher level. However there is a major degree of overlap.

Figure 2: Distribution of serum cholesterol, men aged 30-50, Framingham study

CHD deaths have been quite rare in this age-group, and have now effectively come to an end. 

In Figure 3 we now look at a slightly older age-group of men, between 50 and 62 years, we see that the distribution curves are identical.

Figure 3: Distribution of serum cholesterol, men aged 50-62, Framingham study

In other words, blood levels of cholesterol are meaningless for the prediction of CHD in this age-group. This has been explained clearly in the original papers from the Framingham study, a 30 year follow up of the town in Massachusetts. The message has not been heard but here it is:

Could the message be any more clear than this?

It obviously has implications for families with high cholesterol, FH.

The family in The Netherlands

We can see the average SMR of the family in the Netherlands in Figure 4. It has changed considerably during the years 1830 to 1999. In the 19th century the family members had very significant survival advantage, with SMR for females of only 45 and for males 70. At that time it was clearly a very good thing to have FH. 

Figure 4: Relative standardised mortality rate (SMR) for a large family with familial hypercholesterolaemia in The Netherlands

The study indicates that mortality of the FH family relative to the general population increased only after 1915. In the mid 20th century the SMR reached a peak of almost 200. This was very serious with premature deaths from CHD in the family members and an SMR twice that of the general population. 

The pattern of death in the families with FH follows that of the epidemic of CHD. The rise of the graph coincides with the emergence of CHD in the 1920s. The peak was reached in about 1960-70 and thereafter the SMR, the relative death rate fell. So when there was a high prevalence of CHD, people with FH had a higher incidence of death from it. As the overall death rate from CHD in the population diminished, the relatively high SMR for people with FH came back to the national average. 

The last point on the graph represents the time period 1970-99. This is the time of a 90% drop in CHD deaths in the UK. 1999 was very different than 1970. It is unfortunate that the study did not divide 1970-99 into two periods of 1970-84 and 1985-99. A large drop of SMR might then have become visible.

It would be fascinating to see how the families with FH have fared since about 1980 when the recordings stopped. I have written to the authors of the paper asking if the investigation of the families will continue but they tell me that this is unlikely to happen. It is a unique opportunity lost. We can only speculate that those with FH will in the 21st century revert to a much better health profile and lower SMR than the overall population.

Genetics and environment

This important study is an excellent example of the interaction between genes and environment. A genetic advantage in one time or place might turn to be a disadvantage in other circumstances. 

Genetic assessment of FH is very important but it becomes more complex. I think it advisable to close this Post here and to publish an additional post very shortly. It will be "FH - Genetics & Mutations".

Reference:

Mortality over two centuries in large pedigree with familial hypercholesterolaemia: family tree mortality study.

Sijbrands et al.

British Medical Journal 2003; 322: 1019-23.

Epidemic of diabetes threatens to bankrupt the HNS

Bees on an artichoke flower in my vegetable garden

The lives of plants and animals are intriguing, but it is human life that never ceases to provide us with an endless supply of drama and humour. Newspaper headlines are, I think, generally intended to be serious, but the shock factor often becomes ridiculous. Like so many things in life, do not take them too seriously but try to identify the facts behind the scary headlines.

We are told that the number of people in the UK with diabetes has increased by an astounding 60% in ten years. In another twenty years or so we might all have diabetes. If this is true then there must be a very obvious cause. I have previously suggested that the main cause of diabetes is a visit to the doctor.

As usual the message is vicim-blaming. The epidemic is due to us becoming obese (another epidemic) and eating too much sugar. Surely our diet and other aspects of lifestyle cannot have changed so dramatically during such as short time. But obesity, although generally and officially viewed as the sign that the Grim Reaper will shortly visit us, is not the disadvantage made out.

And then there is the warning that this epidemic of diabetes will bring about the bankruptcy of the NHS. What on earth does this mean? Do people actually believe this?

"Bring down the NHS" - how could this possibly happen? How would we possibly replace it? Will we transfer pay for the million people who work in the NHS into welfare unemployment benefits for them? Will the hospitals close down?

