Methods for health risk assessment of chemicals: are they relevant for alcohol?

A description is given of the principles used for toxicological risk assessment of food additives and environmental contaminants. The terminology is not generally accepted. Hazard identification is to identify whether a substance may cause toxic effects in man. Dose-response assessment is to understand the relationship between exposure and various toxic effects. The shape of dose-response curves is discussed. For most effects (including nongenotoxic carcinogens), a threshold is assumed, under which there is no risk. For genotoxic compounds, no such threshold is assumed. For substances with a threshold, an uncertainty factor is applied to the no-observed-adverse-effect level or the lowest-observed-adverse-effect level, to arrive at an acceptable or tolerable daily intake. In the case of ethanol, some of these principles of toxicological risk assessment may be used and based on epidemiological dose-response data. The critical effects are liver effects, fetal damage, and cancer. All of these effects may be assumed to have thresholds. Possible uncertainty factors are discussed. However, because of the decreased risk of cardiovascular effects at low/moderate intakes and the fact that alcohol intake is a lifestyle factor, it is doubtful whether the application of a no-observed-adverse-effect level/uncertainty factor approach--as used for food additives and environmental pollutants, to which there is unavoidable exposure--is really applicable or meaningful for ethanol.

A new model function for continuous data sets in health risk assessment of chemicals using the benchmark dose concept.

The benchmark dose concept is an alternative way to calculate acceptable daily intakes or other low-risk limits of exposure for nongenotoxic compounds. An effective dose that corresponds to a specific change of effect/response, e.g., 10%, over background, on a given model is calculated and a benchmark dose which is the lower confidence bound for the effective dose can be estimated. One problem with this method is that it has not been possible to model "S"-shaped curves for continuous data with the existing commercial computer programs. We present a new mathematical model for continuous data, which can form S-shaped curves. For a number of data sets on the toxicity of trichloroethene, the introduction of this new function significantly increased the possibilities to model continuous data. A new computer package has also been developed to improve and simplify the calculations of benchmark doses.

Application of the benchmark method to risk assessment of trichloroethene.

An alternative approach for risk assessment of nongenotoxic substances, the benchmark method, has been evaluated and applied to trichloroethene as a test case. The benchmark dose is the dose that corresponds to a specific increase in risk, normally 1 or 10%. Experimental data from the literature on trichloroethene were used for these calculations. Eighty sets of data on effects on liver, kidney, the central nervous system, and tumors were analyzed. All non-observed-effect levels (NOELs) were higher than the benchmark dose corresponding to 1% extra risk, and 42% of the NOELs and 93% of the lowest-observed-effect levels (LOELs) were higher than the benchmark dose corresponding to 10% extra risk. The present study confirms that the benchmark methodology gives a more detailed picture of dose-response relationships than risk assessment using the NOEL/LOEL approach and facilitates comparison between various toxicity studies. However, the polynomial regression models used in the present study quite often failed to fit the experimental data. Despite the advantages with the benchmark approach, several factors must be considered in the risk assessment process. In the case of trichloroethene, a revised risk assessment using the benchmark approach would lead to a similar guideline value as the traditional NOEL/LOEL approach.

DNA damage in lung cells in vivo and in vitro by 1,3-butadiene and nitrogen dioxide and their photochemical reaction products.

A UV-irradiated mixture of 1,3-butadiene and nitrogen dioxide (NO2) was tested for its potency to induce DNA damage measured as single-strand breaks (SSB) in lungs of mice. Both gases were also tested separately. After 16 h exposure a UV-irradiated mixture of 40 ppm butadiene + 20 ppm NO2, but not 20 ppm butadiene + 10 ppm NO2 + UV, induced a significant increase in SSB as measured by the alkaline unwinding technique. There was no increase in the level of SSB using the alkaline elution technique during the same testing conditions. However, after 5 h exposure to 60 ppm butadiene + 30 ppm NO2 + UV both methods demonstrated a significant increase in SSB. Mice were also exposed to butadiene at 80 and 200 ppm for 16 h and at 500 ppm for 5 h. DNA damage was demonstrated in both liver and lung after 5 and 16 h (only at 200 ppm) of exposure using the unwinding technique. Using the alkaline elution assay, a significant increase in the level of SSB in lung and liver was found only after 5 h of exposure. When mice were exposed to 30 ppm NO2 for 16 h or 50 ppm for 5 h, a significant increase in SSB was found with the unwinding technique. Alveolar macrophages from mice were also exposed in vitro to the gas mixture and to butadiene and NO2 separately. In these experiments, the DNA damage was studied with the unwinding technique. A significant effect was demonstrated with 40 ppm butadiene + 20 ppm NO2 + UV. NO2 itself contributed to some extent to the increase. Reasons for the discrepancies between the unwinding and the alkaline elution techniques are discussed.

