Cafestol and kahweol, coffee lipids present in unfiltered coffee brews, potently increase LDL-cholesterol concentration in human subjects. We searched for an animal species in which cafestol similarly increases LDL-cholesterol. Such an animal model could be used subsequently as a model to study the mechanism of action of cafestol and kahweol. Cafestol and kahweol increased serum lipids in African green monkeys (Cercopithecus aethiops), cebus (Cebus apella) and rhesus (Macaca mulatta) monkeys, hamsters, rats and gerbils differently from the increase in human subjects. In African green monkeys, the rise in total cholesterol was less pronounced than that in human subjects. In addition, the increase in total cholesterol was predominantly due to a rise in HDL-cholesterol rather than LDL-cholesterol. Thus, the rise in plasma lipids might illustrate the mechanism in these monkeys rather than the mechanism in human subjects. In other animal species, cafestol and kahweol did not raise cholesterol consistently. The variability in effects on serum lipids could not be explained by the mode of administration or dose of diterpenes, nor by the amount of cholesterol in the diet. In conclusion, we did not find an animal model in which cafestol and kahweol elevate plasma lipoproteins to the same extent as in human subjects. For the time being, therefore, studies on the mechanism of action should be done preferably in human subjects.

The aim of the present study was to examine the effect of reducing saturated fat in the diet, or partly replacing it with unsaturated fat, on the serum lipoprotein profile of human subjects. The study had two intervention periods, 8 weeks (phase 1) and 52 weeks (phase 2). In phase 1, total fat was reduced from 31 to 25 % energy (polyunsaturated fatty acids (PUFA):saturated fatty acids (SFA) ratio increased from 0.2 to 0.4) by reducing the quantity of coconut fat (CF) in the diet from 17.8 to 9.3 % energy intake. In phase 2, subjects were randomised to groups A and B. In group A total fat was reduced from 25 to 20 % energy (PUFA:SFA ratio increased from 0.4 to 0.7) by reducing the quantity of CF in the diet from 9.3 to 4.7 % total energy intake. In group B, the saturated fat content in the diet was similar to group A. In addition a test fat (a mixture of soyabean oil and sesame oil, PUFA:monosaturated fatty acids ratio 2) contributed 3.3 % total energy intake and total fat contributed 24 % energy intake (PUFA:SFA ratio increased from 0.7 to 1.1). At the end of phase 1, there was a 7.7 % reduction in cholesterol (95 % CI -3.6, -12.2) and 10.8 % reduction in LDL (95 % CI -4.9, -16.5) and no significant change in HDL and triacylglycerol. At the end of phase 2, the reduction in cholesterol in both groups was only about 4 % (95 % CI -12, 3.2) partly due the concomitant rise in HDL. The reduction in LDL at 52 weeks was significantly higher in group B (group A mean reduction 11 %, 95 % CI -20.1, -2.0 and group B mean reduction 16.2 % 95 % CI -23.5, -8.9). In phase 2, triacylglycerol levels showed a mean reduction of 6.5 % in group 2A and a mean increase of 8.2 % in group 2B. The reduction of saturated fat in the diet is associated with a lipoprotein profile that would be expected to reduce cardiovascular risk. The reduction of dietary saturated fat with partial replacement of unsaturated fat brings about changes in total cholesterol, HDL- and LDL-cholesterol that are associated with a lower cardiovascular risk.