Drunk driving : a study of psychosocial, and health factors, focusing on young and middle-aged men in a metropolitan area

Abstract

Drunk driving is a common and serious alcohol-related problem in Sweden and many other countries. In Sweden 9,433 persons, of whom 90 % were men, were convicted of drunk driving in 1997. Surveys indicate that many more have driven under the influence of alcohol. The great majority of studies have focused on rather small groups of subjects convicted for drunk driving, or under treatment. Longitudinal prospective population studies have been missing. The general aim of this thesis was to study risk factors at young age for drinking and driving and also social and health problems associated with drunk driving in a longitudinal perspective.

Three sources of data were used:

* The conscription survey 1969/70 of all 8,122 men born 1950-51 from Stockholm County, enrolling for military service, linked to data on drunk driving, public drunkenness, risky driving, hospitalisation, disability pension, sickness absence, criminality, and mortality during a 23 year follow period. (Papers I-IV).

* Analysis of alcohol involvement at death in all 668 deceased persons aged 15-54 years in Stockholm in 1987, using information from death certificates, autopsy, medical and police records and sentences for drunk driving. (Paper V).

* Survey data from a random sample of subjects aged 18-54 years (n=3,064) in Stockholm County with reports of alcohol habits, drinking and driving, and drunk driving (Paper VI).

The analyses in Paper 1-11 showed that risky alcohol use, background in social class 11 or III (the lowest) and indicators of deviant behaviour reported at conscription predicted convictions for drunk driving over the whole 23 year follow up period, although they had a somewhat weaker influence during the latter part of this period. The first paper, with a 7 year follow up showed that the risk factor pattern was similar for convicted with low and high blood alcohol concentrations and also for those sentenced for public drunkenness. The risk for hospitalisation and mortality was significantly increased in multivariate analysis for those convicted for drunk driving (RR=1.5 and 3.4) and risky driving (RR=1.9 and 3.7), with higher risks for hospitalisation for suicide attempts, accidents, psychosis, narcotic and alcohol diagnoses. Drunk driving implied a significantly higher risk for disability pension and high levels of sickness absence. The analysis of those deceased in Stockholm in 1987 revealed that about 10 % of both men and women had an alcohol diagnosis as underlying cause of death. About 60 % of all men and about 30 % of all women, and over 80 % of those convicted for drunk driving had a suspected alcohol-related death. Altogether 28 % of the men and 10 % of the women, and 37 % of the men, and 17 % of the women aged 25-34 years in the Stockholm County sample reported drinking and driving, the majority not being high consumers.

The occurrence of drunk driving implies increased risk for accidents, and is also a signal of risky alcohol use, and often other problems. Each case must be considered seriously for secondary prevention and treatment, and the societal concern must not weaken. Risky driving is also associated with a high morbidity and mortality, and usually not discussed in relation to alcohol and narcotic use. In the efforts of preventing traffic accidents this might be useful knowledge.