Review of the genotoxicity of nitrogen oxides.

Nitrogen oxides (NOx) are formed in combustion processes and are major pollutants in urban air. Relatively few studies on the genotoxicity of NO2 and NO have been performed. These studies indicate that NO2 is genotoxic in vitro, but the effect of NO seems to be very slight. One in vivo study showed chromosome aberrations and mutations in lung cells after inhalation of NO2 (and NO), but tests for chromosome aberrations in lymphocytes and spermatocytes or micronuclei in bone marrow were negative after inhalation of NO2. Based on present studies, there is no clear evidence of a carcinogenic potential of NO2, although lung adenomas were induced in the susceptible strain A/J mouse. The primary metabolites of NOx are nitrite and nitrate. Nitrate seems to be devoid of genotoxic properties, but nitrite is genotoxic in vitro, and there are also positive in vivo results. Cancer studies have been mainly negative. However, carcinogenic nitrosamines have been shown to be formed in vivo after inhalation of NO2. Nitrogen oxides are key components in atmospheric smog formation, which may lead to secondary effects. Strongly mutagenic nitro-PAH compounds are easily formed, and mutagenic reaction products may be formed photochemically from alkenes.

Risk comparisons between limit values for ionizing radiation, PAH, and benzene in Sweden.

The background of regulatory limit values for carcinogenic agents in Sweden is discussed and exemplified with the ambient and occupational air pollutants benzene and PAH (especially benzo[a]pyrene, BaP), ionizing radiation, and radon. The estimated cancer risks at different limit values are compared, as is the estimated number of cancer cases annually due to existing pollutant levels. Although the individual lifetime cancer risks are much higher at the occupational limit values for benzene and BaP than what is recommended for the general public, the estimated number of cancer cases annually is lower at existing pollutant levels. The individual cancer risk at the occupational dose limit for ionizing radiation is comparable to the occupational cancer risk with BaP, but higher than the one for benzene. At the dose limit for the general public, the radiation risk would be higher than what is recommended for individual air pollutants. However, doses from man-made radiation are usually much lower than the dose limit due to optimization. Thus, the estimated number of cancer cases annually due to radiation is low and comparable to the estimated number due to the chemical air pollutants discussed. In contrast, the lung cancer risk with radon in dwellings both at limit values and existing levels is high and comparable to occupational limits for the chemical carcinogens and for radiation. The estimated number of cancer cases annually at existing radon concentrations (1100 cases) is much higher than for the other discussed pollutants (0.03-7 cases each) and also higher than the estimated number of lung cancer cases due to total air pollution (approximately 100 cases).

Review of the genotoxicity of ozone.

Ozone is a powerful oxidant, reactive to biomolecules. In aqueous solution it decomposes to give hydrogen peroxide, superoxide and hydroxy radicals which can take part in secondary reactions. Ozone is a disinfectant that inactivates both viruses and bacteria. Although other reactions are primarily responsible for the inactivation, cellular DNA is also damaged. Ozone is genotoxic to microorganisms, plants and cell cultures in vitro. The results from in vivo cytogenetic studies with laboratory animals after inhalation exposure are contradictory. Chromosome aberrations in lymphocytes, but not SCEs, have been demonstrated in Chinese hamsters but not in mice. Chromatid deletions were induced in pulmonary macrophages in rats. No cytogenetic effects have been reported for bone marrow cells or spermatocytes. The few experimental and epidemiological studies with human subjects do not allow a conclusion on the cytogenetic effects of ozone in lymphocytes in humans. No life-long cancer studies have been performed with ozone. However, after 4 and 6 months of inhalation exposure, lung adenomas were induced in strain A/J mice, but not in Swiss-Webster mice.

Mutagenic activity of photoreaction products formed from chlorinated ethenes and nitrogen dioxide.

