Tuesday, 16 January 2018

High blood cholesterol gives a survival advantage

High cholesterol is good

Previous Posts have shown that the existing dogma that "a high level of cholesterol in the blood is bad" is only true in men below the age of 50 years. In women of this age, the blood level of cholesterol is of no value in determining coronary risk or life expectancy.

The most important source of information is the 30 year follow-up of the observational US Framingham study, published in 1987. The conclusion (effectively ignored) is as follows:

The study did not extend beyond the age of 70 years and so the conclusion is limited. However other studies have shown that beyond this age a high cholesterol level is associated with the best survival. There are three important studies showing this that I have reviewed previously: the Paris study, the Honolulu study, and the Stamford (USA) study. There are several other studies showing the same thing. 

New evidence

Now there is a further study, this from the Aging Research Centre at the Karolinska Institute – Stockholm University.

Serum total cholesterol and risk of cardiovascular and non-cardiovascular mortality in old age: a population-based study 
Liang et al. BMC Geriatrics (2017) 17:294 DOI 10.1186/s12877-017-0685-z 

The study involves the observation of 3090 adults aged 60 years or above, from a population cohort. The average length of follow-up was 7.2 years, that is 23,196 person-years of observation. During this time 1059 participants died, 34.3%.

Compared with those who were still alive at the end of follow-up, and after controlling for age, those who died during follow-up were:

  • older, 
  • less likely to have a university education, 
  • less likely to be former smokers or current smokers (surprising), 
  • more likely to be physically inactive, 
  • more likely to have a lower level of total cholesterol, 
  • more likely to have diabetes, 
  • more likely to have cognitive impairment, 
  • more likely to have mobility limitation. 

Whether or not the participants died or did not die did not differ significantly: 

  • in the male/female proportion,
  • in the prevalence of heavy alcohol drinking, 
  • in the prevalence of obesity, 
  • in the prevalence of hypertension,
  • in use of cholesterol-lowering  medications. 
For the purpose of survival analysis all the participants were divided into three groups based on the blood level of cholesterol at the onset of the study: 

  • less than 5.18mmol/L, 
  • 5.18 to 6.21, 
  • greater than 6.21.

The results are displayed in Figure 1.

Figure 1. All cause mortality based on blood cholesterol level

It is clear in Figure 1 that the highest mortality rate (71.8 per 1000 person-years) is in those with the lowest blood levels of cholesterol, less than 5.18 mmol/L. The lowest mortality rate (35.6 per 1000 person-years) is in those with the highest blood levels of cholesterol, half the death rate than in those with the lowest cholesterol. Why is this not publicised?.

Statin effect

During the past 20 years an increasing number of the population have been commenced on long-term cholesterol-lowering therapy, and especially statins. This has added a complication to cholesterol-related observation studies, but it is good to see that such studies continue, demonstrating an academic scepticism of the often-stated dogma that “The lower cholesterol the better”. My recent Post of evolocumab shows that this simply is not true, but it is as well that other people think the same as I do.

An important point is that a low level of cholesterol in an individual can be a “natural” level, a person-characteristic, or it can be the result of statin or other cholesterol-lowering medication in someone with a naturally higher level.

In the Stockholm study the participants are divided into whether or not they take statins (and other cholesterol-lowering medications), but statin intervention was not part of the study: it was purely observational.

In the results we see mortality displays for all the participants (as in Figure 1), those not taking cholesterol-lowering medications, and those who during the study were taking cholesterol-lowering medications (mainly statins).

We also see analysis of all-cause mortality, cardio-vascular disease (CVD) mortality, and non cardio-vascular disease mortality.

All cause mortality

Figure 2 shows all-cause mortality, the most important outcome measure.
Figure 2. Mortality from all causes related to blood cholesterol level

We can see a similar pattern to Figure 1, but it is a little more complicated.  The blue columns, all participants, are as in Figure 1. But we see in addition the division between those not taking statins (green) and those were taking statins (yellow).

The first group of three columns concerns those with low cholesterol (<5.18). We can see that those not taking statins (green column) have the highest mortality. This is a natural low cholesterol and it is clear that it gives a distinct survival disadvantage – it is a bad thing.

But the yellow column, those taking statins, does not really tell us very much. These participants would have a natural cholesterol significantly higher, but we do not have the information, we do not know the pre-statin cholesterol level. 

We can see in the second group (5.18–6.21) and the third group (>6.21) that having a higher cholesterol is associated with a lower mortality rate, a survival advantage, compared to the first group. The participants in the yellow column of first group (those taking statins) would have a similar natural cholesterol level to the second and third groups.

Cardiovascular mortality

In Figure 3 we can see the same analysis but looking at death from cardio-vascular disease (CVD) – strokes and heart attacks.

Figure 3. Cardiovascular mortality related to blood cholesterol level

The pattern is the same. Those with the lowest blood level of cholesterol have the highest mortality, the opposite of what we have been told. Those taking statins (yellow) in the first group would have a natural cholesterol as in the higher cholesterol groups, and therefore have a similar lower mortality rate. 

Non-cardiovascular mortality

Finally we see in Figure 4 mortality due to to causes other than cardio-vascular disease. Such conditions would be cancers or  pneumonia. 

Figure 4. Non-cardiovascular mortality related to blood cholesterol level
The pattern is the same, but course the number of deaths is lower in the subgroups than in the total numbers shown in Figure 2.

Low cholesterol is not a good thing, but if the low cholesterol is because of taking statins, then the mortality rate is similar to those with higher levels of cholesterol.


This is another good quality study that demonstrates clearly, and contrary to what we are generally told, that a high cholesterol level in the blood is a good indicator for a longer than average life expectancy.

