The problem of taint in pork: 1. Detection thresholds and odour profiles of androstenone and skatole in a model system

Detection thresholds of androstenone and skatole were established using a trained 10-member sensory panel. Odour profiles for these compounds were also developed using a model system comprising a lipid base spiked with synthetic androstenone and skatole. A total of 2160 presentations were used for the determination of sensory thresholds which were obtained using the Ascending Method of Limits with a 3-AFC (Alternate Forced Choice) mode of presentation. Preliminary odour profiles were obtained by three consensus-profiling sessions. Character notes developed were then validated by profiling four classes of model samples containing androstenone, skatole, a mixture of skatole plus androstenone and a mixture of skatole plus androstenone plus indole. Individual sensory thresholds for the ten panel members varied between 0·2 to 1·0μg g-1 for androstenone and 0·008 to 0·06 μg g-1 for skatole, leading to group thresholds of 0·426 μg g-1 and 0·026 μg g-1 for androstenone and skatole, respectively. The important descriptors for the odour of androstenone included; sweaty, ammonia, dirty, parsnip, silage, nosefeel and acrid. Among these, only the sweaty note was found to be a significant (p < 0·05) basis for separating pure samples of androstenone from pure samples of skatole by the panel. The important descriptors for the odour of skatole were; mothball, musty and body reaction. Both the mothball and musty notes were a significant (p < 0·001 to p < 0·05) basis for distinguishing between either pure samples of skatole or mixtures containing skatole from pure samples of androstenone. The results show that androstenone and skatole are synergistic, but only in relation to the descriptors defining intensity of the odours, and not in those defining the character of their individual odours. © 1997 Elsevier Science Ltd.