Effects of partially hydrogenated fish oil, partially hydrogenated soybean oil, and butter on serum lipoproteins and Lp[a] in men.J Lipid Res. 1995 Jun; 36(6):1370-84.JL
We have compared the effects of partially hydrogenated fish oil (PHFO-diet), partially hydrogenated soybean oil (PHSO-diet), and butterfat (butter-diet) on serum lipids and lipoprotein[a] in 31 young men. The three test margarines, which contributed 78% of total fat in the diets, were produced from 70% of butterfat, PHSO, or PHFO, each with 30% of soybean oil. Fat provided about 35% of energy, and trans fatty acids 0.9%, 8.5%, and 8.0% of energy in the butter-, the PHSO-, and the PHFO-diet, respectively. Dietary cholesterol was balanced by the addition of dried egg powder to the PHSO- and the butter-diet; thus all diets contained 420 mg dietary cholesterol per 10 MJ per day. The subjects consumed all three test diets for 19-21 days in a random order (crossover design). The serum levels of total and LDL-cholesterol were significantly elevated on the PHFO-diet (mean values 5.42 and 3.94 mmol/L, respectively) compared to the PHSO-diet (5.11 and 3.58 mmol/L, respectively) but not different from those on the butter-diet (5.32 and 3.81 mmol/L, respectively). LDL-cholesterol was significantly reduced on the PHSO-diet compared to the butter-diet. The level of HDL-cholesterol was significantly lower on the PHFO-diet (0.98 mmol/L) when compared to the butter-diet (1.05 mmol/L) and with border-line significance compared to the PHSO-diet (1.05 mmol/L). The ratio of LDL- to HDL-cholesterol was significantly higher on the PHFO-diet (4.20) when compared to both other test diets (3.85 and 3.65, respectively). No significant differences in triglyceride values were observed. Lp[a] increased and apoA-I decreased significantly after consumption of both the PHSO-diet and the PHFO-diet, compared to the butter-diet. In conclusion, our results indicate that consumption of PHFO may unfavorably affect lipid risk indicators for coronary heart disease at least to the same extent as butterfat. To what extent the observed effects are due to the content of monoene trans, diene trans, or to the long chain saturated fatty acids in PHFO remains to be elucidated.