[Negligible amounts of cholesterol-raising diterpenes in coffee made with coffee pads in comparison with unfiltered coffee]. / Verwaarloosbare hoeveelheden cholesterolverhogende diterpenen in koffie gezet met de koffiepadzetmethode bij vergelijking met ongefilterde kookkoffie.

OBJECTIVE: To determine the amounts of the serum-cholesterol raising diterpenes cafestol and kahweol in coffee made with coffee pads and the Senseo coffee machine as opposed to filtered and unfiltered coffee. DESIGN: Observational. METHOD: In five cities in the Netherlands coffee was purchased in three major supermarkets resulting in a total of 30 samples of coffee pads. The levels of cafestol and kahweol were determined by gas chromatography. As controls, the diterpene levels in filtered and unfiltered coffee were also measured. RESULTS: Coffee prepared using coffee pads contained on average 0.76 mg/l cafestol (95% CI: 0.69-0.82) and 0.85 mg/l kahweol (95% CI: 0.77-0.94). Filtered coffee contained 0.76 mg/l cafestol (95% CI: 0.63-0.88) and 0.81 mg/l kahweol (95% CI: 0.63-0.99). Unfiltered coffee contained 72.5 mg/l cafestol (95% CI: 48.5-96.4) and 71.5 mg/l kahweol (95% CI: 45.0-98.1). CONCLUSION: Coffee prepared using coffee pads and the Senseo coffee machine contained minute levels of diterpenes comparable to those of filtered coffee. Its effect on serum-cholesterol levels is therefore likely to be negligible.

Absorption and urinary excretion of the coffee diterpenes cafestol and kahweol in healthy ileostomy volunteers.

OBJECTIVES: To determine the absorption and urinary excretion of the cholesterol-raising coffee diterpenes cafestol and kahweol in humans. SUBJECTS AND DESIGN: Nine healthy ileostomists consumed a dose of one, two or three cups of French-press coffee together with a standardized breakfast on three separate days in random order. Subsequently, ileostomy effluent was collected for 14 h and urine for 24 h. Stability of cafestol and kahweol was also assessed under simulated gastrointestinal tract conditions. MAIN OUTCOME MEASURES: Absorption of diterpenes, stability of diterpenes during incubation with gastrointestinal fluids, and urinary excretion of diterpenes. RESULTS: Corrected mean absorptions expressed as percentages of the amount consumed and the amount entering the duodenum were 67 and 88%, respectively, for cafestol, and 72 and 93%, respectively, for kahweol. We found losses of diterpenes during incubation in vitro with gastric juice (cafestol, 24%; kahweol, 32%), during storage with ileostomy effluent (cafestol, 18%; kahweol, 12%), and during freeze-drying (cafestol, 26%; kahweol, 32%). Mean excretion of glucuronidated plus sulphated conjugates in urine was 1.2% of the ingested amount for cafestol and 0.4% of the ingested amount for kahweol. CONCLUSIONS: About 70% of the ingested cafestol and kahweol is absorbed in ileostomy volunteers. Possibly, undetected metabolites are present in ileostomy effluent, resulting in lower absorption percentages. Only a small part of the diterpenes is excreted as a conjugate of glucuronic acid or sulphate in urine. Therefore, these compounds are extensively metabolized in the human body.

Separate effects of the coffee diterpenes cafestol and kahweol on serum lipids and liver aminotransferases.

The coffee diterpene cafestol occurs in both robusta and arabica beans. It is present in unfiltered coffee brews and raises serum concentrations of cholesterol, triacylglycerols, and alanine aminotransferase in humans. The effects are linear with the cafestol dose. Unfiltered coffee also contains the related compound kahweol, which occurs only in the major coffee strain arabica. The activity of kahweol is unknown. In a randomized, double-blind crossover study, we gave 10 healthy male volunteers either pure cafestol (61-64 mg/d) or a mixture of cafestol (60 mg/d) and kahweol (48-54 mg/d) for 28 d. Relative to baseline values, cafestol raised mean (+/-SEM) total serum cholesterol concentrations by 0.79 +/- 0.14 mmol/L (31 +/- 5 mg/dL), low-density-lipoprotein (LDL) cholesterol by 0.57 +/- 0.13 mmol/L (22 +/- 5 mg/dL), fasting triacy-glycerols by 0.65 +/- 0.12 mmol/L (58 +/- 11 mg/dL), and alanine aminotransferase by 18 +/- 2 U/L (all P < 0.01). Relative to cafestol alone, the mixture of cafestol plus kahweol increased total cholesterol by another 0.23 +/- 0.16 mmol/L (9 +/- 6 mg/dL) (P = 0.08), LDL cholesterol by 0.23 +/- 0.16 mmol/L (9 +/- 6 mg/dL) (P = 0.09), triacylglycerols by 0.09 +/- 0.10 mmol/L (8 +/- 9 mg/dL) (P = 0.20), and alanine aminotransferase by 35 +/- 11 U/L (P = 0.004). Thus, the effect of cafestol on serum lipid concentrations was much larger than the additional effect of kahweol, and the hyperlipidemic potential of unfiltered coffee mainly depends on its cafestol content. Both cafestol and kahweol raised alanine aminotransferase concentrations, and their hyperlipidemic effect thus seems not to be coupled with their effect on liver cells.

Quantification of oxysterols in Dutch foods: egg products and mixed diets.

A sensitive and specific method is described for quantifying various cholesterol oxidation products in foodstuffs, including 7 beta-hydroxycholesterol, cholesterol-alpha-epoxide, cholestane-triol, 7-ketocholesterol and 25-hydroxycholesterol. A chloroform-methanol extract of the food was fractionated over two successive silica columns. Two fractions containing different classes of oxysterols were then analyzed as trimethylsilyl derivatives by capillary gas liquid chromatography, using on-column injection and a temperature gradient from 70 to 200 degrees C. The detection limit was about 0.5 microgram/g dry weight for egg yolk powder. Fresh egg yolk contained only 1.2 micrograms/g of total oxides per g dry weight, showing that artifactual oxidation during the procedure was minimal. Recovery of 5 pure oxysterols added to egg yolk at levels of 6.5 and 10 micrograms/g was between 93 and 102%. In commercial egg yolk and whole egg powder stored for one year, total amounts of oxysterols ranging from 21 to 137 micrograms/g dry weight were found. In duplicates of mixed Dutch diets, total amounts ranged from 3.6 to 6.2 micrograms/g dry weight. Duplicates containing mostly fried and baked foods did not have higher levels than duplicates in which foods had been prepared by boiling or left raw. We conclude that a normal mixed diet provides only minor amounts of cholesterol oxidation products.