SYNOPSIS. Problem drinkers among drunk drivers were identified by the estimation of gammaglutamyl transpeptidase at arrest. The prevalence of problem ...
BLOOD ALCOHOL CONCENTRATION OR GAMMAGLUTAMYL TRANSPEPTIDASE AS A PREDICTOR OF PROBLEM DRINKING AMONG DRUNK DRIVERS? *
J. A. Dunbar, M.B., Ch.B., M.R.C.C.P, D.M.J. ; M. S. Devgun, Ph.D., MIBiol.; J. Hagart, MSc.; S. A. Ogston, M.A., MSc.; and B. T. Martin, BSc., Ph.D. SYNOPSIS Problem drinkers among drunk drivers were identified by the estimation of gammaglutamyl transpeptidase at arrest. The prevalence of problem drinking among drunk drivers is discussed. Results of a follow-up study showing deterioration in alcohol problems following revocation of the driving license are shown. Implications for changes in the United Kingdom drinking and driving laws are discussed. INTRODUCTION Since May this year the Road Traffic Act in the United Kingdom has been amended to recognize "high-risk offenders" among drunk drivers. By this act high risk offenders are defined as drivers who are convicted of drinking and driving on 2 occasions within 10 years with blood alcohol concentrations above 200 mg/100 ml (43.4 mmol/1) or refusals to provide a specimen. Upon arrest the driver must be investigated by the Medical Advisers of the Department of Transportation for evidence of dependence on alcohol or a drink problem before his driving license is restored. An alternate method of identification may include estimation of gammaglutamyl transpeptidase (GGTP) at arrest (Dungar et al., 1982; Dunbar et al., 1983; Pikkarainen & Pentilla, 1980) . This enzyme is induced by alcohol and generally accepted as a reasonably sensitive indicator of excessive alcohol consumption. In this study we compared the advantages of GGTP over blood alcohol concentration (BAC) and compared the GGTP analyses with information retrieved from the drivers' medical records (Wilkins, 1974).
:k
Department of Community & Occupational Medicine, & Biochemical Medicine, Ninewells Hospital & Medical School, University of Dundee, Dundee DDI 9SY, Scotland, UNITED KINGDOM.
399
SUBJECTS AND METHODS Experiment I To stimulate the "Saturday night drinking driver" we invited 54 healthy drivers (53 men and 1 woman), aged 18 to 51 years (mean 33.3 years), to drink wine, spirits, and beer freely over 5 hours. Before drinking a baseline blood sample was taken; a 2nd sample was taken l h hours after the end of the drinking session and athird 8 hours after the session. Gammaglutamyl transpeptidase activity was measured in all 3 samples and blood alcohol concentration in the 2nd and 3rd samples. Experiment II Motorists who were arrested for drinking and driving in Tayside Region were invited to participate in Tayside Safe Driving Project; 96% of the drivers agreed to participate representing 51% ofthe drivers arrested in the Region. Of these, 96.6% of the drivers were male.The mean age was 32.5 years, with a range from 17 to 70 years. At the time of arrest 5 ml of blood were taken and divided into 3 samples: one was given to the driver, the second was analyzed for blood alcohol concentration by the public analyst, and the third was used to measure the serum activity of GGTP using a centrifugal fast analyzer (37°C). An examination of medical records to detect drugs or diseases known to raise GGTP activity was carried out but none was detected. A 2-page structured questionnaire was sent to the general practitioners of drivers who lived within Tayside. The results of this study are presented in another paper at this meeting. Returned questionnaires were given an alcohol rating score from "1" to "3." ("1" = alcohol problems stated to be present by general practitioners; "2" = 2 or more covert indicators of problem drinking in their medical record, e.g., gastritis, morning shakes, or marital difficulties; "3" = "clean" medical record with fewer than 2 covert indicators or illnesses/injury unassociated with alcohol.) Experiment III A sub-group of 128 subjects, chosen for geographic proximity, gave a second blood sample 5-9 months after arrets. Gammaglutamyl transpeptidase activity was measured again. (We found no statistically significant difference between this sub-group and the overall population of drunk drivers.)
