Histamine-2 Receptor Antagonists Do Not Alter Serum Ethanol Levels in Fed, Nonalcoholic Men

Methods

Our study was done at the State University of New York- Health Science Center at Brooklyn between January and April 1992. Twenty-five healthy men were recruited by posting an advertisement at the medical center. Men with alcoholism were excluded by extensive interviewing by two trained observers and by completion of the CAGE questionnaire [12]. All participants had physical examinations, and their blood counts, serum electrolytes, blood urea nitrogen, creatinine, liver enzymes, and urinalysis were determined.

Study Design

This was a prospective, observer-blinded, randomized, five-arm crossover study. At entry, the participants were randomized to receive 7 days of treatment with each of the following: cimetidine (Tagamet; SmithKline Beecham, Philadelphia, Pennsylvania) 400 mg twice daily; famotidine (Pepcid; Merck Sharp and Dohme, West Point, Pennsylvania) 20 mg twice daily; nizatidine (Axid; Eli Lilly and Company, Indianapolis, Indiana) 150 mg twice daily; ranitidine (Zantac; Glaxo Pharmaceuticals, Research Triangle Park, North Carolina) 150 mg twice daily; and no treatment. Participants were contacted daily to remind them to take their drugs, and tablet counts were done at the end of each treatment week to assess compliance. Participants abstained from alcohol and any other agents, such as aspirin [13], that might affect ethanol metabolism during the treatment week. At 1800 hours in the evening of the last treatment day, the participants ingested the remaining tablet of their assigned drug followed by a 15-minute meal, consisting of a turkey sandwich, potato salad, a roll with two pats of margarine, a cookie, and 170 mL of water. An indwelling catheter was inserted in the arm of each participant. One hour after starting the meal, the participants ingested 0.3 g/kg body weight ethanol (92.42% ethyl alcohol, Grain Processing Corporation, Muscatine, Iowa) in 500 mL of orange juice over 8 minutes. Blood samples and exhaled breath for alcohol determination were taken 0, 10, 20, 30, 45, 60, 90, 120, 150, and 180 min after ingestion of ethanol began. During the 4-hour study period further food and fluids were withheld and smoking was not permitted.

Each treatment period was separated by a 7-day “washout” period. Participants received each study drug in randomized sequence and the procedure described above repeated until all had received each of the study drugs and a control period. At the end of the study, a complete physical examination; repeated evaluation of blood counts, serum electrolytes, blood urea nitrogen, creatinine, and liver enzymes; and urinalysis were performed.

The protocol was approved by the Institutional Review Board at the State University of New York-Health Science Center at Brooklyn on 20 November 1991. Written informed consent was obtained from all participants before entry into the study.

Alcohol Determination

Immediately after completing the ingestion of alcohol, each participant rinsed his mouth and pharynx of residual alcohol by vigorous gargling and expectoration (20 times with water). Breath alcohol was assayed with a Lion Alcolmeter S-D2 (MPD Inc., Owensboro, Kentucky) using the methods recommended by the manufacturer, including the use of a fresh mouthpiece for each determination. Alcolmeters were calibrated using an artificial breath alcohol standard (NALCO, MPD Inc.) before each study session.

Venous blood was drawn into a vacuum tube at the times indicated above and placed in an ice-water bath (4 °C). Immediately after the study session, the tubes were centrifuged at 2100 rpm for 10 min at 4 °C. The serum was aliquoted into 0.5-mL microfuge tubes and stored at −70°C. To determine serum ethanol, 100 µL of sample or standard was placed in a 14-mL vial with 100 µL of 2 M ammonium sulfate [14]. The vials were immediately capped with silicone septae and aluminum crimp seals. Ethanol in the samples was determined using an HS-40 automatic headspace sampler (Perkin-Elmer, Norwalk, Connecticut) connected to a Perkin-Elmer Autosystem gas chromatography system containing a 2-meter 80/100 mesh Carbopak C column and flame ionization detector. A calibration curve was constructed (Perkin-Elmer 1020S integrator) using known standards that were assayed along with each set of test samples. Alcohol measurements by breath analysis and gas chromatography were done by persons who had no knowledge of the randomization sequence.

Statistical Analysis

The area-under-the-curve value for serum ethanol levels determined by gas chromatography was calculated by the trapezoidal method (SigmaPlot, Jandel Scientific, Corte Madera, California) using a Macintosh IIsi computer (Apple Computer, Cupertino, California) and expressed as mmol/L x h. Peak serum alcohol concentration and area-under-the-curve value were compared for each treatment. The data were evaluated using analysis of variance for repeated measures with the Tukey test and paired t-test as appropriate (Systat version 5.0, Evanston, Illinois) [15-17]. Differences in values for serum ethanol or area-under-the-curve between treatment groups were considered significant if “P” was less than 0.05. The sample size was chosen to provide 90% power, α = 0.05, to detect a 20% difference between groups [15-17].

