OBJECTIVE

The cholesterol-raising effect of boiled coffee is caused by diterpenes from coffee oil. In order to identify the diterpene responsible, we analysed the diterpene composition of oils from Arabica (Coffea arabica) and Robusta (Coffea canephora robusta) beans and their effects on serum lipids and thyroid function.

METHODS

During the first 3-week period of a randomized, cross-over trial, 11 healthy, normolipaemic volunteers received per day either 2 g of coffee oil (n = 5) or placebo oil (n = 6). After a 2-week wash-out, the reverse treatments were applied for another 3 weeks. Six subjects received Arabica oil, supplying 72 mg day-1 cafestol and 53 mg day-1 kahweol, and five received Robusta oil, which provided 40 mg of cafestol, 19 mg of 16-O-methyl-cafestol, and 2 mg of kahweol per day. Background diets were constant.

RESULTS

The average serum cholesterol levels rose by 0.65 mmol L-1 (13%) on Arabica oil (P < 0.025; 95% CI, 0.21-1.09 mmol L-1) and by 0.53 mmol L-1 (13%) on Robusta oil (NS; 95% CI -0.36-1.42 mmol L-1). The triglycerides levels rose by 0.54 mmol L-1 (71%) on Arabica (P < 0.005; 95% CI, 0.22-0.76 mmol L-1) and 0.49 mmol L-1 (61%) on Robusta oil (P < 0.005; 95% CI, 0.30-0.68 mmol L-1). None of the effects on serum lipids or lipoprotein cholesterol levels was significantly different between Arabica and Robusta oil. Concentrations of serum total and free thyroxine (T4), triiodothyronine (T3), and thyroid-stimulating hormone (TSH) were largely unaffected.

CONCLUSIONS

Both Arabica and Robusta oil elevated serum lipid levels; therefore, cafestol must be involved and kahweol cannot be the sole cholesterol-raising diterpene. The mode of action of coffee diterpenes does not involve induction of hypothyroidism.