4. SUMMARY AND DISCUSSION

4.1. Impairment as a function of Behavioral Tasks (Figure 15)

An overall summary of the acute dose impairment results, as a function of H1-antagonist generation and behavioral category (or subjective sedation), is presented in Figure 15. As clearly shown, the most sensitive objective measures for detecting sedation and impairment appear to be: the Multiple Sleep Latency Test, critical or adaptive tracking, vigilance, divided attention and some driving measures. On the other hand, the categories of cognitive tasks, perception and psychomotor skills all seem to lack sensitivity overall. This may be due, at least partly, to the greater variability across types of the tasks and measures employed in the studies reviewed. Finally, the subjective measures of sedation appear to be relatively sensitive, at least for the 1st -generation drugs.

Also apparent in Figure 15, and as expected, the 1st -generation drugs generally were found to impair and sedate substantially more often than did the 2nd -generation drugs. However, it is important to emphasize that some findings of statistically significant impairment also were reported for the 2nd -generation drugs, specifically for subjective sedation as well as for all of the behavioral categories except psychomotor skills, perceptual tasks, and vigilance. The greater heterogenity of measures employed across studies for these tasks may partially explain the lack of any significant findings at least for the first two categories. In contrast, however, despite the use of a considerably more homogenous group of vigilance measures across studies, the overall results still showed no significant impairment of vigilance by the 2nd -generation drugs. This is an important finding, given that histaminergic pathways are widespread in the CNS and appear to be related to mechanisms that support alertness and vigilance during the wakeful state (Nicholson et al. 1985). Thus, the newer, 2nd -generation histamine-antagonist drugs which claim to be “non-sedating” actually may reflect a true pharmacological advance at least in terms of eliminating any disruption of vigilance.

On the other hand, the repeated reports of apparent arousal or stimulating effects noted with terfenadine and some of the other 2nd -generation drugs suggest that additional study is needed. Although the newer H1-antagonists appear to be relatively devoid of impairing effects, the findings of faster response times and apparent performance enhancement clearly warrant closer scrutiny. What are the specific pharmacodynamic actions for such effects? And what, if any, are the driving safety implications? Only carefully designed studies, using sensitive and validated measures, can address this issue by examining if such increased arousal is associated, or not, with any concomitant disruption of the ability to continue to focus on the primary driving task. Or, is such increased arousal indicative of influences on physiological systems that are not primarily CNS?

4.2. Comparison with Impairment Findings for Alcohol

As noted earlier, alcohol's effects often are used as a benchmark for evaluating the degree of impairment by medicinal drugs. Therefore, a comparison of the results of this review with those from the first author's recent review of the effects of low to moderate BAC's on driving (Moskowitz & Fiorentino, 2000) is in order. Although neither of the current reviews specifically examined the magnitude of impairment associated with alcohol or the H1-antagonists, the relative sensitivity of the various behavioral categories was summarized in each review. In brief, there are several areas of consistency, as well as discrepancy, across the findings from these two reviews. First, both reviews found support for the sensitivity of the following behavioral categories for detecting driving-related performance impairment: Multiple Sleep Latency Test (i.e., measure of wakefulness or arousal), tracking, vigilance and divided attention. Second, critical flicker fusion and simple reaction time were found to be insensitive measures for detecting alcohol's impairing effects, at least for low to moderate doses. In contrast, these two measures did appear to be relatively sensitive to the impairing effects of the 1st -generation antagonists. This suggests that different behavioral mechanisms may be involved. Thus, experimental studies of the effects of a given drug class must include specific measures related to that drug's actions, and not simply rely on the standard test batteries employed for assessing alcohol's effects.

Finally, in addition to examining impairment as a function of the behavioral tasks, as described above, there also are a number of issues which were not addressed in the current review since relevant studies were limited in availability. These issues are summarized briefly below:

4.3. Repeated Dosing And Tolerance Effects

There was a rather limited number of studies in this review which examined repeated doses. Moreover, they ranged from studies of two doses in one day to three or four doses per day over the course of two weeks. Thus, the wide variability of dosing schedules, as well as the limited number of repeated dose studies available for review, do not permit a systematic evaluation of the effects of repeated doses. Nonetheless, this is a very important issue since most individuals needing a medication do not simply take a single dose of a drug. Partial tolerance to sedation and impairment have been reported after repeated doses of the 1st -generation antihistamines in some studies (e.g., Bye et al., 1977; Walsh et al., 1994) but not in others (e.g., Alford et al., 1989; Brookhuis et al., 1993; Goetz et al., 1989). And evidence both for impairment (e.g., Volkerts et al., 1992) as well as for improved performance (e.g., Vermeeren & O'Hanlon, 1998) have been reported after chronic daily dosing with some of the 2nd -generation antihistamines, apparently due to drug accumulation. In the future, more studies will need to examine more systematically the effects of repeated doses of antihistamines.