Results

Of the 25 participants entered, 2 were withdrawn before starting the protocol; 1 because it was discovered that he did not meet the inclusion criteria and the other because of an unrelated hospitalization. The remaining 23 participants completed the protocol without incident. The mean age of these 23 men was 25.6 years (range, 21 to 35 years); mean weight was 74.3 kg (range, 52.3 to 99.1 kg); and mean height 175 cm (range, 165 to 183 cm). Sixteen participants were white; 6 were Asian (4 Chinese, 2 Korean); and 1 was an American Indian. In terms of alcohol consumption, no one gave an affirmative response to any item in the CAGE questionnaire [12]; 1 participant reported drinking 1 can of beer/d; 7 participants reported drinking 1 can of beer/wk; and 15 participants reported that they did not drink alcohol at all. One participant smoked more than 10 cigarettes/d.

Figure 1 shows the mean serum ethanol concentration curves after no treatment, cimetidine, famotidine, nizatidine, and ranitidine. The curves were superimposable and, when these data were analyzed by analysis of variance for repeated measures, no significant difference (P > 0.2) was apparent between treatment arms. Table 1 shows the peak serum ethanol concentrations, area-under-the-curve observed with each treatment arm, and P values (paired t-tests) when each H-2 blocker was compared to no drug. As indicated, there was no significant difference in peak ethanol levels or area-under-the-curve after treatment with cimetidine, famotidine, nizatidine, ranitidine, or no treatment.

View larger version:

• In this window
• In a new window

Figure 1. Serum ethanol concentrations in 23 men following postprandial ingestion of 0.3 g/kg body weight ethanol after 7 days of cimetidine, famotidine, nizatidine, ranitidine, or no treatment was determined at the times indicated. Ethanol were measured by headspace gas chromatography. Bars represent largest standard error of the mean of the five values at each time point. Effect of H-2 blocker treatment on serum ethanol concentration.

Figure 2 shows histograms for the frequency of within-person differences (treatment minus no treatment) of peak serum ethanol concentration after treatment with each H-2 blocker. Ninety-five percent confidence intervals for within-person differences in peak serum ethanol concentration (in mmol/L) were −0.44 to 0.62 for cimetidine, −0.52 to 0.38 for famotidine, −0.47 to 0.47 for nizatidine, and −0.48 to 0.74 for ranitidine. Figure 3 shows histograms for the frequency of within-person differences (treatment minus no treatment) of serum ethanol area-under-the-curve after treatment with each H-2 blocker. Ninety-five percent confidence intervals for within-person differences in area-under-the-curve (in mmol/L x h) were −0.41 to 1.09 for cimetidine, −0.73 to 0.53 for famotidine, −0.44 to 0.88 for nizatidine, and −0.85 to 0.97 for ranitidine.

View larger version:

• In this window
• In a new window

Figure 2. Histograms showing the frequency (%), within the indicated ranges, of net peak serum ethanol concentrations for 23 participants after treatment with each of four H-2 blockers. Within-person differences in peak serum ethanol concentration.

View larger version:

• In this window
• In a new window

Figure 3. Histograms showing the frequency (%), within the indicated ranges, of net serum ethanol area-under-the-curve values for 23 participants after treatment with each of four H-2 blockers. Within-person differences in serum ethanol area-under-the-curve.

Analysis of simultaneous measurements of breath ethanol corroborated the serum values obtained for peak ethanol concentration and area-under-the-curve. Figure 4 shows a close correlation between peak serum and breath ethanol concentration. Peak breath ethanol concentration (mean ± standard error of the mean, in mmol/L) was 3.68 ± 0.32 for cimetidine; 3.22 ± 0.28 for famotidine, 3.16 ± 0.28 for nizatidine, 3.57 ± 0.33 for ranitidine, and 3.25 ± 0.32 for no treatment. As with peak serum ethanol concentration (see Table 1), when the breath data were analyzed by the paired t-test, no significant differences [P > 0.2] were observed when treatment arms were compared to no treatment. Likewise, no significant difference was found (P > 0.2) in the breath ethanol area-under-the-curve when values after treatment with H-2 blockers were compared to values after no treatment (data not shown).

View larger version:

• In this window
• In a new window

Figure 4. Values shown represent peak ethanol concentrations measured using the Lion Alcolmeter breath analyzer and gas chromatography of serum samples after treatment with H-2 blockers and no treatment in 23 participants. Although labeled Breath Ethanol Peak for clarity, values determined using the Lion Alcolmeter actually represent blood ethanol concentrations. The correlation coefficient (r) was calculated with a least-squares analysis. Comparison of peak breath and serum ethanol concentrations.

Because it has been reported that Asians have decreased gastric alcohol dehydrogenase activity compared to non-Asians [11], we compared the results for the 6 Asian men in our study to those for the 17 non-Asians. The American Indian participant was included in calculations for the non-Asian group because alcohol metabolism in American Indians has been reported to be the same as that in whites [18]. As can be seen in Figure 5, the mean values for peak serum ethanol concentration (top) and area-under-the-curve (bottom) for Asians was significantly less for each treatment arm than the values observed in non-Asians (P < 0.05 to 0.01). Nevertheless, as observed with the composite group, when Asians and non-Asians were evaluated separately, no significant differences were observed in the peak ethanol concentration or area-under-the-curve between treatment arms (P > 0.2).

