V. DISCUSSION AND CONCLUSIONS
The results obtained in this laboratory study demonstrate that major driving-related skills were impaired by BACs as low as 0.02% on some important measures for a majority of Ss who were a broadly representative sample of the driving population. The results also indicate that as BACs rise, the percentage of individuals exhibiting impairment, as well as the magnitude of the impairment, grows. Thus, there is great consistency in the relationship between the degree of impairment and BAC. Throughout the range of 0.02% - 0.10% BAC there is evidence of significant alcohol-related impairment. These findings are consistent with the findings from epidemiological crash data, which have been analyzed with contemporary statistical methods (Allsop, 1966; Hurst, 1973; Zador et al., 2000). Additionally, Ss in the study were examined only on a declining BAC curve, and the results, therefore, underestimate the magnitude of impairment expected to occur during alcohol consumption and absorption when BAC is rising.
Logic suggests that the impairment found at low BACs should be paralleled by crash and fatality data. The relationship, however, may be obscured in on-the-road data by uncontrolled variables, which can be controlled in an experiment. The laboratory data, therefore, yield conclusions about causal relationships, which frequently cannot be detected in epidemiological data.
This study further examined the issue of the universality of the conclusion that impairment exists for many behaviors at BACs as low as 0.02%. The study employed a diverse sample of the driving population as Ss and found no substantial differences between the S groups either in the BACs at which impairment appeared or the magnitude of the impairment. Within the statistical power of the study and within the breadth of diversity of the Ss, there is only random variation in the degree of impairment.
Data from epidemiological studies have suggested that age, gender, and drinking practices do differentially affect impairment. As noted earlier, however, the presence of other co-variates associated with each of these three variables may interfere with examination of the relationship of alcohol and traffic collisions and fatalities. For example, an examination of alcohol effects on traffic deaths as a function of driver age is confounded by the fact that the collision force that would moderately injure a young driver can fatally injure an older driver. Although multivariate statistical analysis can control for some of the co-variates, data from controlled laboratory experiments are better able to clarify the underlying relationship of impairment by alcohol.
Although there were essentially no significant differences in alcohol impairment between age groups, male and female Ss, or light, moderate, and heavy drinkers, it is important to note the restricted range of S characteristics. No Ss were18 years or younger, the age group that showed the greatest increase in crash rates at low to moderate BACs in the Grand Rapids data. The effects of age and lack of driving experience are confounded in the epidemiological data, and due to the legal restrictions on giving alcohol to youth, these laboratory data permit no conclusion about the relationship.
Conclusions based on the laboratory data are further limited by the lack of Ss ages 70 years and above. Although not statistically significant, the Grand Rapids data suggested an increased impairment by alcohol for these older individuals. Finally, and perhaps most importantly, as demonstrated by an ethanol clearance rate of 0.0183% per hour for heavy drinkers, the study did not examine alcohol impairment in very heavy drinkers, that is, alcoholics or alcohol abusers. These individuals were excluded out of ethical concern about administering alcohol to problem drinkers.
The major conclusion of this study is that a majority of the driving population is impaired in some important measures at BACs as low as 0.02% BAC. Although research at BACs below 0.03% has been limited, the scientific literature contains no evidence of a threshold BAC below which impairment does not occur. Nor do the data from this study provide any evidence that the driving skills of a particular category of drivers will not be impaired by alcohol.
Scientific data provide clear evidence that important driving skills are impaired at very low BACs. It falls to society as a whole, and legislative representatives in particular, to assess the costs of and the remedies for alcohol-impaired driving.