First we must ask ourselves, is there really an epidemic of diabetes? Well, we don't know. In fact it is now difficult to know what diabetes is - ask experts to define it! 

Diabetes mellitus (DM) means "plentiful sweet urine". It has been known for a long time, since before chemical testing for glucose, when tasting of the urine was necessary. The WHO gave a clear definition of diabetes in about 1983 and it has never removed this definition. It is "a condition in which a random blood glucose is in excess of 11mmol/l on at least two occasions." 

But there is a continuum of blood glucose, like height or weight, and the problem is to define the cut-off between normal and abnormal (disease state). A cut-off point is decided by a committee (e.g. World Health organisation – WHO) rather than by a bio-scientific marker. In more recent years it has been considered that a cut-off of 11mmol/L is too high and thus too many people with unknown and unexpected diabetes might be missed by medical surveillance. Perhaps a cut-off level of 8.5 mmol/L might be in the public interest, and by applying this the number of people with with diabetes is increased enormously "at a stroke", of the pen of a committee. 

However things have moved on further and now people tell me that they have been diagnosed as having diabetes ("having become a diabetic" to use common but unfortunate language) without any knowledge of blood glucose.

The new measure is not the level of free glucose in the blood, which can vary from minute to minute, but the amount of glucose attached to proteins in the blood, and in particular to haemoglobin. The measure is HbA1c, and it is usually expressed as a percentage  the percentage of haemoglobin in the blood to which glucose is attached. Once again there is no specific demarcation between those with and without diabetes but a continuum, and an arbitrary cut-off is necessary to define the "normal".  

HbA1c has been used for several years as a measure of control of diabetes, but it has now come into use as a method of diagnosis. 

However for the purpose of population screening it is a much more expensive test than measuring glucose. Many responsible for public health suggest that blood glucose is best used for identifying diabetes. Public Health England (PHE) suggests, as quoted in The Guardian (26 August 2015), that 5 million people might have high blood glucose levels. It also suggests that "many of those who become diabetic suffer complications such as blindness and foot amputation". The reality is that "many" is just a tiny proportion, although it is very serious for those afflicted.

There is clearly a great effort being made in the UK, and probably in other countries, to identify as many people as possible with diabetes, with financial rewards for the diagnosing doctors. 

And so a combination of case-finding and redefinition of diabetes to include a much greater proportion of the population has inevitably led to a sudden and great increase in the number of people with diagnosed diabetes - that is, a pseudo-epidemic. It is simply not known whether the true incidence of diabetes is rising.

The reason behind the great effort to identify and treat people with diabetes is to prevent the associated diseases of the circulation. These are:

• disease of the large arteries (atherosclerosis),  macro-vascular disease. This would lead to heat attack (MI, deaths already reducing dramatically), stroke, and major failure of blood supply to the limbs, possibly resulting in leg amputation;

• disease of the small arteries, micro-vascular disease. This is more specific for diabetes and it causes damage to the retina (blindness should now be averted by laser therapy), kidney failure (a major problem in ethnic South Asians living in the UK), and damage to the nerves supplying the feet, with subsequent painless ulceration that might result in amputation of a foot or the front part of it.

These micro-vascular diseases occur also in people with Type 1 diabetes and are more closely associated with a high blood glucose. However these conditions affecting only the minority of people diagnosed with Type 2 diabetes.

There are two predictable and significant consequences of the diagnosis of diabetes being applied to an individual. One is worry and the other is an increase personal health insurance premiums. Is early diagnosis by population screening worth-while? 

This was investigated by the UK medical Research Council, in Cambridge. It looked at 15,089 people between the ages of 40 and 69 years, who had a high risk (top 25%) of developing Type 2 diabetes. They were invited to screening and were compared to 4137 people who were not invited. Follow up was for an average of 9.6 years and after this time there was no significant difference in outcome between the two groups.

This lack of benefit of screening has not deterred diabetes evangelism, and indeed population screening continues to expand in a variety of ways. This important research project was effectively a waste of time and it has been ignored (even though it received a prize of the research project of the year).

