This is a continuation of Blog 6 and provides further information which challenges the conventional wisdom that the cholesterol in the blood(TC) is an indication of the risk of developing heart disease and that if the TC is lowered there will be a corresponding reduction in that risk.
The results of a major investigation in which 5201 men and women recruited from 4 different centres in the US were monitored for an average of 4.8 years during which time there were 646 deaths are shown in Table 1. It is evident that the higher the LDL cholesterol (the so-called “bad” one) the lower the death rate(1).
TABLE 1 LDL CHOLESTEROL AND DEATH RATE
|LDL Cholesterol,mmol/L (mg/dl)||Death rate per 1000 person-years|
In the Honolulu Heart Programme the TC was measured between 1965 and 1968 in 7961 men of Japanese origin who were born between 1900 and 1919. During the period of this study the men were monitored on 3 occasions and there were 2072 deaths. Details of cause of death was recorded and then related to the values which had been obtained for cholesterol. The results (Table 2) confirm that in this group of men the death rate attributed to coronary heart disease is directly associated with the TC level. However for deaths from stroke there are high death rates for both high and low cholesterol values. For cancer, it is evident that the death rate is inversely related to the TC levels. In this study there were twice as many deaths due to cancer at low TC levels when compared with whose TC was above 6.98 mmol/L. For all-cause mortality the lowest death rates were found in the cholesterol range between 4.65 and 6.18 mmol/L (180 and 239 mg/100ml). So there is very good agreement with the MRFIT results shown above. The authors concluded that manipulation of TC levels below the range shown above would not be desirable if it was to result in an increased risk of death from cancer and other diseases(2).
TABLE 2 RELATIONSHIP BETWEEN DEATH RATE AND TOTAL CHOLESTEROL FOR VARIOUS CAUSES OF DEATH
|AGE ADJUSTED MORTALITIES PER 1,000 PERSON-YEARS|
|>6.98 (>269 )||4.87||1.74||3.83||17.31|
A fourth examination was conducted in 1991-1993 when there had been a further 72 deaths. The results are shown in Table 3. Once again it is evident that the highest mortality is observed in those with a low TC level over a 20-year period. However it is apparent that as the participants got older those with a high TC have a much lower death rate than those with the lowest value for blood cholesterol. It is difficult to understand what benefits would be derived if these people were treated to lower their TC levels(3).
TABLE 3 ALL-CAUSE MORTALITY AND CHOLESTEROL (HONOLULU HEART STUDY)
|Mean TC, Mmol/L (mg/100ml)||Relative Mortality Rate|
In a study conducted in New Haven, Connecticut, 977 men and women over 70 years old were followed for 4 years and all deaths recorded and related to TC (Table 4). For deaths from heart disease, the death rate at a TC lower than 5.2 mmol/l was more than double that with a TC value higher than 6.2 mmol/l. For all-cause mortality in women the lowest death rate was observed in those with the highest TC values. In men the lowest death rate was in those with TC values in the range 5.2-6.2 mmol/l.
TABLE 4 DEATH RATES AND BLOOD CHLOESTEROL CONCENTRATION
The authors concluded the results:
“do not support the hypothesis that hypercholesterolemia or low HDL-C are important risk factors for all-cause mortality, coronary heart disease mortality, or hospitalization for myocardial infarction or unstable angina in this cohort of persons older than 70 years”(4)
Essentially similar conclusions arise from a study done by Joel Kauffman who analysed data from 2,227 residents in New York with average age 76 years. About 1/5 of them were taking statin drugs to lower TC and LDL cholesterol. The results in Table 5show that the death rates in those with the lowest levels of TC and LDL cholesterol were found to be approximately twice as high as those with high values.
