106 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS odor. For example, Schmidt et al. (5) showed that VSC are detected instrumentally by GC and flame photometry, and correlated these scores with organoleptic ratings. We have previously utilized the GC instrumental procedure to assess the efficacy of rinses against mouth odor in a clinical study. Antibacterial rinses significantly reduced VSC in mouth air samples obtained three hours post rinsing, while the placebo rinse did not significantly reduce VSC. These results indicated that this method could be used for assessing the effects of active agents against local mouth odor (6). It was of interest to correlate this instrumental (objective method) with the sensory (subjective method) rating procedure. There were several obstacles in transforming this instrumental methodology into a practical tool for clinical trials in the field. The instrumental setup is large and cum- bersome and thus cannot be transported readily to a clinical site. Also, the evaluations are normally done in the morning before subjects eat or drink. This makes the sched- uling inflexible and restricts the number of subjects who can participate in the study. Lastly, no published correlation studies of this nature were conducted in a clinical study framework, between the sensory and instrumental evaluations. To facilitate this type of study, it was very useful to have a device to collect, store, and preserve the VSC mouth air samples. Such a device can then be used to analyze the stored samples upon returning to the laboratory. The storage loops described in this study fulfilled these requirements. The storage system also provided us an opportunity to conduct a true clinical study offsite and to correlate these instrumental measurements with a sensory evaluation done by expert judges prior to the instrumental VSC determination. We obtained strong positive correlation (r = 0.77-0.78) during test and recovery periods of the study but less than a robust correlation (r = 0.22) during the control period. The latter is attributed to a lack of familiarity by subjects with the procedures in the early phase of the study. These subjects initially had difficulty with the organo- leptic and storage sampling procedures. Based on previous studies, we are confident that the judges, sample collector, and analyst were not the factors contributing to the low correlation in the control phase. The noteworthy feature of this study is the development of a factor to predict sensory evaluations using the instrumental readings in this study. This provides us a useful tool to translate instrumental measurements into consumer- perceivable odor intensity. Recently, Rosenberg et al. (6) measured the correlation between the sulfide measure- ment in the mouth air and sensory ratings in forty-one subjects using a portable sulfide monitor. The overall correlation coefficient obtained by the investigators (r = 0.603) was within the range of what we observed, although the instrument used was not of the sensitivity or specificity to sulfides as compared to the GC system used to measure sulfur volatiles in mouth air. CONCLUSIONS The clinical approach used to study the correlation between sensory measurements and instrumental measurements of mouth odor indicated that: 1. There was a good correlation between the instrumental and sensory evaluations for mouth odor. The correlation coefficients were 0.22 for the control and 0.77 and 0.78 for the test and recovery phases, respectively, between the two methods.

MOUTH ODOR EVALUATION 107 2. Results also indicated that the correlation improved as the subjects become familiar with the sampling procedures. 3. Both methods were useful for measuring the intensity of offensiveness of mouth odor. REFERENCES (1) J. Tonzetich, Production and origin of oral malodor. A review of mechanisms and methods of analysis, J. Periodont. 48, 13 (1977). (2) M. C. Solis-Gaffar, J. T. Fischer, and A. Gaffar, Instrumental evaluation of odor product by specific oral microorganisms, J. Soc. Cosmet. Chem., 30, 244-247 (1979). (3) J. Tonzetich, Direct gas chromatographic analysis of sulfur compounds in mouth air in man, Arch. Oral Biol., 16, 587 (1971). (4) M. C. Solis-Gaffar, H. P. Niles, W. C. Rainieri, and R. C. Kestenbaum, Instrumental evaluation of mouth odor in a human clinical study, J. Dent. Res., 54, 35 ! (!975). (5) S.C. Schmidt, S. R. Missan, J. W. Tarbot, and A. Cooper, The correlation between organoleptic mouth odor ratings and levels of sulfur compounds, Oral Surf., 45, 560 (!979). (6) M. Rosenberg, G. V. Kulkarni, A. Basy, and C. A. G. McCulloch, Reproducibility and sensitivity of oral malodor measurements with a portable sulfide monitor, J. Dent. Res., 70, !436 (199!).