Why are we not told this? The information has been available for many years but it has been suppressed. 

There are so many vested interests in perpetuating the myth that cholesterol is effectively poisonous, killing us in proportion to its level in the blood. This is the basis of many academic careers and the rationale of cholesterol-lowering medications, especially statins. 

The self-appointed and powerful Cholesterol Treatment Trialists’ (CTT) Collaboration controls public and government policy, and falsely perpetuates the identity of cholesterol as the "cause" of heart disease and premature death. 

We can see that this is wrong.

In a person over the age of 60 years, a high blood cholesterol is clearly a good thing. Correspondingly a low blood cholesterol level gives a serious survival disadvantage. However this is only true if the low cholesterol level is "natural". 

If low blood cholesterol is the result of statin or other cholesterol-lowering medication, then there is no survival disadvantage and there is no need for concern.

We can conclude that it is not really the cholesterol level of the blood that determines future health and survival, but what we might regard as "the constitution" of the individual that is responsible that natural cholesterol level.

Friday, 5 January 2018

PURE study - vegetables are good

PURE - Fruit, Vegetables, and Legumes 

We have already looked at the large multinational Prospective Urban Rural Epidemiology (PURE) study. The objective of the study was investigate the dietary advice to which we have been subjected during the past half-century. 

The first paper concentrated on fats and carbohydrate. The conclusion was that dietary fat advice has been wrong: a higher proportion of the fat intake of the diet is associated with a lower rate of total deaths, cardiovascular disease (CVD) events, and CVD death. A higher proportion of  carbohydrate intake is associated with worse health outcomes.

Reducing total food intake might be necessary weight control, but reduction of the proportion on calories from fat, although encouraged officially, is not correct.

The second PURE publication in The Lancet is devoted to the relationship between health outcomes and intakes of fruit, vegetables and legumes. 135,335 individuals from 18 countries were included in this analysis. The countries included 3 high-income (Canada, Sweden, and United Arab Emirates), 11 middle-income (Argentina, Brazil, Chile, China, Colombia, Iran, Malaysia, occupied Palestinian territory, Poland, South Africa, and Turkey) and 4 low-income countries (Bangladesh, India, Pakistan, and Zimbabwe). 

The investigation of fats and carbohydrate in the first paper looked at the relative proportions of each in the diet. Inevitably therefore, if the proportion of fat was reduced (on advice), then the proportion of carbohydrate must increase, and vice versa.

The second paper is different in this respect. It looks at the consumption of fruit, vegetables and legumes in absolute measures, the actual amount consumed. 

Because of the large number of participants in the study from many countries, it was not possible to measure the amounts in mass units (eg grams). The measurement used was simply the number of portions per day, recognising that the size of a protion would vary from one person to another. 


During a median follow­-up of 7·4 years, there were 4784 major cardiovascular disease events. Greater fruit, vegetable, and legume intake was associated with a lower risk of major cardiovascular disease events.
Overall, higher total fruit, vegetable, and legume intake was inversely associated with major cardiovascular disease, myocardial infarction, cardio­-vascular mortality, non-­cardiovascular mortality, and total mortality when adjusted for age.
Life expectancy, heart disease, and consumption of fruit, vegetables and legumes
Figure 1
The overall results can be seen in Figure 1 above, a rainbow graph that can be broken down into its constituent parts.

For example, in Figure 2 we see just total deaths. 

Figure 2
The group with the lowest intake of fruit, vegetables and legumes (<1 portions per day), we see a relatively high death rate from total deaths (all causes) of 8% during the study period, compared to 3% in those who consume more than 3 portions per day.

If we look again at Figure 1, we see the pale blue bars which represent cardiovascular events. We can see that there are 4% cardiovascular events in the consuming up to 5 portions of fruit, vegetables and legumes per day, compared to 3% in those in the groups consuming more than  5 portions per day.

All morbidity and mortality events are higher in this with the lowest consumption of fruit, vegetables and legumes.

It is interesting to look at the geographical variations of consumption of fruit, vegetables and legumes. The total amounts can be seen in Figure 3.

Figure 3 Geographical variation of consumption of fruit, vegetables and legumes
The highest consumption is in the Middle East, and this is due to a particularly high consumption of fruit.
It is also found that people who consumed more fruits, vegetables and legumes had higher education, higher levels of physical activity, lower rates of smoking, and higher energy, red meat and white meat intake, and were more likely to live in urban areas. 


A greater fruit, vegetable and legume intake is associated with a lower risk of major cardiovascular disease, myocardial infarction, cardiovascular mortality, non­-cardiovascular mortality, and total mortality in the analyses adjusted for age 
In this study,  3·2 servings is equivalent to 400 g of fruit, vegetables and legumes per day. Many dietary guidelines in North America and Europe recommended intake ranging from 400 to 800 g/day, but for most individuals in poorer countries in other continents these targets are unaffordable .

The study indicates that even three servings per day (375 g/day) show benefit against non­-cardiovascular and total mortality, and indicate that optimal health benefits can be achieved with a more modest level of consumption, an approach that is likely to be much more affordable. 


The PURE study is observational and it does not try to investigate the possible ways in which diet might influence human health.  It is a public health investigation, to look for empirical ways in which diet can be recommended to improve health and survival. A high fruit, vegetable, and legume diet can be encouraged without understanding the mechanism of benefit.
Several mechanisms have been proposed to explain the lower risk of cardiovascular disease with higher consumption of fruits, vegetables, and legumes. This would include cereals as a diet high in fibre has been related to lower CHD risk, but there is no conclusion as to mechanism.