View larger version:

• In this window
• In a new window

Figure 5. Bars represent mean peak serum ethanol concentration ( ) and area-under-the-curve ( ) in 6 Asian and 17 non-Asian men after postprandial ingestion of 0.3 g/kg body weight ethanol after 7 days of cimetidine, famotidine, nizatidine, ranitidine, or no treatment. Error bars represent standard error of the mean. Comparison of peak serum ethanol concentration and area-under-the-curve in Asian and non-Asian participants.TopBottom

Discussion

Impaired judgment and delayed response time caused by alcohol consumption result in countless tragedies each year, including more than 20 000 traffic fatalities in the United States alone [19, 20]. For this reason, determining whether H-2 blockers alter ethanol metabolism is very important. Lieber and colleagues have reported that in nonalcoholic men, cimetidine, ranitidine, and nizatidine alter ethanol metabolism, thereby resulting in major increases in blood alcohol concentrations [2-4, 20]. If these findings were reproduced by others, it could be argued that these drugs should be removed from the market because of the vast number of persons at risk and because drugs like famotidine and omeprazole that are equally effective and do not alter blood alcohol levels are available [4, 21]. Certainly, physicians prescribing drugs that elevate blood ethanol levels to patients who drink might face legal consequences of alcohol-related injury. In fact, in Europe, cases of this sort have already been reported [4].

In our study, as in several others [7-10], the reports that cimetidine and ranitidine increase blood alcohol levels after moderate alcohol consumption [2-4] could not be substantiated. Nevertheless, the observation that famotidine does not alter blood alcohol levels [4, 7, 10] was confirmed. One criticism that has been raised regarding the failure of previous studies to detect an effect of H-2 blockers on ethanol metabolism is that the statistical power to exclude a type 2 error was inadequate [4]. Analysis of the potential for a type 2 error in the present study, that is, falsely accepting the null hypothesis (absence of an effect of H-2 blockers on blood ethanol levels), indicates that such an error cannot explain our findings. The confidence intervals indicate that the probability was 95% of detecting an H-2 blocker-induced increase in peak serum ethanol or area-under-the-curve of greater than 18%. The calculated chance of failing to observe the reported [4] 96% and 41% increases in ethanol area-under-the-curve following treatment with cimetidine and ranitidine, respectively, was less than 1%.

Another criticism of previous studies that failed to show an effect of H-2 blockers on alcohol metabolism is that many participants may have been alcoholics with decreased gastric alcohol dehydrogenase activity [4]. Again, this cannot explain the findings in our study. The participants were all medical or graduate students who were carefully evaluated by two trained observers regarding their history of alcohol use before they became aware of the nature of the study; responded in the negative to all items in the CAGE questionnaire [12]; and reported drinking one can of beer/d or less. In fact, most of the participants (65%) did not use alcohol at all.

One must consider other reasons for these discrepant results. The most probable explanation is differences in study design. It is well known that there are wide variations in peak alcohol concentration and area-under-the-curve between individuals, particularly if one does not control for ethnicity [6, 19, 22]. Differences in alcohol metabolism between individuals may be greater than those reported for the effects of H-2 blockers. For example, in the control arm of our study, an eightfold difference between lowest and highest peak serum value (see Figure 4) and area-under-the-curve was noted between participants, whereas during the course of the study, the parameters varied much less for a given participant. Moreover, the mean peak serum ethanol and area-under-the-curve values for non-Asians [3.2 mmol/L and 4.4 mmol/L x h] were 50% to 60% greater than those for Asians (2.1 mmol/L and 2.7 mmol/L x h). Hence, it is clear that a study of this nature requires a sufficient number of participants to ensure adequate power for the study and that the participants must serve as their own controls [15-17]. Whereas our study involved 23 men who were randomly assigned to each of five treatment arms (including no treatment), studies by Lieber and colleagues included as few as five participants in treatment arms [3] and sometimes made comparisons to different sets of controls [4]. Other differences between studies evaluating the effects of H-2 blockers on alcohol metabolism, such as the choice of meal or the time of day, are less likely to account for the different outcomes.

Finally, we need to comment on the apparent ethnic differences in alcohol metabolism observed in this study. As noted above, for each treatment arm, the mean peak serum ethanol level and area-under-the-curve for Asians was significantly less than that observed for non-Asians (see Figure 5). Previous reports that Asians achieve lower blood ethanol levels than non-Asians after ingesting a given quantity of alcohol have ascribed this to genetic differences in alcohol dehydrogenase [22, 23]. Nonetheless, when the data for Asians and non-Asians were evaluated separately, no detectable effect of H-2 blockers on serum ethanol concentrations was found in either group.

On the basis of the results of this randomized, crossover study, we conclude that neither cimetidine, famotidine, nizatidine, nor ranitidine elevate serum ethanol levels following consumption of moderate amounts of alcohol by fed, nonalcoholic men. Consequently, when treating nonalcoholic men, possible adverse effects on ethanol metabolism need not be a criterion for choosing among the currently available H-2 receptor antagonists.