Effects of Fexofenadine, Diphenhydramine, and Alcohol on Driving Performance A Randomized, Placebo-Controlled Trial in the Iowa Driving Simulator John M. Weiler, MD; John R. Bloomfield, PhD; George G. Woodworth, PhD; Angela R. Grant, BS; Teresa A. Layton, BSN; Timothy L. Brown, MS; David R. McKenzie, MS; Thomas W. Baker, MS; and Ginger S. Watson, PhD Background: Sedating antihistamines may impair driving llergic rhinitis afflicts more than 39 million performance as seriously as alcohol. Apersons in the United States (1). Only about Objective: To compare the effects of fexofenadine, di- 4.8 million persons (12%) take prescription drugs phenhydramine, alcohol, and placebo on driving perfor- for this condition; most go without treatment or mance. self-treat with over-the-counter medications, which Design: Randomized, double-blind, double-dummy, generally contain a first-generation antihistamine. four-treatment, four-period crossover trial. These medications may be effective but carry poten- Setting: The Iowa Driving Simulator. tial risks, including drowsiness and impairment in performing everyday tasks (2–6). These adverse Participants: 40 licensed drivers with seasonal allergic events may be sufficient to dissuade some persons rhinitis who were 25 to 44 years of age. from treating their symptoms. Other patients take Intervention: One dose of fexofenadine (60 mg), di- these sedating drugs, become impaired, and try phenhydramine (50 mg), alcohol (approximately 0.1% nonetheless to perform complex tasks; as a result, blood alcohol concentration), or placebo, given at weekly they are more likely to be involved in collisions (2, intervals before participants drove for 1 hour in the Iowa Driving Simulator. 7, 8). Our goal was to examine automobile driving per- Measurements: The primary end point was coherence, a formance, a complex multiaspect task requiring continuous measure of participants’ ability to match the mental alertness; visual, auditory, and kinesthetic varying speed of a vehicle that they were following. Sec- ondary end points were drowsiness and other driving mea- information processing; eye–hand coordination; and sures, including lane keeping and response to a vehicle manual dexterity. By using the Iowa Driving Simu- that unexpectedly blocked the lane ahead. lator, a unique state-of-the-art facility, we evaluated driving performance measures and self-ratings of Results: Participants had significantly better coherence after taking alcohol or fexofenadine than after taking drowsiness to determine the effects of alcohol and diphenhydramine. Lane keeping (steering instability and first- and second-generation antihistamines on driv- crossing the center line) was impaired after alcohol and ing performance. No previous study has compared diphenhydramine use compared with fexofenadine use. the effects of these drugs in the highly controlled Mean response time to the blocking vehicle was slowest environment of a driving simulator. after alcohol use (2.21 seconds) compared with fexofena- dine use (1.95 seconds). Self-reported drowsiness did not Methods predict lack of coherence and was weakly associated with minimum following distance, steering instability, and left- Study Design lane excursion. During ragweed season, we compared the effects Conclusions: Participants had similar performance when of fexofenadine (60 mg), a second-generation anti- treated with fexofenadine or placebo. After alcohol use, histamine; diphenhydramine (50 mg) (Benadryl, participants performed the primary task well but not the Warner-Lambert Co., Morris Plains, New Jersey), a secondary tasks; as a result, overall driving performance was poorer. After participants took diphenhydramine, first-generation antihistamine; alcohol; and placebo driving performance was poorest, indicating that diphen- on driving performance and self-reported drowsi- hydramine had a greater impact on driving than alcohol ness of persons who were allergic to ragweed. A did. Drowsiness ratings were not a good predictor of im- randomized, double-blind, double-dummy, crossover pairment, suggesting that drivers cannot use drowsiness to design was used (9). The University of Iowa Insti- indicate when they should not drive. tutional Review Board approved the study, and all Ann Intern Med. 2000;132:354-363. See editorial comment on pp 405-407. For author affiliations, current addresses, and contributions, see end of text. 354 ©2000 American College of Physicians–American Society of Internal Medicine participants signed a consent form before participa- completed drowsiness scales. After the drive, study tion in the study. staff determined vital signs. Participants were ob- served until they were sober. To maintain the dou- Inclusion and Exclusion Criteria ble-blinding of the alcohol