Pedersen et al. (Reference Pedersen, James and Brouwer1) express concern that recently published research had downplayed the importance of SFA consumption as a risk factor for CHD(Reference Siri-Tarino, Sun and Hu2). Their main argument is that prospective cohort studies are unreliable. There are of course uncertainties in such studies, but it is difficult to ignore that more than thirty cohort studies have shown that patients with CVD did not eat more SFA than had heart-healthy people; in six of them(Reference Takeya, Popper and Shimizu3–Reference Iso, Sato and Kitamura8), stroke patients had actually eaten less.
To make their case, Pedersen et al. presented a small and biased subset of ecological studies apparently linking reduced consumption of SFA to a low incidence of CHD. However, they neglected to mention the many ecological studies that have documented findings from groups with a high consumption of SFA, but with low rates of CHD, including Masai people(Reference Mann, Spoerry and Gary9), French(Reference Renaud and de Lorgeril10), Italian-Americans(Reference Stout, Marrow and Brandt11) and Polynesians(Reference Prior, Davidson and Salmond12). They also claim that the association between the decline of CHD mortality in Finland and the lowered intake of SFA was causal. However, the decline began in North Karelia 3 years before the start of the cholesterol campaign, and it occurred also in the districts where no advice was given(Reference Salonen, Puska and Mustaniemi13).
Pedersen et al. asserted that SFA with twelve to sixteen carbon atoms are the most potent LDL- and total cholesterol-raising fatty acids. However, other researchers reported that the serum content of these fatty acids is inversely associated with serum cholesterol(Reference Samuelson, Bratteby and Mohsen14), and in seven studies, the content of twelve to sixteen carbon fatty acids in the blood or the fat cells was similar or lower in patients with acute CHD than in healthy people(Reference Kark, Kaufmann and Binka15–Reference Clifton, Keogh and Noakes21). The content of certain SFA in the serum reflects the intake of dairy fat(Reference Smedman, Gustafsson and Berglund22, Reference Wolk, Furuheim and Vessby23), and such intake is inversely associated with BMI, waist circumference, ratio of LDL:HDL and fasting glucose concentration, and positively associated with HDL and apoA-I(Reference Smedman, Gustafsson and Berglund22–Reference Brevik, Veierod and Drevon25). In accordance, a meta-analysis of twenty-five cohort studies showed that the lowest total mortality, cardiovascular incidence and mortality, and incidence of diabetes were seen among those with the highest intake of dairy fat(Reference Elwood, Pickering and Givens26).
Pedersen et al. endorse the many reports emphasising the importance of increasing the intake of PUFA. This advice is not based on randomised, controlled dietary trials, because no such trial has ever succeeded in lowering cardiovascular or total mortality by exchanging SFA with PUFA(Reference Ravnskov27). Rather, the advice is based on statistical calculations using data from unreliable cohort studies. Pedersen et al. refer to a meta-analysis of such trials, the authors of which claimed benefit, but they had excluded two trials, where CHD mortality had increased in the treatment groups(Reference Rose, Thomson and Williams28, Reference Woodhill, Palmer and Leelarthaepin29), and included a trial where a decreased risk was seen only in the participants who increased their intake of fish(Reference Watts, Lewis and Brunt30), and also the Finnish Mental Hospital Study(Reference Turpeinen, Karvonen and Pekkarinen31), a trial which does not satisfy the quality criteria for a correctly performed randomised controlled trial. A reduction of SFA was part of the intervention in three multifactorial trials, but these trials were unsuccessful as well(32–34); in one of these, total mortality was twice as high in the treatment group(Reference Miettinen, Huttunen and Naukkarinen33).
Numerous studies on laboratory animals and human subjects have also shown that an increased intake of PUFA, in particular of the n-6 type, is associated with many adverse health effects such as allergy, asthma, immunosuppression, decreased fertility, pre-eclampsia, encephalopathy and cancer(Reference Dam and Søndergaard35–Reference Leitzmann, Stampfer and Michaud41). In accordance with this, Israeli Jews have a high intake of the ‘recommended’ n-6 type of PUFA (from grains and soyabean oil), and they exhibit a high incidence of cancer and CHD mortality compared with other Western countries(Reference Yam, Eliraz and Berry42).
In conclusion, Pedersen et al. do not provide sufficient evidence to implicate SFA in CHD risk. There is increasingly strong evidence that SFA are not involved(Reference Siri-Tarino, Sun and Hu2, Reference Rose, Thomson and Williams28, Reference Oddy, de Klerk and Kendall43–Reference Volek and Forsythe47).