been reported for ethanol concentrations below 50 mg per liter (L) of blood; thus, blood concentrations of n-butanol below 5 mg/L would be acceptable. A stable blood concentration of n-butanol is attained after 30 min of inhalation at 100 ppm. That blood concentration is reported to be 3 mg/L (Astrand et al. 1976); therefore, an exposure to 100 ppm of n-butanol would not be expected to induce any CNS effects. This is consistent with a second approach that depends on epidemiologic evidence that CNS effects do not occur unless concentrations are well above 100 ppm (Tabershaw et al. 1944).
Long-term inhalation exposures were not found, so we relied on a 90-d oral study in rats in which there were no pathology findings at a dose of 500 mg per kilogram of body weight per day (mg/kg/d) (TRL 1986). Using body surface modeling, a 40% uptake in the respiratory system, and a nominal human inhalation rate, this equates to humans breathing a concentration of 250 ppm. Applying a 10-fold species factor gave an AC of 25 ppm for avoiding systemic effects (80 mg/m3) for exposures up to 90 d. The 180-d AC for such effects was set at half this value (90 d/180 d) to give 12 ppm as the 180-d AC. This was the lowest of the ACs, so the 180-d SMAC was set at 12 ppm or 40 mg/m3. The 7- and 30-d SMACs were set at 25 ppm. The SMACs are presented in Table 5-1.
In summary, the original approach relied on irritation reports in humans, comparison of the relative CNS-depression potency and blood concentrations of ethanol and n-butanol, and rather weak evidence that long-term ingestion of n-butanol does not cause observable pathology in rats.
The toxicity database on n-butanol was, and still is, sparse and not suitable for any of the approaches sanctioned by the National Research Council such as benchmark dose analyses or the “ten Berge” approach for time-dose extrapolations. For example, the human exposure data on irritancy come from three human studies published in the 1940s, and they give only a general idea of the exposure level at which most people would cease to experience irritation. As far as CNS effects are concerned, the data consist of blood concentrations of n-butanol that are deduced to be below the threshold for CNS effects by analogy with ethanol blood concentrations. Fortunately, these are reasonably consistent with an epidemiologic report that CNS effects are not observed at exposures below 100 ppm. For ototoxicity (a new end point), the data on rats show no effect at any exposure concentration (Crofton et al. 1994). Comparative data on n-butyl acetate (n-BA) are used to discount the relevance of putative hematologic and immunotoxicity effects of n-butanol. There are simply no data left on which to apply approaches that require discrete, quantitative end points.
NASA has considered whether genetic differences in alcohol dehydrogenase (ADH) could affect the ability of certain individuals to deal with exposures to n-butanol. Because of the multiplicity of human ADH isoforms, which