400
RESULTS Experiment I Of the sample of 54 drivers, 45 had peak blood alcohol concentrations above the legal limit for driving (80 mg per 100 ml; 17.4 mmol/1) 1% hours after the drinking session, corresponding to an alcohol ingestion of greater than 100 g and, of these, 14 were still above the legal limit 8 hours after drinking. In these cases alcohol ingestion was greater than 300 g. No significant differences in GGTP activities were observed in 9 subjects with abnormal values (greater than 50 iu per liter). By 2-way analysis of variance we found significant variation between the times of testing, which indicated a difference in the mean GGTP activities on the 3 occasions, the main activities being 37. 3, 34. 9, and 36.5 iu per liter, respectively (S.E. of difference between 2 means = 0.54). The mean at 1h hours was about 5% lower than at base line and at 8 hours after ingestion (£ = less than 0.01). This was fairly small compared with a typical inter-individual variation which is of the order of + 30 units. Biochemically these differences were not important and may have been a reflection of hemodilution by alcohol; no differences were, however, observed with subjects consuming different amounts of alcohol. Neither overall mean GGTP activities nor variation in values over time depended on the amount of alcohol ingested (Table 1) . Table 1 Amount of Alcohol Ingested, Blood Alcohol Concentration and Gammaglutamyl Transpeptidase Activity (Mean Values Expressed + SD)
Mean blood alcohol
Alcohol ingested (g)
Nil
Mean
-glutamyltranspeptidase
mg/100 ml)* No of Subjects
After 6% hours
1
(IU/1) After
13 hours
0
0
After At start
6h hours
After 13 hours
20
19
21
51-100
4
46 + 49
10 + 21
35 + 21
31 + 18
35 + 21
101-150
6
111 + 66
25 + 45
50 + 57
48 + 57
48 + 53
151-200
12
140 + 32
14 + 15
25 + 9
22 + 9
24 + 10
201-150
9
162 + 72
43 + 47
39 + 27
36 + 27
40 + 29
152-300 301-400
10
195 + 60
56 + 43
39 + 42
37 + 41
37 + 38
10
238 + 40
108 + 58
39 + 29
37 + 30
38 + 28
401-500
2
275 + 4
115 + 13
59 + 35
57 + 42
58 + 35
* 1 mg/100 ml = 0.217 mmol/1.
401
Experiment II Of all the drivers arrested, incorporated in the first phase of the study, 24% had raised GGTP activity (Figure 1). Drivers with abnormal enzyme activity had blood alcohol concentrations throughout the entire range, including concentrations that were under the legal limit for driving. The correlation between blood alcohol concentration and enzyme activity at the time of arrest was highly significant, although the degree of association was only moderate (r = 0.22; £ less than 0.0001). Experiment III A comparison of serum GGTP activity at the time of arrest with that 5-9 months later, in those who co-operated in the follow-up study, showed a significant association (r = 0.88; £ less than 0.0001; Figure 2). In 3 subjects enzyme activity was high at arrest but in the reference range 5-9 months later; in 5 subjects enzyme activity was in the reference range at the time of arrest but was raised 5-9 months later; in 23 subjects (18%) enzyme activity was high at the time of arrest and continued to be high at follow-up. Of the subjects recalled for follow-up with blood alcohol concentrations over 17.6 mmol/1 (80 mg/100 ml), 25 (22%) had raised GGTP activity at the time of arrest; of these 88% were unchanged at follow-up (Figure 3) . Thirteen (45%) of the recalled subjects with blood alcohol concentrations over 43.6 mmol/1 (200 mg/100 ml), the UKsuggested index of problem drinking, had raised enzymeactivity at the time of arrest and 10 of these had raised activity at follow-up. General practitioners completed and returned 93% of the questionnaires distributed. Twenty-two (13%) had an alcohol rating of "1"; that is, the general practitioners were aware of the patient's alcohol problems. A further 42drivers (24%) had 2 or more covert indicators of problem drinking in their medical records. Alcohol rating and gammaglutamyl transpeptidase activity were significantly related but not blood alcohol concentrations and alcohol rating (Table 2). DISCUSSION Serum GGTP activity is generally considered to be a reasonably sensitive indicator of problem drinking (Whitehead et al., 1978); thus, our results indicate that most high risk offenders have blood alcohol concentrations in the range 17.6 - 43.2 mmol/1 (80-199 mg%). However, the problem drinkers may also be found among drivers with alcohol concentrations below the legal limit. Our results indicate that any attempt to identify high risk individuals 402
based on blood alcohol concentrations would be an arbitrary choice, likely to miss a large proportion of high risk offenders.
Table 2
Number of Drivers with an Alcohol Problem Rating of 1, 2, or 3 Against Blood Concentrations of Alcohol and Against -glytamyltranspeptidase Activity at Time of Arrest
Alcohol problem rating 1 No
2
(%)
3 No (%)
No
Total
(%)
No (%)
Blood alcohol concentrations (mg/100 ml)* 4(18)
10(24)
26(24)
40(23)
81-105
200
TOTAL
22(13)
42(24)
110(63)
174(100)
-glytamyltranspeptidase activity (IU/1)
< >
50
13(59)
33(79)
95(36)
141(81)
50
9(41)
9(21)
15(14)
33(19)
TOTAL
22(13)
42(24)
110(63)
Chi-squared = 1.70, df = 6, |> = greater than 0.0. Chi-squared = 9.09, df = 2, £ = less than 0.025. Conversion:
Traditional units to SI - Blood alcohol: 1 mg/lOOml
403
0.22 mmoi/1.
