UC San Diego

It is widely acknowledged that among the likely causes for building-related complaints, the aggregate effect of a variety of VOCs deserves particular attention (Apter et al. 1994, Hodgson et al. 1994, Kostiainen 1995, Rothweiler and Schlatter 1993). Among the various symptoms evoked, sensory irritation not only figures prominently but also lends itself to psychophysical measurement in humans (Cometto-Muñiz and Cain 1992, Hudnell et al. 1992, Kjærgaard et al. 1992, Mølhave et al. 1991).

 Our lab has a particular interest in studying the functional characteristics of the senses of smell and sensory irritation in humans (Cometto-Muñiz and Cain 1992, Cometto-Muñiz and Cain 1996). Our approach entails a stimulus-strategy and a response-strategy. From the perspective of the stimulus we have chosen to test families of chemicals, typically homologous series (e.g., acetate esters) but also more diverse groups (e.g., terpenes). Homologous series provide a convenient “unit of change”, represented by carbon chain length, along which physicochemical properties change in an orderly fashion, allowing to relate those changing properties with the sensory outcome. From the perspective of the response, we have resorted to separate the olfactory from the trigeminal response of the nose by testing subjects lacking olfaction, i.e., anosmics, for whom odor does not interfere. The applicability to normosmics of the nasal trigeminal responses obtained from anosmics was initially suggested by the similarity of eye irritation thresholds in both groups, and further supported by the similarity of nasal localization thresholds in both groups (Cometto-Muñiz and Cain 1998).

In the studies described, use of a uniform sensory methodology, procedure, and instructions in a small but intensively tested group of subjects has been combined with selection of a wide range of VOCs relevant to indoor air. The results have permitted to build a strong QSAR, based on a solvation model, that describes and predicts nasal pungency and eye irritation thresholds for VOCs using a maximum of five general physicochemical descriptors (Abraham et al. 1998a, Abraham et al. 1998b). Modeling of odor thresholds via the solvation equation has been less successful. This can be taken as an indication that some key steps on the odorant-receptor interaction rely on more specific stimulus properties than those reflected on the five general physicochemical descriptors.

 Finally, we discuss the outcome of experiments comparing the sensory impact of VOCs presented singly and in mixtures of up to nine components. The results showed various degrees of agonism among the components of the mixtures. Such agonism allowed detection of the mixtures when their constituents were present at concentrations below their individual thresholds. As the number and the lipophilicity of the components increased, so did their degree of agonism.