4.4. Timing of Acute Doses Tested

Most studies tested for impairment or sedation within the window of expected peak drug effects, typically at two to three hours post-dose. Some studies utilized repeated test batteries over a five to eight hour period. However, in certain cases the lack of significant findings appeared due to testing either too early, or too late, to capture the peak drug effects. For example, two of the significant findings of impairment by cetirizine only occurred on specific measures and at much later times in the testing session, namely between 6 and 8 hours post-dose (Gengo et al., 1990; Walsh et al., 1992). Such effects clearly would be missed if the testing had only included a more limited number of measures or only earlier post-dose times as many of the other studies had done. Thus, future studies must assess the effects of antihistamines at the optimal post-dose times and employ a comprehensive, standardized test battery of the most sensitive and valid measures of sedation and driving-related impairment.

4.5. Specific Populations Tested

The typical subject population used in the majority of the studies reviewed was healthy volunteers, usually young to middle-aged men. Such a sample is appropriate as an initial step in a research program. However, more systematic research studies are needed to explore further the effects of antihistamines on other populations, including women, the elderly, and symptomatic versus asymptomatic allergy patients. In the latter case, studies are needed to evaluate whether the underlying allergy symptoms might actually contribute to impaired performance and, if so, if an antihistamine might improve performance (cf. Burns et al., 1994).

The effect of gender also may influence the test findings in terms of an inherent confound, namely women being relatively more susceptible to a given drug dose, due to their smaller body size. Indeed, of the very few significant findings of impairment by terfenadine, one was reported in a study which only tested women, and found that only the highest dose, 240 mg, caused driving-related impairment (e.g., Bhatti & Hindmarch, 1989).

Driving is a complex task requiring the integration of visual, psychomotor and cognitive skills. Age, and the various medical conditions and medications that often accompany aging, may compromise many of the skills needed to operate a motor vehicle safely. Elderly drivers are known to have a greater crash fatality risk (i.e., more fatalities when in a crash). A recent study of 3,238 drivers aged 65 and older specifically found that cognitive test performance remained significantly associated with crash risk even after controlling for driver age, race and measures of driving exposure (Stutts et al., 1998). Such findings support the validity of the various driving-related cognitive measures employed in the studies reviewed. However, there were relatively few studies which examined the effects of antihistamines on older subjects. Clearly, this area demands further study.

4.6. Clinical Efficacy Versus Side Effect Profile

Finally, another issue needing further study concerns the design of comprehensive and well-controlled studies which compare several antihistamines, with each drug tested at its indicated therapeutic dose, for clinical efficacy (i.e., using wheal and flare tests, the standard skin reaction measures of peripheral allergic effects), subjective sedation, and behavioral toxicity, all within the same study. In the current review, there is only one example of the use of such an exemplary design. It is the study by Simons et al. (1996; Ref#114) which evaluated the effects of five 2nd -generation H1-antagonists (astemizole, cetirizine, loratadine, terfenadine, ketotifen) in comparison to placebo and to the 1st -generation drug, diphenhydramine, as the positive control. The results showed that:

1) compared to placebo, the 1st -generation drug caused both significant subjective sedation and objective impairment;

2) the 2nd -generation drugs were relatively devoid of significant sedation or impairment, with the exception of cetirizine which caused significant sedation; and

3) even the 2nd -generation drugs showed some evidence of sedation or impairment relative to placebo, although the magnitude of the effects generally were not statistically significant.

It should be noted that the Simons et al. (1996) study is limited by its use of only a single objective measure of impairment, namely the evoked response potential. The results of that single study are notable, however, in that they closely mirror the findings of this current review of the findings across many studies. Thus, despite the limitations noted of the studies in this review, the overall findings do appear to be representative of the effects of the antihistamines.