274 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS usually have a pH of about 12, while the normal pH of the skin is between 4 and 6 (6). The metallic salts of zinc, aluminum, and zirconium used in aero- sol antiperspirants may cause superficial skin infections because they produce narrowing of the ducts of sweat glands which may give rise to apocrine sweat gland occlusion and a true hydradentitis (6, 18). The refrigerant action of propellants in personal or household aerosols can produce cooling or freezing of the sensitive corneal (5) or vaginal (7) tissues if they are used improperly. Many facets of skin metabolism, includ- ing cellular respiration, can be altered by some cosmetic aerosols (20). The propellants have been reported to sensitize the myocardium to arrhythmias caused by anoxia and catecholamines (21), although there is no definitive evidence that this type of cardiac toxicity can result from normal use of any aerosol product (22, 23). The interruption of the normal bacterial flora of several body orifices was formerly a potential problem with vaginal deodorant sprays containing antibacterial agents (24). It is known that once the normal flora of the vagina or rectum is altered (as with tetracycline therapy), an abnormal overgrowth of yeasts and fungi (usual symbionts in these areas) may cause pruritis (25) and/or other bacterial infections. Since most vaginal sprays no longer contain antibacterial chemicals, this problem has essentially ceased to exist. Keratitis, due to a foreign body reaction of the skin, has been reported in response to the presence of materials from several household aerosols that have, in effect, been driven into the skin by the force of the spray (5). An increased formation of pulmonary edema fluid and lipid pneu- monia has been reported following a foreign body response to the inhalation of oil droplets from an aerosol containing a vegetable oil (11). Edema fluid is an ideal medium for the groxvth of pathogens, resulting in aspiration pneumo- nia. A slight reduction in specific airway conductance, posing no clinical danger, has been reported following the inhalation of several bronchodilator aerosols. The bronchoconstriction reported, which was less than that caused by smoking a cigarette, was attributed to the aerosol propellant and/or the surfactant chemicals (sorbitol trioleate and soya lecithin) contained in the two bronchodilator aerosols evaluated (26). PROPELLANT TOXICITY AND ABUSE The controversy regarding the toxicity of aerosol products, in general, and the propellant chemicals they contain, in particular, began several years ago with the publication in the lay press of several deaths due to "sniffing," especially by teen-agers, of the vapors from a wide variety of aerosol products (27). The practice involved the deliberate, deep inhalation of the concen- trated vapors, usually from a balloon or paper bag (28). These reports of aerosol abuse brought the toxic potential of the various propellants to the attention of the aerosol industry. The toxicity of the pro- pollants can be divided into three major categories: toxicity due to the refrig-

CURRENT PERSPECTIVES ON AEROSOL TOXICITY 275 erant action of the propellants (12), which may cause reflex airway obstruc- tion, especially of the larynx, and tissue damage to the delicate mucous mem- branes of the vulva (7) and eye (29) decomposition into phosgene when the vapors come into contact with an open flame (30) and finally, systemic toxicity, chiefly referrable to the cardiovascular system (9,1, 9,2). It is this lat- ter type of toxicity which has been accorded so much unwarranted publicity by the press. The problem of teen-age abuse of aerosols is one over which the aerosol industry has little control, other than to update the warnings printed on the aerosol cans. The Inter-Industry Committee on Aerosol Use has established the Aerosol Education Bureau (31) to administer a safety campaign which is designed to warn teen-agers of the potential lethal consequences of abusing aerosolized products. Soon after the controversy regarding the deaths from aerosol "sniffing" began, Taylor and Harris (21) reported that the exposure of mice to several propellants, followed by asphyxia in a plastic bag, produced sensitization of the myocardium to hypoxia, resulting in arrhythmias such as sinus brady- cardia, atrioventricular block, and T wave depression. They used these ex- periments to postulate that the sudden deaths that followed aerosol abuse by teen-agers could be the result of a toxic action of the propellants used in al- most all aerosol packages, as well as to provide a basis for warning against the possible hazards to frequent users of a variety of aerosol products. It is well known, and has been for some time, that high concentrations of many pro- pellants frequently used by aerosol manufacturers can produce a wide variety of toxic effects. Ataxia, tremors, liver, and kidney damage are among some of the more common findings (32, 33). But these experimental results in animals, especially at the high concentrations studied, bear little or no relationship to the lower concentrations to which the consumer of aerosol products is ex- posed (9,8). Also, there is lack of general agreement as to the accuracy of the data on aerosol toxicity when it is extrapolated from animal studies to humans (22, 34). Following several reports by Taylor and Harris on the cardiac toxicity of aerosol propellants, other investigators attempted to reproduce their findings, with little success. McClure, in 1972, failed to produce significant changes in the heart rate or electrocardiogram of anesthetized mice after the administra- tion of several propellants in aerosol form, followed by asphyxia (35). In general, he found that the cardiovascular effects produced by propellant ex- posure were similar to the cardiovascular effects of asphyxia alone. McClure was also unable to confirm similar findings reported by Taylor and Harris in dogs (21). Egle et al. (23) also attempted to repeat the results of Taylor and Harris. They exposed mice to several propellants, either alone or with nitro- gen-induced asphyxia, and reported no augmentation of the asphyxia-induced bradycardia or atrioventricular block by the several fiuorocarbon propellants