The mutagenicity in Salmonella, strain TA 100, of photoreaction products from chlorinated alkenes and nitrogen dioxide, was studied after a mean reaction time of 40 min with UV-irradiation and an exposure time of 20 hr. Only vinyl chloride (monochloroethene) gave rise to significantly mutagenic photoreaction products. Of the other chloroethenes, 1,1-dichloroethene and tetrachloroethene, but not 1,2-dichloro- or trichloroethene, formed slightly mutagenic products. Bacteriotoxic effects occurred at low doses, especially with tri- and tetrachloroethene.

Genotoxic activity of 1,3-butadiene and nitrogen dioxide and their photochemical reaction products in Drosophila and in the mouse bone marrow micronucleus assay.

The genotoxic activity of a photochemical reaction mixture of 1,3-butadiene and nitrogen dioxide was investigated in vivo in the mouse bone marrow micronucleus assay and the somatic mutation and recombination test in Drosophila (the wing spot test). Butadiene alone was not mutagenic in Drosophila, but induced micronuclei in mice at 10 ppm after 23 h of exposure. Nitrogen dioxide was not genotoxic in either test system. The photochemical reaction products were toxic but probably not mutagenic in Drosophila and not genotoxic in mouse bone marrow. The in vivo results do not confirm earlier in vitro results that demonstrated a strong direct-acting mutagenic activity of the photochemical products in Salmonella.

Mutagenic activity of ultraviolet-irradiated mixtures of nitrogen dioxide and propene or butadiene.

The mutagenic activities of mixtures of nitrogen dioxide and 1,3-butadiene or propene were investigated after uv-irradiation in a small, laboratory-bench scale flow-through gas exposure system. The tester organism was Salmonella typhimurium, principally strain TA100. The photoreaction products from 1,3-butadiene and nitrogen dioxide were more mutagenic than those from propene and nitrogen dioxide. Approximately 0.25 ppm butadiene, compared to 100 ppm propene, was needed to give a significant mutagenic effect with 0.25 ppm NO2 after 6 hr exposure. The influence of different experimental conditions on mutagenic activity was studied using propene plus nitrogen dioxide. Increasing the mean reaction time from 40 min to 1 hr 20 min or 3 hr 20 min by reduction of the flow rate through the 20-liter reaction vessel did not appreciably increase the sensitivity of the system, nor did humidification of the air, omission of the metabolic system (S9 mix), or spreading of the bacteria on the agar surface. Prolongation of the exposure time from 6 to 16 or 24 hr did, however, give an increased mutagenic response. With prolonged exposure, a slight mutagenic effect could also be detected with ethene + NO2 + uv. Ozone addition did not appreciably enhance the mutagenic response.

Absence of mutagenic response to radiation from a video display terminal.

The standard Ames Salmonella test (TA 100) was used to detect the mutagenicity of radiation from a video display terminal. The Ames test is a sensitive assay that detects the ability of a chemical to damage deoxyribonucleic acid. It has also been employed to detect the mutagenicity of electromagnetic radiation. An extremely short distance (62 mm) from a video display terminal and an extremely high electrostatic field strength (250 kv/m) was employed. No mutagenic response was found in this test system.

A method for studying the mutagenicity of some gaseous compounds in Salmonella typhimurium.

A dynamic flow-through exposure system was designed for mutagenicity studies of gaseous compounds in Salmonella. Salmonella typhimurium strain TA100 was the primary tester strain. The dose ranges were 0.5-20% of vinyl chloride, ethene, propene, and 1,3-butadiene, 1-200 ppm of ethylene oxide, 0.5-20 ppm of nitrogen dioxide, and 0.1-3.5 ppm of ozone. The gas flow rate was 250, 500, or 1,000 ml/min, and the exposure time was 6 or 7 hours. Of the tested gases, vinyl chloride, ethylene oxide, and nitrogen dioxide were mutagenic. Ethene, propene, and 1,3-butadiene were not mutagenic in this system. Ozone is bacteriotoxic, and no mutagenic effect could be demonstrated in the nontoxic dose range. The exposure system was considered suitable for studies on gaseous chemicals.

Photochemical formation of mutagenic compounds from alkenes and ozone or nitrogen dioxide.

