Fig. 3.

Mean values (and 95 % confidence intervals) of all intensity estimates (19 participants) of the first half of the experiment (pre: before menthol or placebo control was switched on) compared with mean values of all intensity estimates of the second half of the experiment (during: after menthol or placebo control was switched on) for five different sensations measured in eight experimental conditions. Nicotine intensity ratings show a clear decline from ‘pre’ to ‘during’ indicating desensitization in all experimental conditions without the influence of menthol stimulation. Menthol intensity ratings that were close to zero for all three sensations in the first half of the experiment and continued to stay there in the placebo (no-menthol) condition, but increased to different levels after menthol was switched on except for odor and cooling sensations, while high nicotine stimuli were concomitantly presented (VAS = visual analog scale; EU = estimation unit). The estimated mean values were derived from a linear mixed model for repeated measures using a least-square means statement