treatment, participants Key inclusion criteria were ability to remain for 5 remained for 5 hours or until the blood alcohol hours after the drives, history of alcohol use and level was less than 0.03% after alcohol and after willingness to consume alcohol, age 25 to 45 years, one of the other treatments (selected randomly). An seasonal allergic rhinitis caused by ragweed pollen, unblinded Clinical Research Center nurse with no previous successful use of antihistamine to treat sea- other study role determined alcohol levels by using sonal allergic rhinitis, status as a currently licensed a breath analyzer (Alco-Sensor, Intoximeters, Inc., experienced driver who drove an average of at least St. Louis, Missouri). three times a week for at least 3 years, and 20/20 corrected vision. Key exclusion criteria were medical Treatment Preparation and Randomization conditions that might interfere with ability to per- Capsules (fexofenadine, diphenhydramine, and form the study, pregnancy or lactation, unusual placebo) were blinded and packaged by Hoechst sleep patterns (including those of third-shift work- Marion Roussel, Inc. (Kansas City, Missouri). The ers), excessive alcohol consumption, use of tobacco Division of Pharmaceutical Service, College of Phar- in the past year or excessive caffeine consumption, macy, University of Iowa, Iowa City, Iowa, prepared previous experience in the Iowa Driving Simulator, alcohol and placebo beverages. and a positive result on a drug screening test. Driving Simulation Procedures The Iowa Driving Simulator allowed collection of At visit 1, participants were selected on the basis data on driving performance measures in a manner of inclusion and exclusion criteria. Qualified partic- not available with on-street driving (11, 12). Briefly, ipants drove in the Iowa Driving Simulator for 8 the simulator consists of a domed enclosure mounted minutes; those with a tendency to develop simulator on a hexapod motion platform. The inner walls of sickness were excluded. the dome act as a screen on which correlated images Visits 2, 3, 4, and 5 (treatment visits) occurred are projected. weekly on the same day at the same time. Partici- The experimental drive was conducted in dry pants avoided consuming food or beverages, except weather conditions, with good visibility, on a two- water, for 2 hours before these visits. Participants lane rural highway that was 72.4 km (45 miles) long. completed the baseline drowsiness visual analogue The lane widths were standard (3.66 m [12 ft]) and scale immediately before taking a capsule of fexo- the road surface was standard blacktop. The posted fenadine, diphenhydramine, or placebo; the drive speed limit was 88.6 km/h (55 miles/h) for most of was scheduled to start 2.5 hours later to coincide the course. Vehicles in the oncoming lane simulated with peak levels of antihistamine. Both researchers low-density traffic. Participants practiced driving in and participants were blinded to the treatment the simulator for 8 to 10 minutes before each ex- given. After treatment, participants were permitted perimental drive. The experimental drive consisted to consume only fluids; caffeine, stimulants, and of two phases driven consecutively without interrup- depressants were excluded. Vital signs were deter- tion. In phase 1 (30% of the total driving distance), mined and participants completed the second the driver followed a Volkswagen Golf. Phase 2 drowsiness scale 1 hour after taking the capsule. began when this lead vehicle turned off the main The study beverage was dispensed 60 minutes be- road and participants continued to drive “as you fore the scheduled drive and was consumed over 20 normally would” along the designated route. In the to 30 minutes with a light snack. The dose of alco- first three sessions, the experimental drive ended hol (or placebo alcohol) was derived by using an uneventfully. At the end of the fourth and final algorithm that included the participant’s sex and session, participants encountered a vehicle that un- weight to reach an estimated blood alcohol concen- expectedly pulled out from a driveway into the lane tration of 0.1% (21.7 mmol/L) (10). Male partici- of the experimental vehicle. A truck with trailer pants received the equivalent of 800 mg of absolute simultaneously occupied the oncoming lane. alcohol per kg of body weight, and female partici- pants received 640 mg/kg. Ninety-five percent alco- Outcome Measures hol (or placebo alcohol) was added to a glass, which During the first phase, participants were in- was filled with the participant’s choice of noncaf- structed to maintain a constant distance behind a feinated carbonated soda. Alcohol was swabbed on lead car, which had realistic random velocity fluctu- the rim of each glass to maintain blinding.