In order to investigate the possible formation of mutagenic compounds from alkenes emitted in ambient air, laboratory experiments were performed with Salmonella typhimurium strain TA100 in a small-scale flow-through exposure system. The reaction time for mixtures of alkenes with ozone or nitrogen dioxide was 40 minutes, and the exposure time for bacteria was 6 hours. Ozone gave rise to a small mutagenic effect in combination with 1,3-butadiene or vinyl chloride, with and without ultraviolet (UV) irradiation, but not in combination with ethene or propene. Nitrogen dioxide gave rise to a mutagenic effect in combination with propene, 1,3-butadiene, or vinyl chloride, but only after UV irradiation. The mutagenic activity was highest with butadiene and seemed to be dose-related to the concentration of nitrogen dioxide. Nitrogen dioxide with ethene did not produce a mutagenic effect. A mixture of ethene, propene, and butadiene, tested with ozone or nitrogen dioxide with UV irradiation, did not potentiate each other's mutagenic effect.

Retinol (vitamin A) inhibits the mutagenicity of o-aminoazotoluene activated by liver microsomes from several species in the Ames test.

Retinol (vitamin A) has earlier been shown to inhibit the mutagenicity of o-aminoazotoluene (OAAT) in the Salmonella/microsome assay when OAAT is activated with S9 from Sprague-Dawley rats. The results presented in this paper confirm this and also show that S9 from mice, hamsters and gerbils activates OAAT to mutagenic metabolites detected by Salmonella typhimurium TA100. However, S9 from rabbits is inactive. The S9 fraction from rabbits also shows a low aryl hydrocarbon hydroxylase (AHH) activity. The AHH activity or protein content of the microsomal fraction cannot be used to predict the activating capacity of S9 from the other species. Retinol, added in vitro, inhibits the mutagenic effect of OAAT activated by mouse, gerbil or hamster S9. The strongest inhibition is observed with hamster S9 while the inhibition of mouse and gerbil S9 is lower but still higher than in the rat.

A field study of some swimming-pool waters with regard to bacteria, available chlorine and redox potential.

The waters of one indoor bath and three outdoor baths were examined once an hour during 3 days (bath 1) or 6 days, for available chlorine, redox potential, permanganate number, ammonium, nitrate and total nitrogen, total bacterial count at 22 degrees C., total bacterial count at 37 degrees C. and faecal coliform bacteria. The weather, number of swimmers and the chlorine gas addition were continuously registered, and the pH was checked a few times at each bath. In bath 1, an indoor pool with aluminium sulphate precipitation about once a week and with sand filters back-washed every 2 days, less than 10 bacteria/ml. were found in all samples. In bath 2, an outdoor pool with aluminium sulphate precipitation twice a week and with sand filters back-washed twice a week, also few bacteria were found. In bath 3, an outdoor pool with only filtering through sand filters back-washed about every 14 days, high bacterial counts were found every day except the first, when the filters had been newly back-washed. In bath 4, an outdoor pool with only filtering through sand filters back-washed about once a week, high bacterial counts were found now and then during the first 4 days when the weather was warm, but few bacteria were found the last 2 days when the weather was cold and windy, and there were few swimmers.Values from different analyses on the same sample showed relatively good correlation between the redox potential and the free available chlorine. In bath 3 both the redox potential and the available chlorine were weakly correlated to the bacterial count, but in bath 4 there was no such correlation. No other factors were well correlated with the bacterial count either.The bacterial counts at 22 degrees and 37 degrees C. were of the same order. No faecal coliforms were ever found. Use of these bacteria as indicator organism in swimming pools is criticized.The method of using certain minimum values of the free available chlorine as guarantee for a satisfactory bacteriological quality of the swimming pool water is also questioned. The degree of purity of the water is fundamentally connected with the disinfecting power of the available chlorine.Use of certain minimum values of the redox potential, according to these investigations, seems to be a method of somewhat greater accuracy. Provided that the methods of precipitation are performed correctly and filters are being back-washed often enough, then an automatically registering redox potential device, perhaps connected to the chlorine gas pump, ought to constitute a good control of the hygienic quality of a swimming-pool water. This must, however, always be completed by bacteriological examinations, preferably made at high bathing load.

The effect of saline on the eye irritation caused by swimming-pool water.

In laboratory experiments the acute eye irritation produced by exposure to tap water was not significantly increased when chlorine compounds were added to the water at concentrations of 1 mg./l. The greatest irritation was produced by 2 mg. Cl(2)/l. as NH(2)Cl. The addition of NaCl at concentrations above about 0.5% abolished the irritant effect of tap water, and prevented irritation even when 1 mg. Cl(2)/l. was present.In a field experiment involving two swimming baths, one with fresh and the other with saline water (0.5% NaCl), eye irritation in the saline bath was significantly lower than in the freshwater bath only when the swimming time did not exceed 30 min.