174(100)
Theoretical objections to GGTP have been largely eliminated by this study: there is no binge effect raising the enzyme over a short time course, and the prevalence of diseases and drugs amongst drunk drivers is so low that the problem of false positive does not occur. Indeed, in the larger Tayside Safe Driving Project study of 1047 drivers no cases of raised enzyme could be attributed to disease or drugs. In two further studies, the results of which are being presented at this meeting, it was shown that GGTP accurately reflects the spectrum of social and medical facets of problem drinking. Serum gammaglutamyl transpeptidase activity has an additional advantage over blood alcohol concentration. When measured more than 5 months after arrest it indicates whether or not the driver is controlling his drinking. Although it seems that conviction results in a restoration of normal activity, this occurs in very few drivers: in less than 10% of our subjects. It is far commoner to find continuing abnormal activity or increasing abnormality, indicating that conviction has not moderated the driver's alcohol consumption. Where there is sustained abnormal or rising enzyme activity, but no other known cause of disturbed liver function, it seems unwise to restore the driver's license. A quarter of convicted drivers in this study had raised serum GGTP activity. Given the reported sensitivity, specificity, and predicted value of the test, a third of drivers in the study may well be problem drinkers (Baxter et al., 1980). CONCLUSIONS Among drunk drivers serum GGTP activity measured at the time of arrest is a better indicator of problem drinking than blood alcohol concentration. ACKNOWLEDGMENT This study was funded by grants from the Biomedical Committee, Scottish Home and Health Department and the Association of Police Surgeons of Great Britain. We thank Chief Inspectors Ritchie, Nairn and Palmer; Dr. D. C. Marshall and the Police Surgeons; Professor Knox and the general practitioners of Tayside; Professors Griffith and Mair, Mrs. Bree, Brown, Kynoch, Montgomery and Reid, all of whom helped collect data for this study. Correspondence to Dr. J. A. Dunbar, 325 Strathmartine Road, Dundee DD3 8NE.
404
REFERENCES Baxter, S., Fink, R., Leader, A. R., and Rosalki, S. B. (1980). Laboratory Tests for Excessive Alcohol Consumption Evaluated in General Practice. British Journal on Alcohol and Alcoholism, 15: 164-66. Dunbar, J. A., Hagart, J. , Martin, G. E., et al. (1982). Drivers, binge drinking and gammaglutamyl transpeptidase. British Medical Journal, 285: 1083. Dunbar, J. A., Martin, B. T. , Devgun, M. E. , et al. (1983) . Problem drinking among drunk drivers. British Medical Journal, 286: 1319-1322. Pikkarainen, of arrested drunk (ed.), Alcohol, Almgvist & Wiksel
J. , and Pentilla, A. (1980). Screening drivers for alcoholism. In Goldberg, L. Drugs, and Traffic Safety. Stockholm: International.
Wilkins, R. (1974) . The Hidden Alcoholic in General Practice. London: Paul Elek Scientific Books. Whitehead, T. P., Clark, C. A., and Whitfield, A. G. W. (1978). Biochemical and haematological markers of alcohol intake. Lancet, 1: 978-781.
405
•
350-,
•
•
•
•
91
I
•
290
•
. 1subject • 2-5subjects
#
• 6-9subjects
g 200 ^
150 '
• ?> £
II00
*
'
•
• 50'
• •
••
••
•
•
*
•
•
•
•
•
•
• .i i(•••••I* •• • • • • •
-• •
I • • • • •* •_ • • • •
u it:s ••••••
* • • •
• •• •
•
•
• .
. •••
50 100 150 200 250 300 Blood alcohol concentrations (p h a s e ll mg/100ml
350
Figure 1 . Blood alcohol concentrations and gammaglutamyl transpeptidase activity in drivers at the time of arrest (Phase 1).
450
500
250 • 1 subject
200-1
• 2 - 5 s u b je c ts
• 6-9 subjects 150
100
50
■tl
•• • 50 B0 150 200 250 Serum T-giutomyltrcnsferase (phase 2) IU /I
300
350
Figure 2 . Serum gammaglutamyl transpeptidase activity at the time of arrest (Phase 1) and at follow-up (Phase 2).
406
3S0' ( p h o s e 2 H U /l S eriim 'y-glutom yltronsferose
• 1 subject
300'
200
• 2-5 subjects • 6-9 subjects
25CH
150'
100 50
• •• • • 50
CO
150
200
250
300
Blood alcohol concentrations ( phase 11 mg/100ml
Figure 3. Blood alcohol concentrations at the time of arrest and gammaglutamyl transpeptidase activity at follow-up (Phase 2).
407