Kauffman concluded that:
“ ……. high TC and LDL.C levels are beneficial, certainly in the elderly who are most likely to be given a statin drug. The recent emphasis on the value of lowering LDL.C rather than lowering TC, promoted by health authorities and drug companies in the last few years, is invalidated by this study.”(5)
TABLE 5 RELATIONSHIP BETWEEN DEATH RATE AND CONCENTRATIONS OF TOTAL CHOLESTEROL AND LDL CHOLESTEROL IN BLOOD
|TC, Mmol/L,(mg/100ml)||No of subjects||Deaths(%)||Relative Death Rate|
|< 4.5 (175)||580||97 (16.7)||1.8|
|4.5-5.2 (176-199)||574||78 (13.6)||1.2|
|5.2-5.8 (200-226)||556||57 (10.3)||0.9|
|> 5.8 (226)||567||59 (10.4)||1.0|
|< 2.53 (97.8)||572||90 (15.7)||2.0|
|2.53-3.12 (97.9-120.6)||568||83 (14.5)||1.6|
|3.12-3.72 (120.7-144)||571||65 (11.4)||1.2|
|>3.72 (144)||566||53 (9.4)||1.0|
A recent study in Norway which involved much larger numbers has followed 52,087 men and women aged 20-74 years over a 10-year period. TC levels were measured and details of any deaths which occurred were recorded. The results are shown in Table 6. In both men and women there was no statistically significant increase in the risk of death at higher TC levels. Individuals with a TC of 7.0 mmol/L(298 mg/100ml) or higher were no more likely to die of cardiovascular disease than those with levels below 5.0 mmol/L(194mg/100ml).
In men, there was no increase in the all-cause mortality with raised TC. Those with a TC level between 5.0 and 5.9 mmol/L had the lowest death rate, which was 23% lower than those with a TC below 5.0 mmol/L.
For women the pattern is different. The higher the TC, the lower the risk of dying from all causes. Compared with those with a TC below 5.0 mmol/L , those with the highest TC levels were 28% less likely to die from all causes.
The authors commented as follows:
‘’If our findings are generalizable, clinical and public health recommendations regarding the ‘dangers’ of cholesterol should be revised. This is especially true for women, for whom moderately elevated cholesterol (by current standards) may prove to be not only harmless but beneficial.’’
They went on to conclude:
‘’Our results contradict the guidelines’ well-established demarcation line (5 mmol /L) between
‘good’ and ‘too high’ levels of cholesterol. They also contradict the popularized idea of a positive, linear relationship between cholesterol and fatal disease. Guideline-based advice regarding CVD prevention may thus be outdated and misleading, particularly regarding many women who have cholesterol levels in the range of 5–7 mmol/Litre and are currently encouraged to take better care of their health’’(6).
TABLE 6 VARIATION IN MORTALITY DUE TO ALL-CAUSES AND TO CARDIOVASCULAR DISEASES(CVD) WITH TC IN MEN AND WOMEN (NORWAY)
|TC, Mmol/L (mg/100ml)||Men||Women|
In a subsequent paper the research team provided the information for the different age ranges (Table 7).
This detailed break-down shows that as expected most deaths occur after the age of 60 years. Although the optimum TC level for men aged 60-69 is in TC range 5.0 to 5.9 mmol/L , for those over 70 the lowest death rate is in the higher TC level of 6.0 to 6.9 mmol/L. For women it is very clear that the death rate for the over 60s decreases as the TC increases. For this age range it is evident that the highest death rates are for those with a TC level which is below 5.9 mmol/L . The relatively high death rates for those aged 60+ years at low TC values should also be noted(7).
TABLE 7 ALL-CAUSE MORTALITY RATES (PER 1000 PERSON-YEARS) AND TC LEVELS FOR AGE RANGES INTHE NORWEGIAN STUDY
|TC LEVELS, mmol/L(mg/100ml)|
According to the NHS Choices website TC should be less than 5.0 mmol/L and LDL cholesterol should be less than 3.0 mmol/L. All the information quoted above demonstrates that for people over 60 years old these levels are associated with an increase in all-cause mortality. For older people it is evident that the higher the TC value the lower the death rate. It is difficult to understand why the relevant public policy places so much emphasis on cholesterol reduction.
1. P Fried et al (1998) Journal of the American Medical Association 279 (8) pp585-592.
2. G.Stemmerman et al (1991) Archives of Internal Medicine 151 pp.969-972.
3. I.J. Schatz et al (2001) Lancet 358 9729 pp. 351-355.
4. H Krumholz (1994) 272 (17) pp1335-1340.
5. Joel Kauffman (2007) Journal of American Physicians and Surgeons 12 (1) p7.
6. H Petursson et al (2012). Journal of Evaluation in Clinical Practice 18 (1) pp 159-168.
7. H Petursson et al(2012). Journal of Evaluation in Clinical Practice 18 (1) pp 170-171.