The cholesterol or lipid theory of cardiovascular disease states that dietary cholesterol and saturated fat are the cause of atherosclerosis, which in turn is the cause of cardiovascular disease. This hypothesis was made popular by Ancel Keys in his Seven Countries study, and since then the theory has gained a large following. The main problem with the theory is that it does not stand up to any objective scientific scrutiny, and diligent scientists have exposed it as junk science, modern advances in molecular biology have exposed it as ungrounded, and advances in the understanding of propaganda models have exposed it as a money making scam. The amazing thing is that many professionals still cling onto dietary cholesterol as a probably cause of cardiovascular disease, many through ignorance as their role as low levels repeaters, many through their role as paid agents of shadowy agendas. In this regard, let’s pull apart a propaganda piece from the Ministry of Truth to see why it should never have passed peer review.
The study under scrutiny is a 1981 vintage1, a classic decade for false paradigms and propaganda. The researchers fed healthy subjects three dairy products, skimmed milk, yoghurt or full fat milk, in order to assess their impact on plasma lipoprotein levels. Thirty two high school boys were chosen for the study aged between 16 and 18 years. The study lasted for 5 weeks, and in the first and last week the boys were told to consume their normal diet. The three weeks in the middle of the study, the boys consumed skimmed milk, yoghurt or full fat milk in addition to their normal diet. Diet records were kept for food consumption and were analysed for nutrient intake by computer programme. The authors claimed that the results of the study showed a correlation between the fat content of the dairy product and the change in total cholesterol. In this regard skimmed milk (low fat) was associated with a reduction in total cholesterol, while the yoghurt (medium fat) and cream (high fat) were associated with an increase in total cholesterol.
Problems With This Paper
1. The largest problem with this study is the fact that it is difficult to alter dietary components within a diet without also changing total energy intake. In fact the total energy intake of the boys drinking the skimmed milk was 3518 kcal per day at baseline, but went down to 3346 kcal per day during the experimental three weeks. However, in the yoghurt group total energy intake was 3742 kcal per day at baseline but rose to 4028 kcal per day during the experimental phase. Likewise, the baseline energy intake of the full fat milk group was 3842 kcal per day at baseline and this rose to 4081 kcal per day during the three week experimental period. The authors concluded that no nutrient other than fat was associated with changes in cholesterol levels in the study, which may be true. However, while true, it is certain disingenuous because energy intakes certainly changed and if this was not taken into account, then it is a serious methodological flaw.
2. Diet records were only kept for three days of the week, and of these only two were selected for analysis. Diet records are notoriously inaccurate, and other methods such as the frequent food questionnaire are now used for this purpose because of inherent problems with food diaries (here). Also the analysis of the nutritional intake was done with a computer from estimations in food tables which is also inaccurate. The wisdom of selecting teenage boys to perform this study unsupervised should also be questioned because we all know how reliable teenage boys are, don’t we? Was the reported nutrient intake actually a representation of the actual intake? Very unlikely based on the sloppy methods employed.
3. Another problem with the study was that plasma lipids fell during the initial study week in all groups. This is problematic because the initial week the subjects were supposed to be following their normal diet for a baseline reading. This suggests one of two things. Either the psychological implications of participating in a study caused cholesterol levels of the boys to change, or they subconsciously changed their diet because of the study and the maintaining of food diaries. Based on other evidence, the latter is much more likely, in which case the study is poorly controlled and its data of questionable value.
4. Although serum cholesterol rose initially on both the yoghurt and cream diets, by week four they had returned to baseline. In other words, by week four, despite still consuming the yoghurt and cream supplements, the total cholesterol levels of the subjects were no different from their starting levels (which had actually fallen in the initial week anyway). Therefore it could not be possible for fat or cholesterol to be the cause of the changes to plasma lipoproteins as the dietary intakes had not changed during this time. The authors could quite easily have reported that cholesterol levels fell during a period of high fat and cholesterol intake if they chose to use different semantics because this is what the data showed.
5. The authors measured the plasma levels of the low density lipoprotein (LDL) in this study. However, more is known about this particle now and it shows heterogeneity. In this regard, both large buoyant and small dense particles have been identified. Only the small dense LDL particle shows an association with cardiovascular disease. Studies that have not differentiated between the two particles should therefore be called into question. Likewise total cholesterol is now thought to be a poor predictor of cardiovascular risk, More important is the high density lipoprotein (HDL) to LDL ratio, based on the observation that high HDL plasma levels appear to reduce the risk of cardiovascular disease. The HDL levels of the subjects rose in all treatment groups in this study, suggesting a protective effect of dairy consumption of cardiovascular risk.
6. Dairy products contain substances known to alter fasting insulin levels. The likely candidate is conjugated linoleic acid (CLA) a fatty acid that may improve insulin sensitivity. This factor was not controlled for and so it is not possible to conclude that the fat content of the dairy products was responsible for the effects seen. Also the calcium intake of all subjects doubled during the study. Calcium may reduce fat absorption because it can form insoluble soaps with fatty acids in the gut. Interactions between these nutrients may therefore have affected the results and added complexity to their interpretation. In a similar vein, yoghurt can delay gastric emptying rates and alter the digestive process, which affects insulin levels. As postprandial glycaemia and insulin levels are known modifiers of plasma lipoprotein levels, this is another factor that should have been controlled for.
8. A wealth of data now supports the contention that full fat dairy products as anti-obesogenic foods. Feeding full fat dairy to subjects has been shown to cause weight loss, and those who eat the most dairy have the lowest body weights. Because body weight is one of the single strongest determinant of cardiovascular risk, it is very unlikely that dairy foods are a causative factor in cardiovascular disease.
9. The yoghurt was administered as a drink. However no details are given about the composition of this drink in terms of other ingredients. If other factors were added to the yoghurt in the creation of this drink (such as sugar), these could have been responsible for the effects seen.
10. The subjects lost weight in the initial study week, when dietary intake was supposed to be of the subjects normal pattern. Why was this the case and what effects did it have on the subsequent study week cholesterol levels?
Dr Robert Barrington’s Comments: The conclusions of the authors were that skimmed milk appears to ‘have advantages over yogurt and full cream milk in populations that have dietary hyperlipidemia’. This conclusion in itself demonstrates the bias and subjectivity in every nuance of the authors dialectic, as no where in their experiment did they test anything on subjects with hyperlipidaemia. The conclusions should have read ‘these results show that supplementary dairy products of differing fat levels have complex interactive effects on the plasma lipoprotein levels of healthy boys and the inconsistent changes make drawing conclusions difficult’. Just how papers like this manage to pass peer review and become published is a wonder. The authors even managed to include a sly dig at eggs, just to show their pay masters they were on script.
RdB