However the outcome measure in this study was death, and this is easy to measure for a large population sample. If we are concerned about reduction in morbidity (illness) within the population, such as visual deterioration, kidney disease, or foot amputation,  data collection over a period of ten years for almost 20,000 people becomes a very major and a very costly process, unfortunately prohibitively so.

We are therefore dealing with population screening to identify a large number of people with hitherto undiagnosed diabetes knowing that only a very small proportion might benefit, and this precise proportion is not known or predictable. 

We might just remember that Type 1 diabetes is the result of a failure of insulin production by the Islets of Langerhans in the pancreas. Treatment with insulin is essential. A major feature of  Type 2 diabetes is an increased amount of insulin in the blood, but there is resistance to its effect. The nature of insulin resistance is far from fully understood, but it means that insulin production within the body must increase. A high blood sugar will only occur when the limit of insulin production has been reached, in other words a relative rather than absolute deficiency of insulin.

Professor Edwin Gale of Bristol University has suggested that Type 2 diabetes is a disease in search of a definition. He goes on to suggest that the term Type 2 diabetes is not a specific disease as its influence on different people varies greatly. In some it is associated by a number of serious and more specific disease processes, but in others it appears to do no harm. If diabetes is to be defined by blood glucose levels then we would be intellectually more honest if we were to use the term "idiopathic hyperglycaemia", which simply means an elevated blood glucose of unknown cause.

In other words we have a great deal yet to understand about diabetes. Rather than the present overall protocol  approach, further understanding will hopefully lead to a more individualistic approach to people with a high blood sugar.

But what about the economic catastrophe that is predicted? Could diabetes possibly "bankrupt" the NHS? The answer is clearly "No". The UK NHS is strictly cash limited by the government. An increased spending on diabetes will need to be balanced by reduced spending on other aspects of health care, unless the government produces additional funding.

Identifying more people with diabetes does not itself cost any money, apart from to the sufferers who must pay higher insurance premiums and who might experience restrictions on employment. The health costs arise from the regular surveillance of so many people (perhaps 10% of the population) by a doctor, or increasingly by a specialist nurse and a clinical team. The regular assessment includes blood tests, retinal screening, foot inspection, blood pressure control, cholesterol management, pharmaceuticals. These all necessitate recruitment of large numbers of staff, and if part of a national programme these become mandatory. Of course there are continual demands for more staff, especially from pressure groups such as Diabetes UK. 

An increased number of people with diabetes is of great financial benefit to the people who treat diabetes. This is true in a salaried NHS, but even more in a health system in which the physician is paid by item of service. "Patient mongering", increasing the number of patients from previously well population, becomes essential for maintaining and increasing income. As Gale indicates, formal guidelines for the masses is in the financial interests of the providers. We must be careful that health systems are designed for the benefit of the patients rather than for the benefit of the providers of care. 

The management of diabetes is increasingly driven by protocols, rigid guidelines, especially when treatment is nurse-led. This becomes expensive because it is not targeted on those individuals who are at particular risk.

The idea of diabetes detection is that the serious vascular diseases will be reduced by the identification of people with high blood glucose, which would then be followed by intensive and effective treatment. We should expect therefore that if the case-finding is effective, then NHS costs will go down, not up!


It is in the interests of most people that disease incidence should be reduced, and this would include diabetes. Early diagnosis is not prevention. If we wish to reduce the incidence of diabetes and its associations ("complications") then we might try to encourage the consumption of less food. This is generally not a successful venture and it is based on a great oversimplification of body size and shape. An expansion of this subject can be found in the book "Curvology - the origins and power of female body shape" by David Bainbridge.

There is another approach that might be helpful. This is the association between diabetes, obesity and low levels of vitamin D. I will expand on this in a future post.

References:

Simmons RK et al. 
Screening for type 2 diabetes and population mortality over 10 years (ADDITION- Cambridge): a cluster randomised controlled trial. 
Lancet 2012; 380: 1741-48.

Gale EAM. 
Is Type 2 Diabetes a category error? 
Lancet 2013; 381, 1956-57.

Sunset over Baie de la Somme, France, August 23rd 2015